Efficient Yeast Inactivation and Protein Extraction from Wine Lees Using Pulsed Electric Fields and Ultrasound: A Comparative Energy-Based Approach

Abstract

The valorization of wine lees, a major by-product of winemaking, is gaining attention as part of broader initiatives to promote circular economy and sustainable resource use in the agri-food sector. This study assessed ultrasound (US), pulsed electric fields (PEF), and their combination (PEF + US) as non-thermal technologies for promoting yeast autolysis and recovering intracellular proteins from wine lees. All treatments effectively reduced yeast viability, with populations decreasing from ~7.0 to ~4.7 log CFU/mL within 5 min. PEF alone achieved this microbial inactivation with a low energy input of 25–100 kJ/kg. In contrast, US yielded the highest protein release (~5700 μg/mL after 20 min), although it required a substantial energy input (~19,800 kJ/kg). The combined PEF + US method provided comparable protein yields (~5400 μg/mL) while reducing energy consumption by more than 50%. These results demonstrate that PEF is the optimal method for low-energy microbial inactivation, whereas US and PEF + US are more effective for protein recovery. The synergy of the combined approach offers a balanced and scalable solution for sustainable bioprocessing, reinforcing the potential of hybrid technologies in the green extraction of wine industry by-products and their integration into circular bioeconomy strategies.

1. Introduction

The global wine industry is a significant contributor to agro-industrial waste, generating approximately 20 million tons of by-products annually, equating to about 30% of the total vinified grape mass [1]. Among these by-products, wine lees represent a significant portion, accounting for approximately 25% of the total waste generated during the winemaking process [2,3]. Wine lees are the sediment formed at the bottom of fermentation tanks, primarily composed of dead yeast cells, bacteria, tartaric acid, phenolic compounds, proteins, and other organic materials [4,5,6]. Traditionally considered a low-value waste stream, wine lees represent a potentially rich source of bioactive compounds such as peptides, mannoproteins, and β-glucans, which are increasingly recognized for their technological, cosmetic, and nutritional value [1,2,3]. In this context, the development of innovative strategies to convert wine lees into high-value ingredients aligns with global sustainability goals and responds to the growing demand for functional bioproducts.

A key step in the valorization of wine lees is the induction and control of yeast autolysis. Autolysis refers to the self-degradation of yeast cells, leading to the release of intracellular and cell wall-associated compounds. Beyond its oenological significance, yeast autolysis is also a promising method for recovering proteins and other valuable biomolecules for food and biotechnological applications. However, under conventional cellar conditions, autolysis is a slow process that may extend over several months, limiting its practical efficiency and scalability [7,8].

To address this limitation, several physical, chemical, and enzymatic methods have been proposed to accelerate or enhance yeast autolysis and thus the release of intracellular components. Among these, pulsed electric fields (PEF) and ultrasound (US) are emerging as promising, non-thermal technologies for improving the yield and selectivity of compound extraction from yeast biomass [9,10,11,12]. PEF induces membrane permeabilization (electroporation) by applying short bursts of high-voltage electric fields, thereby facilitating the release of intracellular components [13]. US, on the other hand, relies on acoustic cavitation and shear forces to mechanically disrupt cells and promote mass transfer [14,15]. Both techniques have shown potential in food and biotechnology contexts, and recent research has highlighted their capacity to enhance autolysis in Saccharomyces cerevisiae by triggering endogenous enzymatic activity and improving cell wall permeability [10,11,12,15,16]. Importantly, these methods are compatible with green processing principles and offer opportunities for the energy-efficient recovery of functional compounds from wine lees [17,18,19]. In recent years, both ultrasound and PEF have increasingly been adopted in industrial applications due to their relatively low cost and environmentally friendly nature. Since both technologies are already established in practice, the development of protocols that combine them may offer additional benefits for industrial use [20,21]. To the best of our knowledge, this is the first study to investigate the combined application of PEF and US specifically for yeast autolysis in wine lees, while simultaneously providing a comparative energy-based evaluation that is essential for assessing industrial feasibility and sustainable by-product valorization.

Despite the recognized potential of PEF and US for enhancing yeast cell disruption, few studies have systematically evaluated their combined effect on wine lees autolysis, particularly in relation to protein release, yeast inactivation, and process energy efficiency. Therefore, the aim of this study was to develop and assess a protocol that integrates pulsed electric fields and ultrasound treatments to promote yeast autolysis in wine lees, with the goal of improving protein extraction yields while minimizing energy consumption and preserving process sustainability.

2. Materials and Methods

2.1. Wine Lees Production

In this study, three commercial S. cerevisiae strains (S1, S2, and S3) were used. S. cerevisiae S1 strain is proposed as a starter for secondary fermentation, while both S. cerevisiae S2 and S3 are widely used as inocula by Greek winemakers on various cultivars. All yeast strains have been clarified and preserved at −20 °C in YPD broth supplemented with 30% glycerol (composition per liter: 10 g yeast extract, 20 g bacteriological peptone, 20 g dextrose). Prior to experimentation, the strains were incubated twice on YPD agar at 28 °C for 48 h. Subsequently, pre-cultures were grown in YPD broth at 28 °C for 48 h. The medium for laboratory fermentations was a Moschofilero must, with a pH of 3.7 ± 0.2, a sugar concentration of 190 ± 0.8 g/L, an available nitrogen of 195 ± 0.3 mg N/L, and a total acidity of 5.6 ± 0.1 g/L. All three S. cerevisiae monocultures were carried out in 1000 mL Duran bottles, and each fermentor was filled with 700 mL of must and inoculated with a single yeast strain culture at a concentration of 10⁶ CFU/mL. Fermentations were conducted at 18 °C under static conditions and were considered complete when total sugar concentration was below 2 g/L. This calculation was achieved by weighing the bioreactors and measuring the final sugar concentration. All experiments were performed in duplicate.

Following fermentation, the wines were centrifuged at 5000× g for 10 min. The resulting wine lees were collected, washed once with Ringer’s solution (Sigma-Aldrich, St. Louis, MO, USA), and centrifuged again under the same conditions. The recovered lees were then subjected to microbiological analysis, and their biomass was determined based on wet weight measurement.

2.2. Autolysis

PEF. The PEF system used was a static bench-scale setup consisting of a high-voltage power generator capable of delivering up to 1 kV/cm, a 25 MHz function/arbitrary waveform generator, and a custom-designed electronic switch circuit composed of insulated gate bipolar transistors (IGBTs). The batch processing chamber was based on a coaxial cylinder-type electrode design. It comprised a stainless-steel inner electrode (5 mm in diameter, 165 mm in height) positioned inside a bronze outer cylinder (1 mm wall thickness, 155 mm height, 30 mm outer diameter) with a closed flat base. Two Teflon rings (28 mm in diameter, 10 mm in thickness) were fitted at the top and bottom of the chamber, each with a central hole allowing the electrode to pass through, thereby electrically isolating it from the surrounding bronze cylinder [22].

US. US-assisted processing was carried out using an ultrasonic cell disruptor (model UCD-250, Biobase, Jinan City, China) equipped with a 6 mm diameter probe. The sonication was applied in pulsed mode with 2 s ON and 6 s OFF cycles, operating at 70% of the maximum power output; this value corresponded to the maximum operating voltage tolerated by the probe. The total treatment duration was adjusted according to experimental requirements, maintaining the pulsed regime throughout. All treatments were conducted under controlled aseptic laboratory conditions to ensure consistency and reproducibility.

Dehydrated wine lees were rehydrated using a lysis buffer (containing Tris-HCl 50 mM, NaCl 150 mM, EDTA 1 mM, and Triton X-100 0.05%) at a ratio of 1:7 (w/w), and 10 mL of the homogenized suspension was used for each treatment. Three different treatment methods were applied: US alone, PEF alone, and a combination of both techniques (PEF + US). In each case, the total treatment duration was 20 min. For the combined (synergistic) treatment, samples were alternately subjected to US and PEF, with each cycle consisting of 2.5 min in each system (US and PEF), ensuring equal exposure time to both modalities within each 5 min interval. Samples were collected at 0, 5, 10, 15, and 20 min during each treatment for further analysis. All samples were processed in duplicate under aseptic conditions to ensure reproducibility and avoid contamination.

The treatment duration of 20 min was selected based on preliminary experiments that evaluated the autolytic response of yeast cells under varying exposure times. These initial trials indicated that 20 min was sufficient to induce observable autolysis while minimizing excessive cell disruption. Similarly, the structure of the combined PEF + US cycles was informed by exploratory tests designed to optimize the synergistic effects of both technologies.

2.3. Microbiological Analysis

Microbial analysis was performed as previously described by Tzamourani et al. [23]. Briefly, all samples were appropriately serially diluted and spread onto YPD agar plates. Plates were incubated at 28 °C for 2 days before colony enumeration (CFU/mL). All data correspond to the average values recovered from duplicate fermentations and duplicated technical repetitions (n = 4).

2.4. Protein Content Determination in Autolysates

The protein content in samples before and after autolysis was determined based on the Bradford method [24], using the Quick Start™ Bradford Protein Assay Kit (Bio-Rad, Hercules, California, USA). In brief, samples were first diluted 1:4 with lysis buffer, and volumes of 20 µL were added over 1 mL of Bradford Dye Reagent, previously warmed to room temperature. Samples were incubated at room temperature for 5 min, then the absorbance was read at 595 nm. The protein content, C (μg/mL), was estimated based on a calibration curve, using Bovine Serum Albumin (BSA) as a standard protein (y = 0.0008x + 0.051, R² = 0.998).

2.5. Specific Energy Consumption Calculation

The specific energy consumption (SEC) for each treatment was calculated using the formula:

$$SEC \left({\displaystyle \frac{\mathrm{k}\mathrm{J}}{\mathrm{k}\mathrm{g}}}\right)= {\displaystyle \frac{\left(P\ast t\right)}{m}}$$

where P is the power input in kilowatts (kW), t is the effective treatment time in seconds (s), and m is the mass of the treated sample in kilograms (kg). Aliquots of 10 mg were obtained for each sample using precision weighing techniques to ensure consistency across measurements. For US, the device operated at 250 W in pulsed mode (2 s ON, 6 s OFF) for a total treatment duration of 20 min (1200 s). This corresponds to 150 ON pulses (since each cycle is 8 s), resulting in a total effective ON time of tUS = 150 ∗ 2 s = 300 s.

Thus, the maximum energy input for US was as follows:

$$SECUS = {\displaystyle \frac{(0.25\ast 300) }{0.01}} = 7500 {\displaystyle \frac{\mathrm{k}\mathrm{J}}{\mathrm{k}\mathrm{g}}}$$

For PEF, the system operated at 45 W in a pulsed regime of 1 ms ON and 1 s OFF. Over 20 min (1200 s), 1200 pulses were applied, resulting in a total ON time of tPEF = 1200 ∗ 0.001 s = 1.2 s.

Thus, the maximum energy input for PEF was as follows:

$$SEC\_PEF = {\displaystyle \frac{(0.045\ast 1.2)}{0.01}}= 5.4 {\displaystyle \frac{\mathrm{k}\mathrm{J}}{\mathrm{k}\mathrm{g}}}$$

In the combined treatment (US + PEF), the 20 min duration was divided equally, with alternating cycles of 2.5 min (150 s) in each modality. For ultrasound, 150 s includes 18.75 full 8 s cycles (each with 2 s ON), resulting in an effective ON time of tcomb = 18.75 ∗ 2 = 37.5 s.

Correspondingly, the maximum energy for US in the combined mode was as follows:

$$SEC\_US\_comb = {\displaystyle \frac{(0.25 \times 37.5)}{0.01}} = 937.5 {\displaystyle \frac{\mathrm{k}\mathrm{J}}{\mathrm{k}\mathrm{g}}}$$

2.6. Statistical Analysis

Statistical significance was determined using one-way analysis of variance (ANOVA), followed by Tukey’s post hoc test for pairwise comparisons, with a significance threshold of p < 0.05. The homogeneity of the data variance had previously been tested using Levene’s test. These analyses were conducted using IBM SPSS software (V19).

3. Results and Discussion

The results of this study are presented in three main sections, each addressing a key aspect of the autolysis-enhancing treatments. First, the specific energy consumption associated with each protocol is reported to evaluate process efficiency. Second, the effects of PEF, US, and their combination on yeast viability are examined to assess cellular disruption. Finally, protein release over time is analyzed as an indicator of autolysis progression and intracellular compound recovery. All results produced by the experiments of this study, and used to generate Figures 1–3, can be found in Supplementary Table S1.

3.1. Specific Energy Consumption

Figure 1 illustrates the progression of specific energy consumption (kJ/kg) over the 20 min treatment period for each modality (US, PEF, PEF + US). As expected, energy input for US increased sharply and linearly, reaching the highest value (7500 kJ/kg) at 20 min due to its substantially longer effective ON time and higher power setting. In contrast, PEF exhibited minimal energy accumulation, reaching only 5.4 kJ/kg by the end of the treatment, reflecting its extremely short pulse input duration (1 ms) and lower power input. The combined treatment displayed an intermediate profile, with a final energy consumption of 938.2 kJ/kg, resulting from the shared exposure time between US and PEF. These results underscore the high energy demand associated with ultrasound-based treatments and suggest that the synergistic application of US and PEF may represent a viable compromise between processing efficiency and energy economy.

These results underscore the high energy demand associated with ultrasound-based treatments and suggest that the synergistic application of US and PEF may represent a viable compromise between processing efficiency and energy economy. Notably, the energy consumption values observed in this study are consistent with those reported in previous research [25,26]. Existing evidence indicates that ultrasound (US) treatments typically demand substantially higher energy inputs compared to pulsed electric fields (PEF). For instance, continuous ultrasound-assisted extraction of bioactives consumed around 190 kJ, whereas pulsed ultrasound reduced this to 80 kJ—a 40–68% energy saving [27]. In contrast, PEF treatments achieve effective results with remarkably lower energy demand, such as 2.5 kJ/kg in osmotic dehydration of strawberries, delivering enhanced dehydration and microbial inactivation with minimal energy input [28].

These findings underscore that PEF is markedly more energy-efficient than US, reinforcing the advantage of prioritizing PEF—or integrating US only when synergistically optimized—for sustainable food processing. This alignment supports the reliability of the present energy measurements and confirms the potential of combining both technologies for more sustainable processing.

3.2. Yeast Viability and Disruption Dynamics

All alcoholic fermentations were completed successfully within 8–9 days. The wine lees reached a concentration of viable cells from 8.5 ± 0.1 logCFU/mL to 8.9 ± 0.1 logCFU/mL. Biomass production was enumerated at 7.5 ± 1.5 g, 6.2 ± 0.4 g, and 6.7 ± 1.1 g for S. cerevisiae strains S1, S2, and S3, respectively.

Figure 2 illustrates the evolution of viable yeast population (log CFU/mL) during autolysis under three different treatment modalities (PEF, US, and PEF + US) as a function of time (Panel A) and strain (Panel B). All wine lees were diluted appropriately to reach the initial population (t = 0 min) of 7.1 ± 0.1 logCFU/mL. Following the onset of treatment (5–20 min), a significant and consistent reduction in viable yeast cells was observed across all modalities (p < 0.05). At the 5 min mark, specific energy consumption differed importantly between treatments: ultrasound (US) reached 1875 kJ/kg, the combined PEF + US treatment consumed approximately 234.6 kJ/kg, and pulsed electric fields (PEF) alone accounted for only 1.35 kJ/kg. This corresponds to an energy reduction of approximately 87.5% in the combined treatment compared to ultrasound and over 138.000% higher energy demand for ultrasound compared to PEF. Interestingly, no statistically significant differences were detected between the three methods at any time point beyond zero, suggesting that all autolytic approaches induced rapid inactivation during the first 5 min. Efficient recovery of intracellular compounds from S. cerevisiae using PEF depends on the treatment’s ability to permeabilize the cytoplasmic membrane through electroporation. The degree of electroporation is affected by key PEF parameters, such as field strength and treatment duration [29]. After 5 min of treatment, the lack of further yeast population reduction indicates a plateau phase, which implies that yeast cells either reached a disruption threshold or residual subpopulations were more resistant to further mechanical or electrical stress. Previous studies have shown that both heat-mediated and PEF-mediated autolysis can yield higher DNA extraction efficiencies when performed at elevated temperatures (50–60 °C) and over prolonged durations [30,31].

Figure 2B presents the strain-specific mean viability across treatments, calculated over the 5–20 min interval (including t = 0). Significant differences were observed among the three S. cerevisiae strains. Strain S1 consistently exhibited the lowest residual population post-treatment, suggesting higher susceptibility to autolytic disruption. Conversely, strain S3 maintained higher cell viability, especially under PEF and PEF + US treatments, indicating a more robust cell wall architecture or lower intrinsic autolytic activity. The cell wall component of S. cerevisiae is composed of 80–90% polysaccharides, which typically make up 15 to 30% of the cell dry mass [32]. During autolysis, respiratory enzyme activity declines while hydrolytic enzyme activity increases [33]. The activity of cell wall-degrading enzymes, such as glucanases and proteinases, is modulated by genetic differences among yeast strains, highlighting the strain-specific nature of cell wall integrity and autolysis [34]. As a result, the cell wall becomes porous and collapses, leading to the release of the intracellular compounds, such as proteins, DNAs, and RNAs, and other intracellular products into the surrounding medium [35,36,37]. More specifically, yeast strains spontaneously produce both endogenous and secreted proteases with different biochemical properties and exhibit diverse extracellular proteolytic patterns in response to external stimulations [36]. Therefore, these findings support the expectation of strain-specific autolytic responses. Notably, among the tested strains, S. cerevisiae S1 demonstrated the highest autolytic capacity. This commercial S. cerevisiae strain is proposed as a starter for secondary fermentation, and its strong autolytic capacity is a valuable technological characteristic for sparkling wine production, as the autolysis in the bottle is a marker of wine complexity [38]. The interaction effect between treatment type and strain was statistically significant (p < 0.05), further reinforcing the importance of selecting appropriate yeast strains for industrial applications.

Previous investigations into yeast inactivation have provided quantitative insights into the specific energy requirements of US and PEF treatments. In probe-based ultrasound experiments applying power levels between 74 W and 117 W to 30 mL S. cerevisiae suspensions (~0.03 kg), the specific energy input ranged from approximately 298 kJ/kg to 1146 kJ/kg, with a maximum of around a 4-log reduction [39]. In contrast, PEF treatments have been shown to require comparatively lower energy to achieve similar inactivation levels. For example, specific energies in the range of 10–30 kJ/kg were adequate to achieve a 5-log reduction of S. cerevisiae when combined with moderate temperature (~55 °C) [40]. Furthermore, yeast eradication in buffer media was shown to increase with energy inputs from ~22 to 615 kJ/kg, with complete inactivation observed at the higher end of this range, though cell resealing effects occurred post-treatment in complex media [41]. These findings align well with the current study, in which US requirements were an order of magnitude higher than PEF, while combined treatments considerably reduced total energy inputs while still achieving effective cell disruption.

3.3. Protein Recovery Dynamics and Energy Considerations

During induced autolysis of yeast cells from wine lees, techniques such as PEF, US, or their combination disrupt cell walls and release intracellular proteins. Measuring the total protein concentration in autolysates indicates the extent of cell lysis and the effectiveness of the method employed [9]. In this study, PEF, US, and their combination were tested for lysing yeast cells in wine lees over a maximum of 20 min, and the total proteins released were measured.

When considering all strains (Figure 3A), a steady increase in protein concentration was observed for US and PEF + US treatments, reaching the highest values at 20 min, with only the US treatment being significantly superior to PEF (p < 0.05). However, when excluding strain S3, the protein release advantage of both US and PEF + US over PEF became statistically clearer (Figure 3B). This confirms that PEF alone is less effective for rapid protein extraction, while US—alone or in combination—enhances autolysis efficiency, particularly in strains with higher inherent susceptibility (S1, S2). The findings reinforce the idea that strain-specific performance can influence average treatment outcomes, a critical point for industrial-scale optimization.

Different US treatment techniques have also been assessed to accelerate the autolytic process in S. cerevisiae within a model wine system in the study by Blanco-Huerta et al. [9], compared to high hydrostatic pressure (HHP) treatments up to 10 min, and US was found to promote a faster release of proteins, but without mentioning the energy cost. The environmental scanning electron microscopy (ESEM) of the treated lees that was used demonstrated that the impact on the yeast cell surface was more pronounced after exposure to US compared to treatments involving HHP [7]. Additionally, in the study by Dimopoulos et al. [10], which evaluated the effects of PEF on the progress of yeast S. cerevisiae autolysis under various treatment conditions, it was found that the final yield of autolysis, in terms of protein release, was not affected by PEF. Despite being the most energy-efficient approach, the PEF method appears, based on the results of the present study, to be ineffective in promoting the release of high protein concentrations during short-term applications. Although it can be used for a long-term result, while natural autolysis proceeds at a notably slow rate, yeast autolysis initiated by PEF treatment facilitated the release of 80% of the total protein content within just three weeks of lees aging [42].

As mentioned, samples treated with US for 20 min showed the highest protein concentration, with a statistically significant difference from those treated with PEF (p < 0.05). Applying US to lees caused a rapid, exponential rise in protein concentration in the model wine at the beginning of the treatment. This was followed by a slower, decelerating phase. While both traditional autolysis and US-assisted lysis boosted protein release, a significant increase in protein release was observed only with US [43]. Overall, the highest protein concentrations were released using the US method at the maximum treatment time of 20 min. The combined application of PEF and US resulted in similarly high protein concentrations compared to ultrasound alone, while demonstrating significantly lower energy demands. Specifically, at 20 min of treatment, ultrasound alone consumed approximately 7500 kJ/kg, whereas the combined PEF + US approach required only 938.2 kJ/kg—an approximate 87.5% reduction in energy consumption. Given that no statistically significant differences were observed between the two methods (p > 0.05), the synergistic use of PEF and US appears to be a promising, energy-efficient, and environmentally sustainable strategy for biomolecule extraction from wine lees. These findings are in agreement with previous studies showing that coupling PEF and US enhances extraction efficiency while reducing energy input. Ntourtoglou et al. [19] demonstrated that PEF pretreatment prior to ultrasound significantly improved polyphenol recovery from grape stems. Together, these results validate the potential of the PEF + US combination as a sustainable approach for enhancing extraction processes in oenological and biotechnological applications.

In the literature, several studies using methods based on pressure (e.g., bead milling, HHP) or on waves (US, PEF) have been described as physical extraction methods for releasing proteins from S. cerevisiae [44]. Among them, ultrasonication appears to be more effective for yeast cell lysis, enabling the extraction of periplasmic, membrane-bound, and insoluble recombinant proteins from yeast cells [45]. Although there has not been an extensive comparison of the energy required by each method, this factor appears to be decisive for the choice of method on an industrial scale. As ultrasound methods have operational and economical limitations, such as amplitude and energy consumption, high-pressure methods seem to be currently the most widely accepted by the industries, but they often exhibit poor selectivity and require costly maintenance [44,45].

Aiming for efficient inactivation and protein extraction with low energy in a short period, this work proposes the combination of PEF and US. According to our results, when ultrasound (US) and pulsed electric fields (PEF) are applied together, the apparent synergy is less about higher yield (since our results show no significant increase in protein extraction compared to each technique alone) and more about achieving similar yields with lower energy input in short test times. This may be due to mechanisms that complement each other at different stages of cell disruption. Applying PEF, short, high-voltage pulses cause the transmembrane potential to exceed and form hydrophilic pores in the plasma membrane. This does not always cause complete cell rupture, but it reduces the mechanical resistance of the membrane–wall complex [10,11,38]. Ultrasound induces acoustic cavitation; bubble collapse near cell surfaces generates microjets, high local shear, and shockwaves that mechanically disrupt cell walls and membranes, accelerating mass transfer. For yeast, cavitation is a primary mechanical driver of protein release [12,13,14]. Combining PEF and US, PEF “primes” cells by weakening the membrane–wall interface and creating pores that concentrate stress and reduce the force required for rupture. Consequently, when US cavitation events occur, the mechanical energy needed to create a critical crack or irreversible breach in the envelope is lower than for intact cells. Conversely, cavitation stresses can enlarge electropores or convert reversible pores into irreversible damage, so each method effectively lowers the intensity needed for the other. This bidirectional assistance may explain how combined treatments can reach the same protein yield with less total energy input. Of course, the exact physical mechanisms occurring remain insufficiently understood and warrant further investigation.

4. Conclusions

This study demonstrated the effectiveness of ultrasound (US), pulsed electric field (PEF), and their combined application (PEF + US) in inactivating yeasts and extracting proteins from wine lees. All treatments enhanced protein release, with PEF + US achieving protein yields similar to ultrasound alone, but with significantly lower energy consumption. In parallel, a marked reduction in yeast viability was observed in all treatments, particularly in the early stages, indicating effective microbial disruption. The observed strain-dependent differences in viability emphasize the importance of testing multiple yeast strains to ensure consistent and reliable inactivation outcomes. The combined approach, applying short and alternating exposure to each technology, maintained aseptic conditions and offered a balanced, non-thermal solution. While the combined PEF + US treatment demonstrated promising results in terms of energy efficiency, its scalability and industrial feasibility remain to be validated. Future studies should explore the performance of this approach under continuous processing conditions and assess potential challenges related to equipment integration, cost, and process control. Overall, the synergy between PEF and US presents a promising and sustainable strategy for valorizing wine industry by-products, combining efficiency, energy savings, and process safety. These findings support the broader use of hybrid technologies for green extraction and integration into circular bioeconomy models.