Sustainable bioconversion of excess grape must into polyhydroxyalkanoates by Cupriavidus necator DSM 545 and Hydrogenophaga pseudoflava DSM 1034

TL;DR

This paper explores using surplus grape must to produce bioplastics (PHAs) with two bacterial strains, showing grape must is a sustainable and effective feedstock.

Contribution

The study is the first to investigate red and white grape musts as substrates for PHAs production by C. necator and H. pseudoflava.

Findings

C. necator DSM 545 achieved up to 61.5% PHB of cell dry weight in batch cultures using grape must.

H. pseudoflava DSM 1034 reached 67.9% PHB on grape must but performed poorly in fed-batch conditions.

Grape musts are promising feedstocks for sustainable PHAs production in circular economy frameworks.

Abstract

Wine production generates significant quantities of by-products each year, among which surplus grape must is notable for its high content of fermentable sugars, organic acids, and polyphenols. As wine consumption declines and grape must surpluses grow, identifying sustainable valorization strategies becomes increasingly critical. This study investigates for the first time the potential use of red and white grape musts as substrates for polyhydroxyalkanoates (PHAs) production by two well-characterized bacterial strains, Cupriavidus necator DSM 545 and Hydrogenophaga pseudoflava DSM 1034, under both batch and fed-batch fermentation regimes. Both musts supported microbial growth and PHAs accumulation. In batch cultures, C. necator DSM 545 achieved a PHB content of up to 61.5% of cell dry weight (CDW), while H. pseudoflava DSM 1034 reached 67.9% PHB on grape must, with yields and biomass comparable to or exceeding those obtained with synthetic sugar-based media. Under fed-batch conditions with red must, C. necator DSM 545 sustained growth and PHB production across multiple feeding feeding periods, outperforming the control medium. Conversely, H. pseudoflava DSM 1034 displayed initial growth but failed to increase biomass over time, suggesting that this strain may be poorly suited for use in fed-bacth-based applications, likely due to nutrient depletion or the accumulation of inhibitory compounds. Overall, these results proved that grape musts are promising feedstocks for sustainable PHAs production. Their integration into circular economy frameworks offers a valuable opportunity for waste recovery and the development of bioplastics within the agri-food industry, especially in light of the increasing grape must surpluses recently experienced worldwide.

Surplus red and white grape musts were tested as substrates for PHAs production

Batch cultures of C. necator DSM 545 and H. pseudoflava DSM 1034 accumulated PHB

Fed-batch setting confirmed C. necator DSM 545 performance with red must

H. pseudoflava DSM 1034 showed limited fitness to prolonged fed-batch conditions

Grape musts are promising, cost-effective feedstocks for PHAs production

Figures

Growth of *C. necator* DSM 545 (A) and *H. pseudoflava* DSM 1034 (B) in fed-batch mode. □, dotted line: MM + red grape must; ○, continuous line: MM + pure sugars (glucose and fructose). Results are the mean of three replicates ± standard deviation

Full text

Introduction

Growing environmental and economic concerns related to the availability of petroleum have driven the transition from conventional petroleum-based plastics to more sustainable, biodegradable alternatives. Conventional plastics pose serious environmental challenges due to their resistance to degradation and their persistent accumulation in natural ecosystems. It is estimated that millions of tons of plastic waste enter the oceans each year, contributing to marine pollution, harming wildlife, and disrupting ecological balance (Geyer et al. 2017; Jambeck et al. 2015). A significant amount of plastic waste is either incinerated or landfilled, worsening environmental problems due to its large volume, wide distribution, and non-biodegradable composition. These long-lasting materials also break down into microplastics, which are increasingly detected in food chains and even human tissues, raising additional concerns for both environmental and public health. Therefore, there is an urgent need to replace these “conventional” plastics with biodegradable and more sustainable polymers such as polyhydroxyalkanoates (PHAs). PHAs have garnered significant attention as a promising alternative to petroleum-based plastics due to their biodegradability, compostability, and biocompatibility (Sabapathy et al. 2020). The applications of these plastics included biomedical devices, pharmaceuticals, electronics, construction, automotive components, and packaging (Anjum et al. 2016; Sirohi et al. 2020). PHAs are synthesized by a variety of microorganisms as energy and carbon storage materials within their cytoplasm, especially under stressful conditions like an excess of available carbon or the limitation of one or more essential nutrients, such as nitrogen (Ganesh Saratale et al. 2021; Povolo et al. 2012). Beyond its primary role as a carbon and energy reserve, PHAs also contribute to increased cellular resistance against various environmental stresses, such as oxidative stress, nutrient limitation, and osmotic pressure (Obruca et al. 2021, 2018). Although microbial PHAs are commercially produced worldwide, their higher production costs compared with those of conventional plastics restrict their widespread application (Favaro et al. 2019; Rodríguez et al. 2021). Thus, for an environmentally friendly microbial PHAs synthesis and to achieve cost-effective large-scale production, multiple strategies can be pursued, including the selection of more efficient microbial strains or the optimization of process conditions. One potential solution could be the use of cheaper raw materials as substrates (Brojanigo et al. 2022; Ganesh Saratale et al. 2021; Kovalcik et al. 2020).

The winemaking process has gained much attention worldwide in recent years. In 2024, the global wine production was 226 million hectolitres (OIV 2025). Recent trends indicate a decline in wine consumption, driven by factors such as changing lifestyle habits, health concerns, and shifting market preferences (Anderson 2023; OIV 2025). Indeed, in 2024, wine consumption reached approximately 214 million hectoliters, representing a 3.3% decrease compared to 2023 (OIV 2025). This decline has led to an estimated surplus of 1,180,000 tons, which disproportionately affects major producers such as Italy, France, and Spain, and the situation appears likely to worsen in the coming years (OIV 2025).

As a consequence, there is a growing risk of surplus grape must, which may represent an underutilized by-product (OIV 2025). This trend is further reinforced by the increasing consumer preference for quality over quantity. In the case of alcoholic beverages, this shift has been particularly evident, with global recorded alcohol consumption peaking in the mid-1970s and subsequently declining. Wine consumption has shown the steepest decline, followed by spirits (Anderson and Pinilla 2017). These evolving consumption patterns suggest that the volume of unused or downgraded grape must is likely to increase, presenting a valuable opportunity for resource recovery through sustainable applications. Several agro-industrial residues, such as whey (Favaro et al. 2025; Koller et al. 2007), molasses (Omar et al. 2001; Santimano et al. 2009), lignocellulosic hydrolysates (Pan et al. 2012), and starchy (Brojanigo et al. 2020; Haas et al. 2008) have been previously exploited as low-cost substrates for PHAs production. Compared with these feedstocks, grape must provides a naturally sugar-rich matrix that requires minimal pretreatment, although its high acidity and phenolic compounds could impose limitations on microbial performance. These characteristics highlight surplus grape must as an attractive and largely untapped feedstock for PHAs biosynthesis. The valorization of this excess grape must through biotechnological processes, such as the production of PHAs, offers a sustainable alternative for waste management and resource recovery.

Grape must, the juice extracted from crushed grapes during the winemaking process, is complex mixture of water, fermentable sugars, organic acids (such as tartaric and malic acid), minerals, polysaccharides, proteins and phenolic compounds. When not fully absorbed by the wine market or redirected to alternative uses, must accumulates as surplus material that requires efficient and sustainable management. Nowadays, the excess grape must is typically managed through two conventional ways. It is either processed by the food industry for the production of grape juice or sweet wine intended for human consumption, or concentrated to increase its sugar content, yielding rectified concentrated must (RCM), commonly used for the natural enrichment of wines. Traditional winemaking practices, such as the addition of RCM and tartaric acid supplementation, are routinely employed to adjust the sugar and acid content in grape musts. In recent years, membrane-based technologies, including ultrafiltration, dialysis, and reverse osmosis, have emerged as innovative alternatives to conventional techniques (El Rayess et al. 2024; Notarnicola et al. 2015). These methods address key challenges in must and wine processing, such as stabilization, clarification, chemical composition adjustment, and alcohol content control, supporting the sustainability and the circular economy strategies in the wine sector. Although these practices offer a practical means of handling surplus must, they are limited in terms of economic return and do not fully exploit the substrate’s potential for value-added biotechnological conversion. In alignment with circular economy principles, grape must could be repurposed in biotechnological processes to produce valuable bioproducts such as PHAs, bioethanol, organic acids, and other bioproducts. These applications could contribute not only to reducing waste but also to fostering resource efficiency and sustainability within the agri-food sector (Basaglia et al. 2021; Favaro et al. 2019; Gupte et al. 2022).

This study aimed to evaluate the potential of excess grape must as a substrate for growth and PHAs production by two well-known bacterial strains: Cupriavidus necator DSM 545 and Hydrogenophaga pseudoflava DSM 1034.

Materials and methods

Microorganisms and inocula

C. necator DSM 545 and H. pseudoflava DSM 1034, wild-type strains supplied by DSMZ (German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany), were maintained at − 80 °C in 25% glycerol stocks. C. necator DSM 545, uv mutants obtained from C. necator H16 (Schlegel et al., 1965), and H. pseudoflava DSM 1034 can grow, among other carbon sources, on glucose, fructose, glycerol, and glucose, fructose, lactose and sucrose, respectively (Lopar et al. 2014; Povolo et al. 2013).

In all experiments, the bacterial inocula were prepared in 125 mL flasks containing 20 mL of Ramsey Mineral Medium (MM) (Ramsay et al. 1990) supplemented with 20 g/L glucose, by transferring a loop of cells grown on a Petri plate containing Nutrient agar (HI Media). The composition of the MM medium for the flask cultures included for each liter of distilled water: 6.7 g Na_2_HPO_4_·7H_2_O; 1.5 g KH_2_PO_4_; 1 g (NH_4_)2SO_4_; 0.2 g MgSO_4_·7H_2_O; 60 mg iron ammonium citrate; 10 mg CaCl_2_·2H_2_O; and 1 mL of trace elements solution. Trace elements solution contained per liter of distilled water: 0.3 g H_3_BO_3_; 0.2 g CaCl_2_·6H_2_O; 0.1 g ZnSO_4_·7H_2_O; 30 mg MnCl_2_·4H_2_O; 30 mg NaMoO_4_; 20 mg NiCl_2_·6H_2_O; 10 mg CuSO_4_·5H_2_O.

MM medium, trace elements and glucose solution were separately autoclaved at 121 °C for 20 min.

After 24 h at 30 °C under shaking, the bacteria were centrifuged at 3000 × g for 10 min, pellets were washed with sterile NaCl 0.9%. and re-suspended with the same solution. The bacterial suspensions were then added as inocula for the experiments described below to achieve an initial optical density of OD600nm = 0.250.

Grape musts

White and red grape musts from Vitis vinifera var. Glera and Merlot, respectively, were kindly provided by Coccitech SRL (Treviso, Italy). Grape musts were centrifuged for 10 min at 3000 × g, the pH of the supernatant adjusted to 6.5 using 5 M NaOH, then autoclaved at 121 °C for 20 min and added to the flasks for the experiments as described in paragraph 2.3 Batch and Fed-batch bacterial cultivations.

The sugar, ethanol, glycerol, and total soluble polyphenol contents of the must were determined using the methods described in paragraph 2.5 Chemical analyses.

Batch and fed-batch bacterial cultivations

C. necator DSM 545 and H. pseudoflava DSM 1034 growth and PHAs accumulation were first evaluated in red and white grape musts in batch mode. The growths were conducted in shaken 125 mL flasks containing 30 mL of MM, with white or red grape musts added to achieve a total sugar concentration (glucose and fructose) of 20 g/L. Every 12 h, samples were taken to analyze the residual sugar concentrations. After 72 h of growth, cells were separated by centrifugation, and the pellets were frozen at − 80 °C. PHAs in the bacterial pellet were then determined.

Fed-batch fermentations were conducted with red grape must in agitated 500 mL flasks with a working volume of 100 mL of MM. 10 mL of red grape must was added to reach a total sugar concentration (glucose and fructose) of 20 g/L. Every 48 h, the same volume was taken as sample for CDW, PHAs and residual sugar concentrations Subsequently, cultures were fed with red grape must corresponding to 10% of the working volume to restore the initial volume. Since the must mainly supplied carbon but was deficient in nitrogen, a nitrogen-source solution was supplemented to maintain approximately one-third of the nitrogen concentration present in the initial mineral medium (MM). Specifically, 1 mL of a 33 g/L concentrated nitrogen-source stock solution was added to the 100 mL culture after each feeding period, resulting in an effective nitrogen concentration of 0.33 g/L, corresponding to one-third of the initial 1 g/L nitrogen level in the MM, in accordance with fed-batch strategies previously described (Povolo et al. 2013).

Reference experiments were conducted in MM containing a solution 10 g/L glucose and 10 g/L fructose. Growth was monitored by measuring optical density at 600 nm (OD_600nm_) using a spectrophotometer (Spectronic Genesys™ 2PC, Thermo Fisher Scientific Inc., USA). At the end of the experiments, pellets were lyophilized to quantify cell dry weight (CDW): Falcon tubes were pre-weighed, wet biomass weight noted, and re-weighed after lyophilization to determine biomass. PHAs in bacterial pellets and residual sugar concentrations in the supernatants were then analysed as described below.

PHAs analysis

For the quantification of PHAs, approximately 10 mg of the lyophilized pellet was then subjected to methanolysis according to the method of Braunegg et al. (1978). In short, 1 mL of chloroform, 1 mL of 5 g/L benzoic acid in chloroform solution and 2 mL of methanol solution with 3% sulfuric acid were added to the pellets in glass vials sealed with a Teflon tape. After the incubation in the oven for 4 h at 100 °C, the organic phase was separated, and 3-hydroxybutyric acid (3HB), 4-hydroxybutyric acid (4HB), and 3-hydroxyvalerate acid (3HV) monomers were analyzed by gas chromatography. A Thermo Finnigan Trace GC was employed with a flame ionization detector (FID) and an AT-WAX column. Helium was used as the gas carrier at a 1.2 mL/min flow rate. The injection port, the detector and oven temperatures were set at 250 °C, 270 °C and 150 °C, respectively. Benzoic acid was used as the internal standard, whereas the external standards, poly(3-hydroxybutyric acid), poly(3-hydroxybutyric acid-co-3-hydroxyvaleric acid), and poly(3HB-co-11.2 mol% 4HB) were purchased from Sigma-Aldrich (Milan, Italy).

PHAs yields were calculated by dividing the amount (g) of produced PHAs and biomass by the amount (g) of consumed sugars.

Chemical analyses

Sugars, ethanol and glycerol content in the musts and supernatants of culture contents were determined by High Performance Liquid Chromatography (HPLC) using the methods described in (Costa et al. 2023; Gupte et al. 2024). Before injection, samples were filtered through an acetate cellulose 0.22 μm filter (Millipore). The analyses were performed using a Shimadzu Nexera HPLC system equipped with a RID-10A refractive index detector (Shimadzu, Kyoto, Japan). The chromatographic resolution was achieved using a Phenomenex Rezex ROA-Organic Acid H + (8%) column (300 mm × 7.8 mm). The column temperature was set at 60 °C, and the analysis was performed at a flow rate of 0.6 mL/min using isocratic elution with 5 mM H_2_SO_4_ as the mobile phase.

Total soluble polyphenols were measured using the Folin-Ciocalteu method as described in (Caminiti et al. 2025) and were reported as gallic acid equivalent per liter (mg GAE/L).

Results and discussions

Grape must chemical characterization

As reported in Table 1, the composition of the musts was consistent with that reported in the scientific literature (Barata et al. 2011): as expected, the main sugars here identified were glucose and fructose, present at similar concentrations in both juices, with slightly higher levels observed in the white must, where a small amount of ethanol was also detected. This likely indicates the onset of fermentation, further supported by the simultaneous occurrence of glycerol. The total polyphenol content (expressed as mg GAE/L) was comparable to values reported for grape musts (Wang et al. 2016), with higher levels in the red (318.08 ± 15.79 mg GAE/L) compared to the white must (238.65 ± 14.54 mg GAE/L). This finding is consistent with expectations due to the greater phenolic extraction from red grape skins. During winemaking, only a fraction of the polyphenols is transferred from grapes to wine. This strongly depends on how well the liquid must comes into contact with the solid parts of the grape bunches, as well as on the specific grape variety and its degree of ripeness. Additionally, microbiological, chemical, and physical factors can further alter both the structure and concentration of polyphenols during the following fermentation, clarification, and aging (Giovinazzo and Grieco 2015).Table 1. Chemical characterization of red and white grape must. Results of sugars, ethanol and glycerol are reported in g/L. Total soluble polyphenols are reported in gallic acid equivalent per liter (mg GAE/L)Red grape mustWhite grape mustGlucose79.57 ± 4.4686.59 ± 5.64Fructose92.38 ± 4.7799.34 ± 6.50Ethanolnd1.56 ± 0.11Glycerolnd0.42 ± 0.02Total soluble polyphenol318.08 ± 15.79238.65 ± 14.54nd, non-detectable

Batch cultivation

Autoclaved red and white grape musts were tested for C. necator DSM 545 and H. pseudoflava DSM 1034 in batch mode to evaluate their suitability as growth substrates (Figs. 1 and 2). Sterilization of grape must can induce precipitation of tannins and other colloidal materials. In this study, a slight formation of particulate matter was observed visually after autoclaving, but it did not seem to interfere with biomass measurements. To verify this, non-inoculated controls were subjected to the same sterilization procedure in each experiment, the sterilized broth was centrifuged, and the resulting pellet was processed identically to the inoculated samples. The dry weight of these controls was negligible, confirming that the autoclave-induced precipitate did not contribute to the measured cell mass. Therefore, separation of supernatant and particulate phases before inoculation was unnecessary.Fig. 1C. necator DSM 545 growth curve and glucose and fructose consumption on A MM amended with red grape must, B MM amended with white grape must, and C MM containing glucose and fructose solution. ●: OD_600nm_; ▲: glucose; ■: fructose. Results are the mean of three replicates ± standard deviationFig. 2H. pseudoflava DSM 1034 growth curve and glucose and fructose consumption on A MM amended with red grape must, B MM amended with white grape must, and C MM containing glucose and fructose solution. ●: OD_600nm_; ▲: glucose; ■: fructose. Results are the mean of three replicates ± standard deviation

As a reference, growth was also performed in MM containing 20 g/L of sugars (10 g/L of glucose and fructose, respectively) (Figs. 1C and 2C). A sugar concentration of 20 g/L was selected as the experimental reference parameter because it is widely recognized as an appropriate and standardized flask condition for both C. necator DSM 545 and H. pseudoflava DSM 1034. This choice ensures reproducible and well-controlled growth dynamics while enabling direct comparison with previously published studies (Povolo et al. 2013; Favaro et al. 2025).

Both red and white grape musts successfully supported the growth of C. necator DSM 545 (Fig. 1A, B) and H. pseudoflava DSM 1034 (Fig. 2A, B), demonstrating their suitability as substrates for microbial cultivation.

C. necator DSM 545 and H. pseudoflava DSM 1034 reached the stationary phase after 48–50 h. The pH of the media, measured at the end of the experiments, consistently ranged between 6.7 and 6.9, indicating that the culture medium possessed sufficient buffering capacity to maintain stable pH conditions throughout microbial growth. Furthermore, both bacteria demonstrated enhanced growth in media supplemented with grape musts compared to the control: although the growth trends in the musts were comparable to those obtained in MM and sugars, the final OD_600_ monitored in the musts were always slightly higher than those detected in the benchmark MM supplemented with glucose and fructose. These evidences suggest that grape musts might probably contain additional components likely derived from the plant that enhance bacterial growth such as macroelement (Ca, K e Mg), microelement (Mn, Cu e Zn), vitamins, amino acids as reported by Richardson et al. (2021).

In this study, the polyphenols don’t seem to exert inhibitory effects under the tested conditions. It is worth noting that high concentrations of polyphenols can have antimicrobial effects, potentially inhibiting the growth and metabolic activity of certain microorganisms involved in fermentation (Coppo and Marchese 2014; Daglia 2012).

Throughout cultivation, glucose and fructose concentrations were monitored. As expected, in grape musts and benchmark experiments, nearly all the glucose was consumed during the logarithmic growth phase by both microorganisms. Remarkably, they were able to utilize glucose and fructose simultaneously. However, after 72 h of cultivation, residual fructose was still detectable in the culture supernatants from both C. necator DSM 545 and H. pseudoflava DSM 1034 cultures. Specifically, exhausted broths of C. necator DSM 545 contained residual fructose concentrations of 1.64 g/L in red grape must and 3.56 g/L in white grape must. Under the same conditions, H. pseudoflava DSM 1034 broths showed residual fructose levels of 0.57 g/L and 3.09 g/L, respectively. These findings suggest that both microorganisms preferably utilized glucose over fructose.

The metabolism of C. necator DSM 545 is extremely versatile, and the bacterium can utilize a broad spectrum of substrates for growth, including carbohydrates (Costa et al. 2022; Povolo et al. 2010), organic acids (Caminiti et al. 2025; Liu et al. 2016), glycerol (Dimou et al. 2015), aromatic compounds, and even carbon dioxide (Bellini et al. 2022; Costa et al. 2023). The preference for glucose over fructose could be attributed to the higher efficiency of glucose transport systems and the direct integration of glucose into the glycolysis pathway. In contrast, although metabolizable, fructose requires preliminary isomerization or phosphorylation steps that reduce efficiency, leading to slower assimilation. As a result, glucose is typically consumed more rapidly, while fructose tends to persist during mixed-sugar fermentations (Bellini et al. 2024, 2022). Similar consideration could be attributed to H. pseudoflava DSM 1034 (Auling et al. 1978; Koller et al. 2007).

After 72 h of incubation, CDW and PHB content were determined (Table 2). No 4HB or 3HV were detected in either bacterium. The absence of copolymer units clearly indicates that the musts did not provide suitable metabolic precursors for the biosynthesis of these monomers, thereby restricting polymer formation to the homopolymeric PHB.

Table 2CDW, PHB (A), biomass and PHA yields and productivity (B) obtained with C. necator DSM 545 and H. pseudoflava DSM 1034 after 72 h in red and white grape musts and sugars solution (glucose + fructose)ACDW (g/L)PHB (g/L)PHB (%)C. necator DSM 545 MM + red grape must8.23 ± 0.084.90 ± 0.7759.50 ± 8.80 MM + white grape must7.42 ± 0.184.56 ± 0.1761.48 ± 0.81 MM + 20 g/L glucose + fructose7.12 ± 0.084.20 ± 0.3358.99 ± 3.94H. pseudoflava DSM 1034 MM + red grape must7.45 ± 0.224.73 ± 0.2363.46 ± 1.23 MM + white grape must8.28 ± 0.035.62 ± 0.0267.87 ± 0.00 MM + 20 g/L glucose + fructose6.98 ± 0.025.23 ± 0.6274.94 ± 3.09BPHB yields (g PHB/gconsumed sugar)Biomass yields (g biomass/gconsumed sugar)Biomass productivity(mg/L h)Polymer productivity(mg/L h)C. necator DSM 545 MM + red grape must0.2660.44811468 MM + white grape must0.2810.45811363 MM + 20 g/L glucose + fructose0.2520.4279958H. pseudoflava DSM 1034 MM + red grape must0.2430.38310366 MM + white grape must0.3340.49311578 MM + 20 g/L glucose + fructose0.3030.4049773Results are the mean of three replicates ± standard deviation

C. necator DSM 545 achieved PHB content of 61.48% of the CDW and a PHB concentration of 4.56 g/L with white grape must; slightly higher results were observed for red grape must, thanks to a higher biomass production.

Noteworthy, the benchmark broth supplemented with pure glucose and fructose prompted slightly lower biomass as well as PHB production (Table 2).

H. pseudoflava DSM 1034 accumulated PHB up to 67.87% of CDW with a PHB concentration of 5.62 g/L in white grape must and 63.46% of CDW with a PHB concentration of 4.73 g/L in red grape must. This strain performed better in white grape must in terms of both CDW and PHB production, although the mineral medium containing sugars yielded a higher PHB percentage.

As reported in Table 2, the yields in terms of PHB per g of consumed sugars using grape must as a substrate were higher than those previously reported for the same microbial strains grown on wine lees in flasks and under similar conditions (Caminiti et al. 2025), highlighting the superior effectiveness of grape must in supporting both cell growth and biopolymer accumulation. Biomass yields per g of consumed sugars were comparable to those reported in the same study. Overall, both findings indicate that grape must supports PHB accumulation more effectively than wine lees.

These results are comparable to those obtained in flasks by other authors, such as Nygaard et al. (2021), who optimized PHB production by C. necator ATCC 17697 in batch bioreactor mode using pure fructose and ammonium sulfate. Similar polymer content but slightly lower biomass concentrations compared to those presented in this study were previously reported from sucrose by H. pseudoflava DSM 1034 (Povolo et al. 2013), whereas the use of lactose, even with yeast extract supplementation, resulted in lower biomass and PHB. Instead, H. pseudoflava ATCC 33688, on a hardwood hydrolysate containing glucose and xylose, accumulated up to 84.0% PHAs after 72 h, with a final PHAs concentration of 5.44 g/L (Blunt et al. 2023).

Fed-batch cultivation

Based on the promising results obtained in batch mode with both strains, a fed-batch strategy, simulating microbial growth conditions more similar to those in a bioreactor, was then adopted using red grape must which supported the highest PHB levels. The growth was monitored for 192 h and, as a benchmark, MM containing glucose and fructose was also evaluated. The chosen cultivation period was based on the observation that both strains reached the stationary phase approximately after at least 48 h (Figs. 1 and 2). We selected a longer cultivation period based on the understanding that the strains used, such as C. necator and H. pseudoflava, primarily accumulate the polymer during the stationary phase. Feeding with must or the sugar solution was carried out every 48 h, which allowed to provide sufficient substrate for growth, since even by the second feeding period, some sugars remained unconsumed in the medium. Similar to the flask experiments, the pH of the spent medium remained largely unchanged, staying around 6.8.

Grape must supported the growth of C. necator DSM 545 better than the reference medium (Fig. 3A). Throughout the cultivation process, OD_600_ showed a continuous upward trend, culminating in a final value of 66 after the fourth feeding period. In contrast, in the benchmark condition, growth plateaued after the second feeding period, with OD_600_ stabilizing at approximately 33. This limited growth is likely attributable to nutritional constraints inherent in the synthetic reference medium. The consistently increasing OD_600_ values of C. necator DSM 545 in grape must suggest that this strain greatly benefits from the complex composition of the substrate. The presence of amino acids, vitamins, minerals, and potentially growth-promoting compounds in grape must may have supported not only initial growth but also sustained biomass accumulation over multiple feeding periods. In contrast, the plateau observed in the reference medium after the second feeding period indicates that, despite sufficient carbon sources, the limited nutritional complexity may have imposed metabolic bottlenecks. This suggests that nutrient-rich, complex media like grape must, compared to minimal medium, can better maintain high cell density in prolonged cultivation and support the full biosynthetic potential of C. necator DSM 545.

Fig. 3. Growth of C. necator DSM 545 (A) and H. pseudoflava DSM 1034 (B) in fed-batch mode. □, dotted line: MM + red grape must; ○, continuous line: MM + pure sugars (glucose and fructose). Results are the mean of three replicates ± standard deviation

Fed-batch conditions are inherently more stressful, and repeated feeding periods can lead to the accumulation of metabolic by-products with inhibitory or toxic effects. From the second feeding period onward, a slowdown in fructose consumption, and to a lesser extent glucose, was already evident in the must. A similar pattern was observed in the simple sugar solution, where fructose, despite being the preferred carbon source, was utilized less efficiently starting from the second feeding period.

In red grape must, C. necator consumed a total of 56.84 g of sugars (20.80 g fructose + 33.84 g glucose). Under the same operating mode but with the simple sugar solutions, only 39.48 g of sugars were metabolized. This higher overall sugar consumption explains the greater biomass formation observed with grape must, which evidently provides a more supportive nutritional environment that enhances both growth and PHA accumulation in C. necator.

In contrast, Hydrogenophaga pseudoflava DSM 1034 exhibited substantially lower robustness under the same fed-batch conditions and underwent a rapid initial growth phase during the first 48 h in both media (Fig. 3B), reaching an OD₆₀₀ of approximately 19 by the end of the first feeding period. After the first feeding period, its capacity to metabolize both glucose and fructose declined sharply, indicating a limited tolerance to the increasingly stressful environment. Thereafter, on pure sugars, growth entered a relatively stable phase with no significant further increase in OD₆₀₀ values. A similar trend was also observed on grape must but with a marked decline from the second feeding period. The drop in OD₆₀₀ in grape must may even point to sensitivity to certain compounds that accumulate over time in natural substrates or to nutrient imbalances affecting cell viability or division. Alternatively, H. pseudoflava DSM 1034 may have a less robust stress response, making it less suited for extended cultivation under fed-batch conditions, particularly in a complex medium like grape must. Possible reasons could also include the limited adaptability to prolonged exposure to complex substrates, or a more restricted metabolic flexibility compared to C. necator DSM 545.

Regarding the possibility that residual sugars accumulated in the medium during fed-batch cultivation of H. pseudoflava DSM 1034 and reached inhibitory concentrations, additional fed-batch experiments were performed (data not shown). In these experiments, cells were centrifuged and re-suspended in fresh medium at the end of each feeding period, preventing the accumulation of sugars or other metabolites in the supernatant. The biomass concentrations obtained were comparable to those achieved in the standard fed-batch runs. This indicates that elevated sugar concentrations were not responsible for the limited growth of H. pseudoflava DSM 1034. Taken together, these results allow to rule out sugar inhibition as a major factor. Rather, the data suggest that H. pseudoflava DSM 1034 is intrinsically less tolerant to the physiological stress associated with repeated feeding periods, which likely explains its markedly lower biomass accumulation relative to C. necator DSM 545 under fed-batch conditions.

The different behavior of the two strains highlights important differences in the metabolic capabilities and nutritional requirements of C. necator DSM 545 and H. pseudoflava DSM 1034, particularly under repeated fed-batch conditions. In addition, it underlines the different sensitivity to environmental stressors of the two evaluated strains and the importance of fermentation strategy and medium composition in sustaining long-term microbial activity.

At the end of every feeding period, every 48 h of cultivation, CDW and biopolymer contents were determined for both bacteria (Tables 3 and 4). These data align with the observations resulting from Fig. 3, confirming the reported growth trends. Indeed, for C. necator DSM 545 (Table 3), both CDW and PHB concentrations in the reference medium remained relatively stable along the four feeding periods, with only a slight increase at the end in CDW and PHB. In contrast, when grown on red grape must, all the parameters rose along with time. Indeed, CDW in the sugar solution ranged from 7.44 g/L to a peak of 9.36 g/L at 144 h, followed by a slight decline at 192 h. On the contrary, in red grape must, CDW steadily increased from 8.35 at 48 h to 20.67 g/L at 192 h. A similar trend was observed for PHB in terms of both concentration and accumulation.Table 3CDW and PHB accumulated by C. necator DSM 545 during the fed-batch with red grape must and sugars (glucose + fructose) as a benchmark. Results are the mean of three replicates ± standard deviationCDW (g/L)Time (h)MM withsugarssolutionMM withred grapemustFeeding period 1487.44 ± 0.288.35 ± 0.04Feeding period 2969.35 ± 0.1012.61 ± 0.07Feeding period 31449.36 ± 0.2917.00 ± 0.01Feeding period 41928.79 ± 0.2420.67 ± 0.43PHB (g/L)Time (h)MM withsugarssolutionMM withred grapemustFeeding period 1486.88 ± 0.017.83 ± 0.39Feeding period 2966.84 ± 0.069.79 ± 0.05Feeding period 31448.26 ± 0.3513.74 ± 0.69Feeding period 41928.27 ± 0.5318.03 ± 0.21PHB (%)Time (h)MM withsugarssolutionMM withred grapemustFeeding period 14892.55 ± 3.6193.80Feeding period 29673.15 ± 1.4477.60 ± 1.34Feeding period 314488.20 ± 0.9980.83 ± 4.09Feeding period 419294.00 ± 5.6687.25 ± 2.62Table 4CDW and PHB produced by H. pseudoflava DSM 1034 during the fed-batch with red grape must and sugars (glucose + fructose) as a benchmark. Results are the mean of three replicates ± standard deviationCDW (g/L)Time (h)MM withsugarssolutionMM withred grapemustFeeding period 1485.54 ± 0.205.99 ± 0.06Feeding period 2964.92 ± 0.285.75 ± 0.13Feeding period 31444.70 ± 0.296.88 ± 0.23Feeding period 41926.06 ± 0.165.73 ± 0.43PHB (g/L)Time (h)MM withsugarssolutionMM withred grapemustFeeding period 1484.12 ± 0.043.63 ± 0.04Feeding period 2963.40 ± 0.302.61 ± 0.33Feeding period 31442.52 ± 0.043.65 ± 0.38Feeding period 41923.62 ± 0.523.32 ± 0.13PHB (%)Time (h)MM withsugarssolutionMM withred grapemustFeeding period 14874.40 ± 1.8960.64 ± 1.33Feeding period 29669.06 ± 2.2945.27 ± 4.66Feeding period 314453.66 ± 2.4052.97 ± 3.73Feeding period 419259.97 ± 4.1458.00 ± 2.03

Conversely, H. pseudoflava DSM 1034 displayed a different trend. On both broths, CDW and PHB declined after the first feeding period. In grape must, H. pseudoflava DSM 1034 exhibited high CDW values after the first feeding period, which remained relatively stable throughout the process reaching a maximum value of 6.88 g/L at 144 h. PHB content slowly decreased along the process, dropping from 3.63 g/L at 48 h to 3.32 g/L at 192 h in grape must. Similar behavior was found with sugar solution.

While for C. necator DSM 545 only 3HB was found, for H. pseudoflava DSM 1034, a small amount of 3HV, ranging from 1 to 3%, was detected with both substrates (data not shown).

During the fed-batch cultivation, the consumption of glucose and fructose by C. necator DSM 545 and H. pseudoflava DSM 1034 was also monitored (Table 5A and B). At the end of the four feeding periods, C. necator DSM 545 metabolized 39.48 g/L of sugars (19.19 g/L of glucose and 20.29 g/L of fructose) when cultivated on pure sugar solution, and 56.64 g/L of sugars (33.84 g/L of glucose and 20.80 g/L of fructose) when grown with the grape must. With C. necator DSM 545, despite the efficient sugar consumption in the first feeding period, residual glucose and fructose gradually accumulated over time, suggesting that the amount of sugars supplied exceeded the strain’s ability to consume them efficiently. This effect was observed with both pure sugar solutions and grape must, although it was less pronounced in the latter, likely due to its more balanced nutrient composition, which may have facilitated efficient substrate assimilation.Table 5. Glucose and fructose monitoring at the end of each feeding period during the fed-batch cultivation of (A) C. necator DSM 545 and (B) H. pseudoflava DSM 1034 in sugars solutions and red grape must, (C) biomass and PHB yields at the end of growth(A) C. necator DSM 545Glucose (g/L)Fructose (g/L)Time (h)MM with sugars solutionMM with red grape mustMM with sugars solutionMM with red grape mustFeeding period 1480.89 ± 0.13 (9.11)0.55 ± 0.05 (9.45)2.53 ± 0.28 (7.47)2.86 ± 0.42 (7.14)Feeding period 2967.06 ± 0.01 (3.83)3.29 ± 0.08 (7.26)7.98 ± 0.25 (4.55)9.19 ± 0.01 (3.67)Feeding period 314413.22 ± 1.07 (3.84)4.27 ± 0.72 (9.02)13.05 ± 0.86 (4.93)13.17 ± 0.67 (6.02)Feeding period 419620.81 ± 0.01 (2.41)6.16 ± 0.55 (8.11)19.71 ± 0.10 (3.34)19.20 ± 3.29 (3.97)Total consumed sugars (g/L)19.1933.8420.2920.80(B) H. pseudoflava DSM 1034Glucose (g/L)Fructose (g/L)Time (h)MM with sugars solutionMM with red grape mustMM with sugars solutionMM with red grape mustFeeding period 1482.64 ± 0.24 (7.36)4.74 ± 0.08 (5.26)nd (10.00)1.15 ± 0.21 (8.85)Feeding period 2965.38 ± 1.72 (7.26)13.39 ± 0.14 (1.35)7.57 ± 0.66 (2.43)9.94 ± 0.16 (1.21)Feeding period 31447.86 ± 1.85 (7.52)21.94 ± 0.09 (1.45)14.93 ± 1.44 (2.64)18.85 ± 0.02 (1.09)Feeding period 419610.94 ± 1.00 (6.92)28.99 ± 0.43 (2.95)22.94 ± 0.50 (1.99)26.45 ± 0.18 (2.40)Total consumed sugars (g/L)29.0611.0117.0613.55(C)Biomass yields (g biomass/g consumed sugar)PHB Yields (g PHB/g consumed sugar)C. necator DSM 545 MM with sugars solution0.2230.209 MM with red grape must0.3780.330H. pseudoflava DSM 1034 MM with sugars solution0.1310.078 MM with red grape must0.2330.135nd, non detectableResults are the mean of three replicates ± standard deviation. The value in parentheses indicates the amount (g/L) of consumed substrate during the feeding period

In contrast, as expected, H. pseudoflava DSM 1034 exhibited lower sugar utilization effectiveness on both substrates. In both media, glucose and fructose concentrations rose sharply in the later feeding periods, suggesting that the feeding strategy consistently outpaced H. pseudoflava’s metabolic uptake.

Biomass and PHB yields were also assessed at the end of the process (Table 5C). For C. necator DSM 545, the use of red grape must resulted in significantly higher PHB (0.330 g/g) yields compared to the pure sugars solution (0.209 g/g), highlighting its potential as a more efficient carbon source. Moreover, the fed-batch strategy, compared with the batch (Table 2), promoted PHB accumulation over biomass growth, as reflected by yields in both substrates, indicating that a greater proportion of substrate was channelled toward polymer production rather than cell proliferation.

Although very promising, these results indicate that for C. necator DSM 545, the fed-batch strategy and the use of grape must need further adjustments and several improvements could be explored. Firstly, fine tuning the feeding rate to better match the strain’s metabolic capacity could prevent excess substrate build-up. Additionally, implementing real-time monitoring of sugar concentrations would allow dynamic adjustment of nutrient supply. The optimization of the carbon-to-nitrogen (C/N) ratio may also enhance both cell growth and PHB accumulation. Finally, ensuring that the inoculum is well-adapted and metabolically active at the start of the process, could further improve substrate utilization efficiency.

On the contrary, biomass and PHB yields of H. pseudoflava DSM 1034 in fed-batch mode, were significantly lower compared with those in batch. Therefore, the applied feeding strategy did not prove suitable for this strain. Once again, these findings suggested that H. pseudoflava DSM 1034 was less suited than C. necator DSM 545 for sustaining biomass growth and PHB accumulation over multiple feeding periods under fed-batch conditions, or that alternative process optimizations might have been required for improved performances. This observation is further supported by the apparent absence of commercial-scale processes utilizing fed-batch or continuous cultivation with H. pseudoflava as reported in (Koller et al. 2007; Koller and Mukherjee 2022). One possible explanation could be that the stressful conditions within the fermenter could not be favorable for the growth of this microorganism, indicating that H. pseudoflava may not be a suitable candidate for biotechnological applications in such industrial systems.

In contrast, C. necator is widely employed in industrial PHAs production processes (Koller and Mukherjee 2022), due to its robustness and efficiency under continuous or fed-batch conditions. A fed-batch fermentation strategy with C. necator DSM 545 has been effectively optimized in previous studies (Bellini et al. 2024), highlighting its suitability for PHB production: the process achieved up to 10 g/L of PHB (83% of CDW) in just 48 h with sugar-based syrup (SBS) from cereal waste and acetate-rich waste medium (AWM) from gas fermentation. Even with SBS alone, substantial yields were obtained, confirming the strain's strong performance under industrial conditions.

Conclusions

This work demonstrates that surplus red and white grape musts from the winemaking industry are effective, low-cost carbon sources for microbial PHB production. Both C. necator DSM 545 and H. pseudoflava DSM 1034 grew efficiently and accumulated PHAs in batch cultures. H. pseudoflava DSM 1034 displayed particularly high PHB yields in white must, while C. necator DSM 545 maintained consistent productivity across both matrices.

In fed-batch mode with red must, C. necator DSM 545 sustained biomass and PHAs accumulation over four feeding cycles, outperforming the control.

Overall, the study confirms the technical feasibility of transforming winemaking surpluses into valuable biopolymer precursors, positioning grape must as a promising substrate within circular bioeconomy frameworks.

The growing availability of off-market musts driven by evolving consumer preferences amplifies the potential of this approach. Additional unconventional sugar sources, such as Botrytis-affected musts and deteriorated table grapes, further expand the spectrum of wine-related feedstocks for PHAs.

Process optimization and scale-up studies will be essential to fully exploit the potential of grape must–based feedstocks: given that the present results were obtained in flask experiments, significantly different and potentially improved outcomes may be achieved under controlled bioreactor conditions, including higher biomass accumulation, improved PHAs yields, and shorter cultivation times. Future research should strongly focus on controlled bioreactor-based fermentations, where key operational parameters such as pH, dissolved oxygen, feeding rate, and substrate concentration can be precisely regulated. Tailored feeding strategies, pre-treatments, or co-substrate supplementation could be systematically evaluated in fermenters to enhance robustness, productivity, and process reproducibility. Such conditions would allow a more accurate assessment of strain-specific physiological responses, particularly under fed-batch regimes, and help disentangle substrate-related effects from process-induced stress.

Overall, the valorization of grape must by-products for PHAs production represents a promising strategy to reduce agri-food waste and decrease reliance on fossil-derived plastics, contributing to climate mitigation and the advancement of circular agri-food systems.