Background and Objective: Yeasts are widely reported as valuable sources of intracellular bioactive compounds with important uses in food, pharmaceutical and biotechnological industries. Efficient recovery of these compounds depends largely on the structural complexity of yeast cells and the diversity of extraction strategies developed to access various molecular targets.

Material and Methods: This review provided a comprehensive overview of the extraction and isolation techniques such as enzymatic, mechanical and green technologies, used in the recovery of bioactive compounds from yeast cells. Wide arrays of biomolecules, including vitamins, pigments, lipids, proteins, peptides and minerals, are discussed in the context of their extraction techniques from yeast cells.

The novelty of this review depends on its integrated assessment of current technologies, mapping the evolution of yeast extraction methods and proposing further directions, involving nanotechnology, hybrid extraction systems and enzyme-assisted green protocols.

Results and Conclusion: This study provided a comprehensive review on bioactive compound recovery methods from yeast cells, highlighting advancements in ultrasonic, enzymatic, microwave-assisted and other extraction methods for improved efficiency. Furthermore, physical techniques [e.g., high-pressure homogenization (HPH) and pulsed electric field (PEF)] and green extraction technologies (e.g., supercritical CO₂) offered enhanced yields and decreased ecological footprints. Further directions necessitate optimizing compound specificity, industrial scalability and integrating novel platforms such as nanotechnology and advanced enzymatic systems for efficient, sustainable and commercially viable uses.

Keywords: Bioactive compounds, Encapsulation, Extraction, Functional foods, Yeast cells

1. Introduction

Yeast cells have emerged as innovative carriers for the microencapsulation of bioactive compounds, demonstrating significant advantages in the controlled release and protection of sensitive core substances. This encapsulation approach not only enhances the stability of bioactive materials but also minimizes alterations to the sensory characteristics of food products; thereby, maintaining consumer acceptability [1].

The sources demonstrate strong consensus on yeast classification. Approximately 1500 yeast species are described, with Saccharomyces as the most extensively studied genus [2]. Key species include Saccharomyces cerevisiae (S. cerevisiae) (the preeminent industrial ethanologen), S. byayanus, S. pastorianus and nonconventional yeasts such as Brettanomyces, Hanseniaspora and Pichia [3]. Industrial uses span multiple sectors: fermented beverages (beer, wine and sake) [4], bioethanol production [3], pharmaceuticals (insulin and vaccines), food additives and probiotics [5].

From various yeast species, S. cerevisiae has achieved prominence as an effective encapsulating agent due to its unique structural and biochemical characteristics [6]. The S. cerevisiae is a unicellular eukaryote widely used in biotechnology, fermentation and bioactive compound production. Its robust multilayered cell wall (composing of β-glucans, mannoproteins and chitin) provides mechanical stability and limits access to intracellular contents. Close the wall, the plasma membrane regulates molecular transport and homeostasis. Yeast cells contain membrane-bound organelles, including the nucleus, mitochondria, endoplasmic reticulum, Golgi apparatus and vacuole, which compartmentalize metabolism and store bioactive molecules such as proteins, enzymes, lipids, pigments and vitamins [2]. Figure 1 demonstrates the schematic structure of yeast cells.

Research shows diverse uses of yeast-derived bioactive compounds. Tan et al. documented yeast cell-derived delivery systems successfully encapsulating flavors, vitamins, carotenoids and phenolics [7]. Moreover, yeast-derived peptides with antioxidant, ACE-inhibitory and antidiabetic bioactivity were identified [8]. Sarkar et al. highlighted yeast β-glucans with immunomodulatory and metabolic health benefits [9]. The most robust human evidence is from a study, which reviewed oral yeast-derived β-glucans in multiple trials. Particularly, yeast-derived β-glucans significantly decreased upper respiratory tract infection incidence in susceptible individuals, with studies using doses ranging 7.5–1500 mg d^-1. Additional benefits included increased salivary IgA and IL-10 levels in various populations [10]. In addition, yeast cells were verified as effective microencapsulation carriers with improved bioavailability and stability, compared to other wall materials. However, the sources highlight that optimal dosing and preparation standardization need further investigations [1].

For more than five decades, S. cerevisiae has been used as a coating material for a diverse range of flavors and essential oils, reflecting its versatility and functional potentials in food technology [6]. Vegetable oils and bioactive compounds such as Zataria multiflora Bois. [11], curcumin [12], berberine [13, 14] and flaxseed oil [15] have widely been used for various industrial and therapeutic uses due to their well-documented pharmacological and nutritional benefits. Yeast cells, specifically S. cerevisiae, are a unique and appealing coating material for encapsulation in food and pharmaceutical industries. This is primarily due to their unique cellular structure and established role in human nutrition, where they are recognized as qualified presumption of safety (QPS) material by regulatory authorities [16]. However, many bioactive substances, including those derived from yeasts, face stability issues when exposed to harsh industrial and environmental conditions. This instability can lead to significant degradation or loss of efficacy, presenting a significant challenge in the use of these bioactive compounds. In response, the technology of microencapsulation has emerged as a promising solution to protect these molecules, enabling their controlled release and stability during storage and processing [17].

Microencapsulation involves a series of critical steps; one of the most critical ones includes extraction of bioactive compounds from the coating materials. Figure 2 summarizes the encapsulation and extraction procedures of bioactive compounds from S. cerevisiae cells.

Recent advancements in extraction technologies have significantly enhanced the efficiency and specificity of recovering bioactive compounds from yeast cells. From these, pulsed electric field (PEF) treatment has shown significant promise for selective extraction of the intracellular proteins and vitamins of yeast cells as rich sources of bioactive compounds by facilitating the release of small molecules while preserving the integrity of macromolecules within the cells [18]. Additionally, high-pressure homogenization (HPH) and PEF have been demonstrated effective in extracting a range of bioactive components of S. cerevisiae cells such as glutathione, proteins and mannoproteins with PEF offering a gradual controlled extraction process [19].

Furthermore, various innovative extraction techniques of bioactive components from yeast cells as a natural source, including ultrasound-assisted extraction, microwave-assisted extraction and supercritical fluid extraction, have been investigated for their potentials to improve extraction yields and decrease environmental effects. These methods offer distinct advantages for enhanced selectivity, efficiency and sustainability, when compared to traditional extraction techniques.

Emerging extraction technologies such as green extraction methods and nanotechnology-based approaches include significant promise for the recovering bioactive compounds from natural sources [20]. To moderate the release of hazardous waste into the environment, researchers have increasingly adopted green chemistry approaches. These methodologies avoid the use of toxic reagents, energy‑intensive processes, costly solvents and high‑temperature reaction conditions; thereby, minimizing environmental effects and resource consumption [21].

This study aimed to provide a comprehensive review of the various extraction methods used to recover bioactive molecules from yeast cells, whether originally from the cells as a natural source or previously inserted. By categorizing the techniques based on the types of substances extracted, the study aimed to highlight the advancements and challenges in the field, as well as the potential uses for these compounds in the food and pharmaceutical industries.

Despite extensive studies on yeast bioactive extractions, comparative analyses that unify conventional and emerging methods within a sustainability framework are still limited. This review addressed this gap by consolidating diverse extraction strategies and assessing them. In particular, it emphasized the convergence of physical and green extraction approaches with innovative technologies such as nanotechnology and enzyme assisted systems.

1. Materials and Methods

The following keywords were used to search for this review studies using a number of electronic search engines and databases, including PubMed, Scopus and Google Scholar: extraction, isolation, encapsulation, microencapsulation, bioactive compounds, yeast cell and S. cerevisiae. The authors collected every published original study and review that looked into techniques for extracting bioactive substances from yeast cells. The most relevant articles were included regardless of the time of publication.

1. Results and Discussion

The extraction of bioactive compounds from yeast has been investigated using the structural and biochemical characteristics of target molecules. The results under synthesize representative techniques were reported in various yeast species for the recovery of nucleic acids, proteins, peptides, pigments, polysaccharides, vitamins, lipids and mineral elements. Emphasis was on the underlying extraction principles, methodological diversity and primary functional uses, providing an integrated overview of current technological approaches used to achieve yeast-derived bioactive compounds for analytical, nutritional and industrial purposes.

3.1. Nucleic Acids

3.1.1. DNA

Current S. cerevisiae DNA extraction methods are too time-consuming and labor-intensive or produce variable low-quality DNA. Therefore, they are not well appropriate for comprehensive colony screening using polymerase chain reaction (PCR). To address this issue, the glass bead Chelex 100 preparation method, also known as gas chromatography (GC) prep technique, has been introduced. In the presence of a metal chelating resin, glass beads are vortexed with S. cerevisiae cells from liquid cultures or colonies to lyse them. Using this method allows for the extraction of high quantities of genomic DNA from multiple samples in approximately 12 min [22].

In another investigation, EtNa, a rapid inexpensive DNA extraction method that acts well for yeasts and bacteria at a variety of concentrations, is introduced. The EtNa is based on the lysis of hot alkaline ethanol. This technique uses heating in an ethanol alkaline solution to extract single-stranded DNA from yeasts and bacteria with similar effectiveness. A crude DNA pellet is produced by centrifugation; however, purification is possible by direct addition to silica columns. When the bacteria identities as unknown and it is critical to avoid ignoring or favoring any of them, this process can be used [23].

An effective technique has been created for the direct extraction of yeast genomic DNA, which involves silica-adsorption of DNA on microcolumns. In this method, a protracted lysis is carried out using hot detergent after an enzymatic cell-wall destruction phase. For later qPCR assays that assess mixed yeast populations in artisan Mexican mezcal fermentations, the resultant extracts created good templates [24].

In a fully automated protocol, it is demonstrated that the extraction of S. cerevisiae DNA with magnetic beads eliminates the need of hazardous reagents (e.g., chloroform and phenol), incubation of samples at 100 °C (e.g., boiling) and glass beads for mechanical cell disruption. This protocol consists of five basic steps of RNA and cell wall digestion, cell lysis, DNA binding to magnetic beads, washing with ethanol and elution [25].

3.1.2. RNA

Traditional yeast extraction methods have been developed for laboratory strains with modest cell densities that grow relatively quickly. These frequently lack the integrity needed for repeated culture samples. Then, these procedures are typically ineffective for non-laboratory strains or cultures with sluggish growth conditions. Using intense bead beating, a mechanically chemical disruption process has been prepared that can reliably disrupt S. cerevisiae cells (> 95%), regardless of their metabolic condition or cell cycle. Glass beads are used in this technique to break the cells. These are disrupted and then incubated in a temperature-controlled mixing block for optional RNase treatment. After adding phenol, chloroform and isoamyl alcohol in a 25:24:1 ratio, the tubes are inverted and centrifuged and the aqueous phase is transferred to fresh microcentrifuge tubes. A Mini Disk Rotor is used to allow the nucleotides to precipitate after 99.5% ethanol is introduced to the aqueous phase [26].

It is difficult to isolate RNA from S. cerevisiae cells because the thick inflexible cell wall must be broken first. The method described involves heating and freezing cells in a cycle while phenol and the sodium dodecyl sulfate (SDS) detergent are present. Low levels of salt are present during the extraction process, allowing DNA to be extracted from the interface, while RNA presents in the aqueous phase after the phenol and aqueous phases are separated using centrifugation [27].

In an experiment, extraction with solution of formamide and ethylene diamine tetra acetic acid (EDTA) was modified to separate RNA from entire S. cerevisiae cells using a readily scalable method that did not need enzymes, phenol or mechanical cell lysis. Without the need for alcohol precipitation, RNA extracted with formamide-EDTA could be used directly to gels for electrophoretic analysis. Compared to traditional procedures, the formamide-EDTA extraction of S. cerevisiae RNA was quicker, safer and further cost-effective. It was also carried out better overall and significantly boosted throughputs [28].

3.2. Proteins

The use of PEF treatment is suggested as an alternative method for the efficient extraction of proteins and other bioactive intracellular compounds from baker's yeasts, with a focus on decreasing the nucleic acid content in the final product [18]. The effectiveness of three physical extraction methods (HPH, PEFs and heat treatment) for achieving diverse biomolecules such as amino acids, proteins and mannoproteins from S. cerevisiae have been studied as well. The findings suggested that HPH and PEFs were the most effective methods [19].

3.2.1. Enzymes

Two-stage method of buffer autolysis was used to extract glutathione reductase from baker's yeast cells treated with toluene for 1 h at 40 °C. After removing the toluene, the cells were autolyzed in buffer for 72 h at 4 °C. A second autolysis stage was carried out for 96 h. The enzyme was then purified using affinity chromatography, achieving a 786-fold increase in purity with 80% recovery [29].

3.2.2. Peptides

In extraction of DesPro(2)-Val15-Leu17-aprotinin from S. cerevisiae’s culture supernatants, an aqueous two-phase partitioning method was used with chymotrypsin as an affinity ligand. This technique used a polyethylene glycol/salt mixture, where the aprotinin-chymotrypsin complex accumulated in the salt-rich (bottom) phase, driven by hydrophobic interactions. The complex could be dissociated by adjusting the pH to a lower value and chromatographic procedures were then used to separate the recombinant aprotinin and protease, ensuring effective purification [30].

3.3. Pigments

Natural pigments have achieved significant attention in industrial uses due to their numerous health benefits and their ability to enhance the nutritional content of products. Unlike synthetic dyes, which include potential health risks, natural colors are favored for their therapeutic characteristics. Marine yeasts have been identified as potent producers of such pigments. In a study in India, four marine yeast isolates were reported to produce carotenoid pigments, with isolate KSB1 yielding a carotenoid concentration of 856 µg g^-1. The pigments were characterized through Fourier transform infrared spectroscopy (FTIR) and high pressure liquid chromatography (HPLC) analyses, revealing promising antioxidant and antibacterial characteristics [31].

Further research into astaxanthin extraction from Xanthophyllomyces dendrorhous demonstrated an efficient three-stage process involving yeast growth, cell wall disruption and pigment extraction using solvents such as ethanol, methanol and acetone. The optimal extraction conditions were reported as 10% ethanol concentration, temperature of 30 °C and yeast cell density of 10 g l^-1 extracted for 24 h, resulting in the highest yields [32].

In another study focusing on bioreactor optimization for astaxanthin production, two commercial enzymes, Glucanex and Accelerrase 1500, were used to lyse the yeast cells. Results revealed that Glucanex achieved 100% astaxanthin extractability, far outstripping the conventional dimethyl sulfoxide (DMSO) and acetone methods. Additionally, use of supercritical CO₂ as an extraction solvent resulted in a 2.5-fold increase in astaxanthin yield [33]. Moreover, a novel eco-friendly technique, using ultrasonic treatment followed by centrifugation, was developed to extract carotenoids from Rhodotorula glutinis. This method resulted in 82% recovery of carotenoid content, producing an extract rich in carotenoids that could be used in cosmetics, pharmaceuticals and food products [34].

3.4. Saccharides

Yeasts from various industries serve as economical raw materials for mannan extraction. Yeast‑derived mannan is primarily used as a prebiotic in animal feed. Although no standardized extraction protocol is available, methods are selected based on the intended use and cost considerations. From the available approaches, the acid–alkaline method is the most widely adopted and cost‑effective method [35]. Other extraction methods, including chemical, physical and enzymatic techniques, have been investigated as well. From these, an alkaline thermal method produced the highest yield (58.82%), though it resulted in a low mannose concentration. In contrast, autolysis followed by hydrothermal treatment provided an extract with a higher mannose concentration of 59.19%. Particularly, enzymatic hydrolysis led to the highest prebiotic activity, showing the potential for improved functional uses [36].

Mannoproteins and beta-glucans are key components of yeast cell-wall matrix particles. Using alkaline-acid extraction technique, 1→3-β-D-glucan was successfully isolated from the yeast cell wall. Chemical analysis, including IR spectroscopy, verified the absence of proteins or additional carbohydrates in the final product, highlighting alkaline-acid technique as the most effective technique for isolating β-glucan. Furthermore, the dilute alkali Sevage technique was used for extracting mannan oligosaccharides, offering promising results for further uses [37].

3.5. Vitamins

3.5.1. Vitamin A

Vitamin A, essential for vision, immune function and skin health, is traditionally synthesized chemically from petroleum-based substrates. However, a biotechnological approach involving S. cerevisiae has been developed to produce vitamin A from xylose, a non-edible sugar. This bioprocess was further optimized by using a two-phase in situ extraction technique and olive oil or dodecane as extractive agents. The results showed significant two-fold increases in vitamin A production, achieving a final titer of 3350 mg l^-1, including retinol (1256 mg l^-1) and retinal (2094 mg l^-1). This approach offered a promising solution for the limitations of vitamin A production, including intracellular storage capacity and precursor availability [38].

3.5.2. Vitamin B

The liquid chromatography/mass spectrophotometry-time of flight (LC/MS-TOF) technique coupled with stable isotope dilution assays has been used to measure key B vitamins, including thiamine, nicotinic acid, nicotinamide, riboflavin, pantothenic acid, pyridoxine and pyridoxal. Yeast powder served as the model food matrix in this study. Several enzymatic treatments, including acid extraction, were assessed to set the most effective methods for converting complex vitamers into forms that could be quantified using isotope-labeled standards. The enzyme preparations, including α-amylase, β-glucosidase and takadiastase, successfully released thiamine and riboflavin. However, enzymes were unable to release pantothenic acid and nicotinamide from their precursors NAD (+) and CoA. Furthermore, hydrochloric acid extraction at 121 °C for 30 min improved pyridoxal release but was detrimental to pantothenic acid [39].

In another study, a method was developed to extract thiamine from S. cerevisiae biomass, yielding a highly fluorescent thiamine derivative. The technique involved bead-beating S. cerevisiae biomass in 0.1 M HCl with silica spheres, which preserved thiamine pyrophosphate in its biologically active form. Compared to conventional heat treatments, this method was verified further effective in preserving various thiamine vitamers in their natural proportions [40]. Optimal conditions for vitamin B₁ extraction from yeast biomass included pH 1.0–1.5 using papain enzyme with optimal conditions for synthetic vitamin B₁ conservation at pH 4.0–4.6 [41].

3.5.3. Folic Acid

Folic acid (vitamin B₉) is widely incorporated into fortified foods, particularly flour and rice. However, its inherent instability is a major challenge, as degradation decreases its bioactivity. In recent decades, encapsulation of folic acid within yeast cells as protective matrices has successfully emerged as a promising strategy to enhance its chemical stability [42].

In an investigation to produce a concentrated folic acid conjugate from brewer's yeasts, a novel extraction technique combining alcohol and water was developed. The initial extraction used 45–50% alcohol, mildly acidified to remove most extractives. After treating with 60% alcohol to eliminate unwanted compounds, the pH was adjusted to 3 and 70% anhydrous alcohol was added to concentrate the extract further. Then, a relatively inert precipitate was removed, and the pH was adjusted to 6.0, facilitating the precipitation of the active conjugate. This method yielded a 40%, resulting in a 400-fold concentrate containing 22 mg of the conjugate per gram of solid S. cerevisiae cells [43].

3.6. Lipids

Multiple extraction and analysis techniques for the extraction of S. cerevisiae lipids have been developed since none of them provide for the thorough and quantitative identifications of various molecular lipid species, even for an organism such as S. cerevisiae that has a relatively simple lipidome. A number of extraction techniques have been explained as follows.

Ultrasonic aided extraction (43±0.33%, w/w) has produced the maximum lipid content from Trichosporon sp. biomass in a study, compared to the traditional Soxhlet (30±0.28%, w/w) and binary solvent [chloroform: methanol, (2:1, v/v)] (36±0.38%,w/w) techniques, respectively. With frequency of 50 Hz and power of 2800 W, the established process parameters of ultrasonic-assisted extraction combined with a chloroform/methanol solvent system produced a conversion efficiency of 95–97% within 20 min at 30 °C. These findings supported the idea that ultrasonic-assisted extraction is a potentially environmentally friendly extraction method that uses less energy, time and solvent without sacrificing the quality of the lipids. Using this green extraction method, the process of producing biodiesel may become further affordable and environmentally benign [44]. In another investigation, lipids were extracted from the Rhodosporidium toruloides yeast strain using enzyme-assisted method. Straight from the culture without dewatering, nearly 96.6% of the total lipids were extracted from R. toruloides cells at atmospheric pressure and room temperature (RT). This was accomplished using microwave heat pre-treatment, enzymatic treatment with the recombinant β-1,3-glucomannanase, plMAN5C and extraction with ethyl acetate. As a result, this procedure may greatly decrease the energy use and costs associated to extracting lipids from yeast [45].

Lipids may be extracted in as fast as 10 min using a microwave-assisted method. This technique preserves product yields equivalent to traditional approaches while increasing the extraction rate by 27 times. Furthermore, this technique integrates extraction and cell disruption in a single step, which significantly streamlines sample handling, cuts down on analysis time and minimizes sample loss during sample preparation. Using this procedure, 7 ml of chloroform-methanol (2:1, v/v) and internal standard were mixed with freeze-dried cells in an extraction tube. The tube was thoroughly vortexed after being flushed with nitrogen gas. This was transferred in the microwave reaction vessel with 30 ml of Mili-Q water and sealed using screw cover. The microwave digestion method was used to heat the vessel. For 10 min, the microwave extraction temperature was ramped up to 60 °C and set at this temperature. Following RT cooling, 0.73% (w/v) NaCl was added and the mixture was then vigorously vortexed. The organic phase was transferred into a fresh clean extraction tube after the sample was centrifuged for 10 min to allow for phase separation [46]. In a separate experiment, an exceptionally high yield was achieved using the S. cerevisiae total lipid extraction method, appropriate for further LC-MS and thin-layer chromatography (TLC) analyses. This procedure involved breaking S. cerevisiae cells with glass beads using mechanical bead mill and CHCl3/CH3OH (2:1, v/v) as the solvent for lipid extraction. To ensure effective lipid extraction, the procedure was carried out five times [47].

To increase the extraction yield, various methods of extracting lipids from Yarrowia lipolytica yeast were investigated in another study. To increase the effectiveness of lipid recovery, a number of available extraction methods such as bead milling, microwave and ultrasound methods were assessed. The most effective enhanced extraction technique was seemingly bead milling, while the best pre-treatment technique for lipid extraction from yeast was cold drying under pressure, which yielded twice as much as traditional maceration. In contrast to the cold drying pretreatment, which appeared as the most energy-intensive method, bead milling extraction was the best option for lipid recovery, according to research of energy consumption and environmental effects. A kinetic analysis of bead milling extraction revealed that the most critical phase included break down of the yeast cell wall, enabling improved mass transfer of lipids from the yeasts to the solvent [48].

3.7. Minerals

3.7.1. Zinc

Yeast cells are well documented for their capacity to bioaccumulate metal ions from aqueous environments through various physicochemical and biochemical mechanisms, including surface adsorption, intracellular absorption and metabolism-dependent uptake. Incorporation of soluble inorganic salts into the culture media during yeast cultivation facilitates the assimilation of significant quantities of target metals such as zinc. For zinc extraction, cells were centrifuged, washed to remove residual media and dried. Digestion was carried out using concentrated nitric and sulfuric acids under controlled heating to fully oxidize the organic matrix and release intracellular zinc [49].

3.7.2. Chromium

To facilitate the extraction of chromium from yeast cells, the cells were treated with 0.1 N ammonium hydroxide (NH₄OH) for 30 min at pH levels exceeding 8.0; a condition under which, a majority of chromium was recoverable. This procedure resulted in the extraction of approximately 85% of the labeled chromium. In contrast, ethanol treatment yielded a significantly lower extraction efficiency, removing less than 5% of the total chromium content. Furthermore, over 80% of the labeled chromium was released following cell lysis induced by a 12-h treatment with teichozyme-Y. However, the resulting extract showed diminished bioactivity, particularly when sonication was used post-treatment [50].

3.7.3. Iron

Iron-enriched yeast biomass presents a promising potentially safer alternative for dietary iron supplementation aimed at preventing iron-deficiency anemia. To export intracellular iron content for quantification purposes, the biomass was first harvested by centrifugation; after which, the supernatant was discarded. The resulting cell pellet was washed three times with distilled water (DW) to eliminate loosely bound or surface-associated iron. The purified yeast biomass was dried to a constant mass. Then, the dried sample was subjected to acid digestion with nitric acid, followed by heating at 140°C to ensure complete decomposition of organic matter and release of intracellular iron for further analysis [51]. The summary of yeast bioactive compound extraction strategies is present in Table 1. In addition, all the techniques for extracting bioactive compounds from yeast cells are summarized in Figure 3.

Extraction of a wide range of bioactive molecules—such as proteins, peptides, lipids and vitamins—from yeast cells has led to advancements in the development of functional foods, nutraceuticals and pharmaceutical uses. Techniques such as PEFs for protein extraction, as well as enzymatic methods for glutathione reductase and peptides, have demonstrated promising results for yield and efficiency [19, 29, 30]. The potential for extracting high-value compounds such as folic acid [43] and natural pigments from marine yeasts further extends the versatility of yeast-based systems in various industrial contexts [31–33]. Despite these promising developments, challenges are still in optimizing extraction processes to achieve higher yields and improve efficiency within various types of bioactive molecules. Some bioactive substances, especially those embedded within the yeast cell wall, make difficulties in extraction efficiency and preservation of their bioactivity.

Bioactive compound extraction from yeast cells can broadly be categorized into conventional and emerging techniques. Conventional chemical extraction methods have long been used but often suffer from limitations in yield, selectivity and environmental sustainability [52]. In contrast, emerging techniques, including PEF [18], HPH [19], ultrasound-assisted extraction, supercritical fluid extraction and microwave-assisted extraction [20], offer significant improvements in efficiency, selectivity and eco-friendliness. These advanced methods not only enhance the yield and purity of bioactive compounds but also align with the growing demand for greener further sustainable bioprocesses.

Overall, extraction of yeast-derived bioactive compounds is determined by the physicochemical characteristics, cellular localization and stability of the target molecules. Mechanical disruption methods are commonly used to overcome the rigid yeast cell wall and release intracellular components such as proteins, peptides and lipids. Enzymatic treatments provide selective and mild conditions, making them particularly appropriate for recovering functional cell-wall polymers and structurally susceptible biomolecules. Chemical extraction techniques are typically used for compounds with specific solubility characteristics, including pigments, vitamins and certain polysaccharides, although process conditions must carefully be controlled to prevent degradation. Emerging and green technologies such as ultrasound, microwave and supercritical fluid extraction enhance efficiency while decreasing solvent use and preserving bioactivity. Overall, method selection reflects a balance between effective cell disruption, compound stability and process efficiency.

Further research should focus on the following areas:

1. Optimization of extraction processes: Tailoring extraction techniques to maximize the yield and bioactivity of specific compounds, particularly those embedded in the yeast cell wall, is critical. A combination of enzymatic, physical and green extraction methods may offer synergistic effects that improve efficiency and sustainability.
2. Scalability and commercial viability: To make yeast-based encapsulation and extraction methods commercially viable, further studies are needed to scale up these processes while maintaining cost-effectiveness and efficiency. This includes optimizing reaction conditions, decreasing energy consumption and improving reproducibility of large-scale extractions.
3. Integration of nanotechnology: Use of nanotechnology in extraction and encapsulation can revolutionize the field by enhancing the efficiency of bioactive compound recovery and providing further precise controls over the release of encapsulated molecules [53]. Future research should investigate the potential of nanoparticles, nanocarriers and nanostructured materials in improving yeast-based encapsulation and extraction technologies.
4. Essential comparative assessment of extraction techniques: Further research should prioritize a comprehensive comparison of various extraction methods to quantify their respective yields and efficiencies. Essential quantitative data, including extraction yield, recovery percentage and the concentration of bioactive compounds, are critical for assessing the practical and industrial applicability of these techniques.

By addressing these challenges and investigating novel routes of research, further advancements in yeast-based encapsulation and extraction techniques further enhance their applicability and performance in various industries, providing ways for the development of further sustainable, innovative and efficient solutions for bioactive compound preservation and delivery.

1. Conclusion

In conclusion, the technologies discussed in this study demonstrate significant advancements in improving the stability, efficiency and sustainability of bioactive compound extraction from yeast cells. Use of physical techniques such as PEFs and HPH with emerging green extraction methods includes significant potential for enhancing extraction yields while minimizing environmental effects.

While the potential of yeast-based microencapsulation and extraction techniques is substantial, further research is needed to address several challenges. These include optimizing extraction protocols for specific bioactive compounds and improving the scalability of the current methods. Additionally, the integration of emerging technologies such as nanotechnology and advanced enzymatic approaches may offer novel pathways for enhancing efficiency and sustainability of the bioactive compound recovery from yeast cells.

The distinctive contribution of this review depends on its comprehensive analysis of extraction mechanisms by categorizing the techniques based on the types of substances extracted while highlighting the advancements and challenges in this field.

Background and Objective: Staphylococcus aureus is a significant foodborne and zoonotic pathogen. This study aimed to enhance the anti-Staphylococcus aureus activity of egg-white lysozyme through heat treatment and synergistic combinations with natural antimicrobials.

Material and Methods: The lysozyme was purified from egg white via ammonium sulfate precipitation and cation-exchange chromatography, yielding a homogeneous protein. Anti-Staphylococcus aureus activity of native lysozyme, heat-treated lysozyme and its combination with ferulic acid or Mycobacterium smegmatis acyltransferase was assessed, respectively. All experiments were carried out in triplicate and statistical analysis was carried out using SPSS software.

Results and Conclusion: The specific activity of lysozyme to Micrococcus lysodeikticus was 27,407.4 U.mg^-1. The lysozyme IC₅₀ against Staphylococcus aureus was 300.8 µg.ml^-1, with transmission electron microscopy verifying bacteriolytic action. Heat treatment under optimized conditions (90°C, 15 min, pH 6.2) significantly enhanced the lysozyme antibacterial activity by 35.1%, which was correlated with structural changes evidenced by circular dichroism spectroscopy. Furthermore, synergistic effects were observed when heat-treated lysozyme was combined with ferulic acid or Mycobacterium smegmatis acyltransferase (MsAcT), leading to prolonged inhibition and decreased viable bacterial counts. The findings of this research demonstrated that structural modifications and combinatorial strategies could effectively improve the efficacy and application potential of lysozyme as a natural antimicrobial agent in food safety.

Keywords: Egg-white lysozyme, Ferulic acid, Heat-induced fibrillar aggregates, Mycobacterium smegmatis acyltransferase, Staphylococcus aureus

1. Introduction

Staphylococcus aureus is a Gram-positive bacterium widely spread in nature and a common foodborne pathogen. The enterotoxins (SEs) of the bacteria show significant heat resistance, rendering them difficult to completely eliminate through conventional cooking methods and posing a significant risk of food poisoning. Furthermore, S. aureus facilitates cross-infection between humans and animals via the food chain through contamination of animal feed, subsequent infection of animals and transmission to humans, establishing it as an important zoonotic pathogen. The S. aureus has been detected in various animal species and a wide range of food products, with its prevalence continuously increasing on a global scale. Therefore, S. aureus is still a high-priority target in food safety monitoring [1, 2].

To combat S. aureus contamination in food processing, diverse biological control strategies have been investigated. These include inhibition using lactic acid bacteria (LAB) probiotics [3], Lysostaphin [4], bacteriophages [5,6], essential oils (EO) [5] and lysozyme [7,8]. From these, egg-white lysozyme has widely been used due to its high catalytic activity, simple preparation and cost-effectiveness [7,9]. It inhibits S. aureus through two primary mechanisms of (1) lytic mechanism as lysozyme hydrolyzses β-1, 4-glycosidic bonds to peptidoglycan, which causes cell wall damage, induces cell lysis and results in bactericidal activity; and (2) the non-lytic mechanism, where under denaturing conditions, it suppresses growth through inherent protein characteristics such as hydrophobicity and cationic effects [8,10]. While the antimicrobial characteristics of native egg-white lysozyme are well-documented, its efficacy under common food processing conditions, particularly those involving heat, needs further investigation. Moreover, strategies to enhance its activity, especially in a heat-treated state, through combination with other natural antimicrobial agents are still under-investigation. Under various reaction conditions, the interactions between other natural antibacterial agents and egg-white lysozyme can lead to various effects on its antimicrobial activity. For example, in an alkaline solution, theaflavin covalently binds to egg-white lysozyme, resulting in significant decrease of its antibacterial activity [11]. However, in an amyloid fibril hydrogel, the interaction between epigallocatechin gallate (EGCG) and egg-white lysozyme significantly broadens the antibacterial spectrum of egg-white lysozyme [12]. Therefore, an in-depth investigation into the interaction conditions between natural antibacterial factors and egg-white lysozyme is greatly important for enhancing the antibacterial efficiency of egg-white lysozyme.

This research aimed to isolate and purify egg-white lysozyme using chromatography and to assess the effect of heat treatment on its anti-S. aureus activity. Furthermore, the study prepared a synergistic combination of heat-treated lysozyme with ferulic acid and acyltransferase to develop an enhanced strategy for suppressing S. aureus growth. The experimental results demonstrated that heat-induced structural modification (fibrillar aggregation) enhanced lysozyme antimicrobial mechanism beyond native peptidoglycan hydrolysis. Moreover, it has first been reported that a synergistic combination of heat-treated lysozyme with ferulic acid and MsAcT against S. aureus significantly improves the efficacy and time of inhibition.

1. Materials and Methods

2.1. Strains, biochemical reagents and chemical reagents

Micrococcus lysodeikticus CGMCC 1.4547 and S. aureus CGMCC 1.282 were purchased from China General Microbiological Culture Collection Center, China. Bovine serum albumin (BSA) and standard protein molecular weight marker were purchased from Takara Biomedical Technology, China. Moreover, CM-Sepharose fast flow chromatography column was purchased from GE Healthcare, China. Ferulic acid, caffeic acid, gallic acid and N-acetylglucosamine (NAG) included analytical grade unless otherwise specified and purchased from Shanghai Macklin Biochemical Technology, China.

2.2. Purification of egg-white lysozyme

Egg white was initially diluted by 50-fold using 50 mmol.l^-1 pH8.0 tris-HCl buffer and the resulting protein solution was ultrasonicated for 10 min. The supernatant was collected by centrifugation at 12,000 rpm for 5 min at 4 °C. Solid ammonium sulfate was added to the supernatant to achieve 40% (w/v) saturation and the mixture was set to precipitate for 2 h at 4 °C. After centrifugation, the precipitate was removed and the collected supernatant was further loaded onto a CM-Sepharose fast flow column (5 × 20 cm) that was pre-equilibrated with 50 mmol.l^-1 tris-HCl buffer (pH 8.5). The lysozyme was eluted with a 5-fold column volume of 50 mmol.l^-1 tris-HCl buffer (pH 8.5) with increasing concentrations of NaCl (0.1 and 0.5 mol.l^-1) at a flow rate of 20 ml.h^-1, respectively. The active fractions were pooled. Protein concentration was assessed using Bradford method with BSA as a standard. The homogeneity of the purified egg-white lysozyme was assessed using sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE).

2.3. Activity assay of egg-white lysozyme

The cell lysis activity of lysozyme from egg white to M. lysodeikticus was quantitatively assessed using spectrophotometeric assay method, which was described by the National Standard of the People's Republic of China (GB/T 30990-2014, Determination of Lysozyme Acyivity) as well as Naveed et al. [13]. Briefly, after incubation overnight, M. lysodeikticus was transferred into the fresh LB liquid media at 1% (v/v) and incubated at 37 °C for 2 h at 220 rpm. The cells were harvested by centrifugation and resuspended in 50 mmol.l^-1 Na₂HPO₄-Na₂HPO₄ buffer (pH 6.2) to a final concentration of 5 × 10⁵ CFU.ml^-1. The reaction system contained 2.5 ml of cell suspension and 0.5 ml of lysozyme solution, while the control system received 0.5 ml of inactivated lysozyme solution. The reaction was carried out at 50 mmol.l^-1 Na₂HPO₄-Na₂HPO₄ buffer (pH 6.2) for 5 min at 25 °C and the absorbance of the reaction system was continuously monitored at 450 nm. One unit of lysozyme activity (U) was defined as the quantity of enzyme necessary to decrease OD₄₅₀ by 0.001 per minute under the standard assay condition.

2.4. Anti-Staphylococcus aureus activity of egg-white lysozyme

Briefly, S. aureus in the logarithmic growth phase was used as an indicator microorganism to assess antibacterial activity of the purified lysozyme. To set the growth curve, overnight-cultured S. aureus was subcultured into the fresh LB liquid media at a 1% (v/v) inoculum and incubated at 37 °C for 14 h at 220 rpm. Samples were collected every 30 min to measure the optical density (OD) of the culture broth at 600 nm. The growth curve was plotted with OD₆₀₀ on the y-axis and incubation time on the x-axis.

The antibacterial assessments were carried out using a method described by Carrillo et al. [14] with slight modification. Following 2 h of subculture, S. aureus suspensions were adjusted to 5 × 10⁵ CFU.ml^-1 through serial dilution in LB liquid media. Aliquots of S. aureus suspensions were mixed with equal volumes of the purified lysozyme at various concentrations and incubated at 37 °C for 6 h at 220 rpm. The OD₆₀₀ of the mixture was recorded after a 6-h subculture. The control group used 50 mmol.l^-1 Na₂HPO₄-Na₂HPO₄ filter-sterilized buffer (pH 6.2) instead of lysozyme. Antibacterial activity of the purified lysozyme to S. aureus was assessed using the antibacterial ratio. Antibacterial ratio was calculated using the following equation of R = (A - B) ÷ A × 100%; in which, R was the antibacterial ratio (%); A was the OD₆₀₀ value of the control group; and B was the OD₆₀₀ value of the experimental group. The antibacterial effect curve was plotted with the antibacterial ratio on the y-axis and the logarithmic of the purified lysozyme concentration on the x-axis. The IC₅₀ was defined as the purified lysozyme concentration, which resulted in a 50% decrease in the level of the antibacterial ratio, compared with untreated groups after a 6-h treatment. The IC₅₀ value was calculated using GraphPad Prism software. The cell morphology of S. aureus after a 6-h lysozyme treatment (IC₅₀ concentration) was reported using transmission electron microscopy (TEM).

2.5. Effects of heat treatment on anti-Staphylococcus aureus activity of the purified egg-white lysozyme

To assess the anti-S. aureus effects of the heat-treated egg-white lysozyme, a specific concentration of the purified lysozyme solution was incubated at various temperatures for a set duration before assessing its anti-S. aureus activity using water bath. In this study, three key parameters were primarily investigated, including heat treatment temperature, time and initial lysozyme concentration. The temperatures were set at 37, 50, 60, 70, 80, 90 and 100 °C, respectively. The treatment times were set at 5, 10, 15, 60, 120 and 240 min, respectively. The initial lysozyme concentration were set at 601.6, 1000, 2000, 4000 and 6000 μg.ml^-1, respectively. The three highlighted factors were optimized using one-factor-at-a-time method. After heat treatment, equal volumes of the lysozyme solution and S. aureus suspension were thoroughly mixed and co-cultured at 37 °C for 6 h at 220 rpm. The antibacterial ratio was then quantified. Untreated lysozyme was used as control group.

To assess the effect of heat treatment on lysozyme molecular structure, circular dichroism (CD) spectroscopy was used to analyze changes in its secondary structural components before and after thermal exposure. The CD spectra were assessed using Jasco J-1500 spectropolarimeter (Jasco, Japan) with a 1-mm cell in the far-UV region from 190 to 300 nm. The concentration of lysozyme was 0.1 mg.ml^-1 in 5 mmol.l^-1 phosphate buffer (pH 6.2).

2.6. Inhibitory effects of phenolic acids combined with heat-treated egg-white lysozyme on Staphylococcus aureus

2.6.1. Compatibility screening of various phenolic acids with egg-white lysozyme

In this experiment, the inhibitory effects of egg-white lysozyme combined with three phenolic acids (ferulic acid, caffeic acid and gallic acid) at various concentrations on S. aureus growth were assessed, respectively. The final concentrations of each phenolic acid were 200, 400, 800 and 1000 μg.ml^-1, while the final concentration of egg-white lysozyme was set at 300.8 μg.ml^-1. The anti-S. aureus activity analysis and antibacterial ratio calculation methods were based on those described in Section 1.4. For the control group, phenolic acids or lysozyme were replaced with 50 mmol.l^-1 Na₂HPO₄-Na₂HPO₄ buffer solution (pH 6.2).

2.6.2. Effects of the heat-treated lysozyme and ferulic acid combination on Staphylococcus aureus growth

The method for assessing the growth curves of S. aureus was based on Section 1.4. The heat treatment procedure for egg-white lysozyme was carried out according to Section 1.5, with the final concentration of heat-treated lysozyme adjusted to 300.8 μg.ml^-1. For the control group, ferulic acid or lysozyme were substituted with 50 mmol.l^-1 Na₂HPO₄-Na₂HPO₄ buffer solution (pH 6.2).

2.7. Numberized subsection inhibitory effects of Mycobacterium smegmatis acyltransferase combined with heat-treated egg-white lysozyme on Staphylococcus aureus growth

The mechanism by which, M. smegmatis acyltransferase (MsAcT) combined with heat-treated egg white lysozyme inhibited the growth of S. aureus is illustrated in Fig. 6.1. The preparation of the purified MsAcT was based on the methods described by Jia et al. [15]. Briefly, recombinant Escherichia coli BL21(DE3) strain was inoculated to Luria-Bertani (LB) broth and grown at 30 °C. The IPTG was added to culture broth to the final concentration of 1 mmol.l^-1, when the OD₆₀₀ reached 0.6–0.8. After 14 h interval, cell pellet was collected, resuspended using loading buffer (20 mmol.l^-1 pH 7.4 Na₂HPO₄-NaH₂PO₄, 20 mmol.l^-1 imidazole and 500 mmol.l^-1 NaCl) and then lysed using sonication. The supernatant from the cell lysate was collected and directly loaded on the HisTrap HP affinity chromatography column pre-equilibrated with loading buffer, respectively. Recombinant protein was eluted with a linear gradient of 20 ml of 20–500 mmol.l^-1 imidazole in the buffer with a flow rate of 0.8 ml.min^-1. The fractions with pure MsAcT were pooled and dialyzed against 20 mmol.l^-1 Na₂HPO₄-NaH₂PO₄ (pH 6.2) buffer overnight at 4 ^◦C. The hydrolysis activity of MsAcT to NAG was assessed according to Jiang et al. [16] and Muzzarelli and Rocchetti [17]. In brief, the reaction mixture contained 20 μg.ml^-1 NAG in 20 mmol.l^-1 Na₂HPO₄-Na₂HPO₄ buffer (pH 6.2). Appropriately diluted MsAcT was added into the reaction mixture to create a linear dependence of the reaction rate to protein concentration. The reaction was carried out at 37 ◦C and the kinetics was detected for 3 h at 202 nm.

The specific procedure for inhibiting S. aureus growth using MsAcT combined with heat-treated egg-white lysozyme was carried out as follows: MsAcT, heat-treated egg white lysozyme and buffer solutions were sterilized using 0.22-μm filters. Then, 200 μl of appropriately diluted log-phase S. aureus cells were mixed with 200 μl of 1 mg.ml^-1 MsAcT solution, while the control group received 200 μl of 50 mmol.l^-1 NaH₂PO₄-Na₂HPO₄ solution (pH 6.2). After incubation at 37 °C for 3 h, 200 μl of the heat-treated egg-white lysozyme (final concentration of 300.8 μg.ml^-1) were added to the experimental and control groups and mixed thoroughly, followed by incubation at 37 °C for 3 h. The samples were then centrifuged at 12,000 rpm for 3 min to separate the supernatant and the bacterial cell pellet. The protein concentration in the supernatant was measured to calculate the increase in protein, while the number of viable bacteria in the pellet was assessed according to the National Food Safety Standard of China (GB4789.2-2022, Food Microbiological Examination: Determination of Total Bacterial Count).

1. Results and Discussion

3.1. Purification and characterization of egg-white lysozyme

The purification of egg-white lysozyme was achieved through ammonium sulfate precipitation and CM-Sepharose fast flow column chromatography, resulting in 8.7-fold purification and yield of 41.0% (Table 1). Cation exchange chromatography verified particularly effective, accounting for 5.8-fold increase in specific activity. The purified enzyme was verified as homogenous using SDS-PAGE, showing a single band at 14.3 kDa (Fig. 1, Lane 6), which was similar to the molecular mass of egg-white lysozyme [18]. Its specific activity was assessed as 27,407.4 U.mg^-1 using M. lysodeikticus cells as substrate. This value was less than that reported by Chen et al. [18], a discrepancy likely attributable to the use of whole cells in this assay instead of the isolated cell walls used in the highlighted study.

A gradual loss of total activity was observed throughout purification. Particularly, lysozyme was detected in pellets at a relatively low ammonium sulfate saturation (40%), suggesting the potential formation of insoluble aggregates under these conditions, as previously documented [19, 20]. For practical uses, it is critical to state that the growth state of M. lysodeikticus, affected by culture conditions and equipment, significantly affects the assessed specific activity [11, 13, 21]. Therefore, standardizing the substrate by setting the growth curve and harvesting log-phase cells under local laboratory conditions is essential for accurate activity assessment. In this study, the buffer solution for the assessment of the egg-white lysozyme activity according to the National Standard of the People's Republic of China (GB/T 30990-2014) requirements included 50 mmol.l^-1 Na₂HPO₄-NaH₂PO₄ buffer (pH 6.2). In practical fields, particularly in the food processing industry, buffer solutions with pH 6.2 are rarely used.

Therefore, it is essential to assess optimal pH and temperature under the specific conditions necessary for the target uses, when assessing using effectiveness of egg-white lysozyme.

3.2. Anti-Staphylococcus aureus activity and mechanism of the purified egg-white lysozyme

The growth curve of S. aureus was characterized, identifying a log phase from 1 to 8 h (Fig. 2A), similar to the previous reports [22]. Cells from the exponential phase (2 h) were used for the assays. The IC₅₀ of native lysozyme against S. aureus was 300.8 μg.ml^-1 (Fig. 2B). The antibacterial efficacy of lysozyme is affected by the composition and sequence of bacterial cell wall peptidoglycan, as well as the physiological state of the enzyme [23, 24]. Lysozyme fights microbes through bacteriostatic, bactericidal and bacteriolytic mechanisms [25]. The transmission electron microscopy (TEM) images provided direct evidence of the bacteriolytic action, showing damaged S. aureus cell walls with distinct light/dark contrast and the collapse of cells, leading to leakage of intracellular contents (Fig. 2C, arrows). Despite this significant effect, the inhibitory activity of the native lysozyme is difficult to sustain over extended times (Figs. 2A, 5A), highlighting a limitation for its use as a standalone antimicrobial agent.

3.3. Enhancement of anti-Staphylococcus aureus activity using heat treatment and structural changes

Heat treatment under optimized conditions (600 μg.ml^-1, 90°C, 15 min, pH 6.2) enhanced the anti-S. aureus activity of lysozyme by 35.1%, achieving 82.5% inhibition compared to the native enzyme (Fig. 3A–C). This increase in activity after thermal denaturation was similar to that against other microbes such as SARS-CoV-2 and Bacillus subtilis [26, 27, 28]. Structural analysis revealed the reason behind this enhancement as a significant rearrangement of secondary structure occurred, with α-helix content decreasing from 35.59 to 23.60% and β-sheet, β-turn and random coil structures increasing (Fig. 3D, Table 2). This unfolding and proliferation of β-sheets drived the formation of fibrillar aggregates [30,29], which were postulated to perforate microbial membranes, a mechanism distinct from the native enzyme peptidoglycan hydrolysis [27]. This suggested that structural modification wa a viable strategy to improve the efficacy of lysozyme.

3.4. Synergistic anti-Staphylococcus aureus effects of heat-treated egg-white lysozyme and phenolic acids

From the phenolic acids (ferulic, caffeic and gallic acids), ferulic acid showed the highest inhibition (52.2%) at 200 μg.ml^-1, though differences diminished at higher concentrations (1000 μg.ml^-1), where all acids reached ~99% inhibition (Fig. 4A–C). Ferulic acid is reported to inhibit S. aureus by suppressing tetK and MsrA efflux pumps on the bacterial membrane [31,32].

A combination of 400 μg.ml^-1 ferulic acid with native lysozyme (300.8 μg.ml^-1) showed a synergistic effect, increasing inhibition by respectively 12.3 and 29%, compared to either compound alone (Fig. 4A). This synergy was further increased, when ferulic acid was combined with heat-treated lysozyme, reducing bacterial biomass (OD₆₀₀) by additional 18.7%, compared to the combination with native enzyme (Fig. 5A). The optimal protocol involved sequential addition of heat-treated lysozyme added at time zero, followed by ferulic acid after 6 h. This not only delayed the entry into the log phase by an additional hour but also decreased final biomass by 48.8% (Fig. 5B). This demonstrated that combining lysozyme with other antibacterial compounds, particularly after structural modification, could significantly enhance and prolong its inhibitory effect, potentially broadening its antibacterial spectrum [12,33]. However, when designing such combinations, concentration and addition sequence of egg-white lysozyme and phenolic acid, biocompatibility, reaction condition and differences in the mechanisms of action must carefully be addressed. Otherwise, adverse effects may occur. For example, in a solution with pH 9.0, phenolic acids (e.g., theaflavin) form covalent bonds with egg-white lysozyme, resulting in a 20% decrease in the antibacterial activity of egg-white lysozyme [12,33].

3.5. Synergistic inhibition using MsAcT and heat-treated lysozyme

The enzyme MsAcT hydrolyzes acyl groups on peptidoglycan structural units, theoretically facilitating subsequent hydrolysis by lysozyme (Fig. 6A). Experimental data verified this synergy. Treatment with heat-treated lysozyme alone increased the extracellular protein concentration by 0.115 mg.ml^-1, indicating cell lysis. When combined with MsAcT, the protein release increased further substantially by 0.208 mg.ml^-1 (Fig. 6B). This was verified by colony counts, which showed a decrease in viable cells from 1.69 × 10⁸ CFU.ml^-1 (lysozyme alone) to 1.41 × 10⁸ CFU.ml^-1 (combination treatment) (Fig. 6C). This is a significant finding, as bacterial deacylation of peptidoglycan is typically a resistance mechanism that decreases hydrophobic interaction with lysozyme [10]. However, the present results demonstrated that external enzymatic deacylation by MsAcT could instead sensitize cells to lysozyme. It is challenging to explain this novel result at the molecular level. Since the lysozyme used in this research was achieved through separation and purification from egg-white and its 3-dimensional (3D) structure has not been resolved, molecular docking techniques cannot be used to investigate the differential interactions between NAM/NAG and deacetylated NAM/NAG with the lysozyme. Additionally, MsAcT is a promiscuous biocatalyst possessing multiple catalytic activities such as hydrolytic activity and perhydrolysis activity [15]. Further in-depth investigation is needed to assess which catalytic activity dominates during its synergistic action with egg-white lysozyme in inhibiting S. aureus.

When egg-white lysozyme acts synergistically with other natural antibacterial agents to inhibit the growth of S. aureus, biocompatibility is one of the primary key parameters. Environmental conditions such as buffer pH [10,24], ionic strength [10,20,24] and temperature [14,27] of the egg-white lysozyme solution can alter its conformational state to various degrees; thereby, decreasing its antibacterial activity. Polyphenolic compounds may modify and disrupt the conformation of egg-white lysozyme, diminishing or eliminating its antibacterial activity [11, 12, 21, 31]. When other antibacterial proteins (e.g., proteases) act synergistically with egg-white lysozyme to inhibit the growth of S. aureus, the degradation of lysozyme by proteases can directly lead to its inactivation [14]. This study investigated the synergistic inhibitory effect of egg-white lysozyme and ferulic acid on growth of S. aureus under a sequential interaction mechanism, where two antibacterial agents were sequentially added to the reaction system to inhibit the bacterial growth. The experimental results indicated that the sequential interaction mechanism could partially overcome the issue of biocompatibility and maximize the antibacterial efficacy [Fig. 5B].

With the public in-depth understanding of the bacterial resistance mechanisms, growing awareness of environmental hazards caused by antibiotic misuse and heightened self-health consciousness, prohibition of antibiotics in animal feed and breeding processes has become a global interest. Therefore, it is vital to actively search for various alternatives to antibiotics [34, 35, 36]. From various antibiotic alternatives, egg-white lysozyme is a promising candidate due to its excellent biosafety. This study demonstrated that structural modification through heat treatment and strategic combination with other antimicrobials such as ferulic acid or enzymes such as MsAcT could significantly enhance its efficacy and time of action against S. aureus. Further studies should focus on clarifying the detailed mechanisms of these synergistic interactions and investigating other modification strategies such as glycosylation and lipophilization [10,24] to fully realize the potential of lysozyme as an effective antibiotic alternative.

1. Conclusion

In conclusion, this study successfully purified and characterized egg-white lysozyme and demonstrated that its anti-S. aureus activity could significantly be enhanced through thermal denaturation and synergistic combinations. Heat treatment induced structural reorganization of lysozyme, increasing its β-sheet content and enabling membrane-perforating antibacterial mechanisms distinct from its native lytic activity. Furthermore, combining heat-treated lysozyme with ferulic acid or MsAcT resulted in the synergistic effects, delaying bacterial growth and improving bacteriolytic efficacy. The results highlighted the potential of modified lysozyme as part of integrated natural preservation systems to improve food safety. Further studies should focus on explaining detailed mechanisms of synergy and investigating other modification methods to further broaden the applicability and effectiveness of lysozyme-based antimicrobial interventions. To better promote this process in the field of food processing, follow-up research should thoroughly investigate its use effects, affecting factors and long-term stability in real food matrices (e.g., meat and dairy products). Additionally, a comparison of production costs and profitability with other antibacterial processes is necessary.

1. Acknowledgements

This research was supported financially by the National Natural Science Foundation of China (nos. 31370802 and 31870787) and Fujian Province Natural Science Foundation (no. 2021J01169).

1. Declaration of competing interest

The authors declare no conflict of interest.

1. Authors’ Contributions

Conceptualization, W.H. and X.X.; methodology, W.H.; software, Q.W.; validation, L.J., .J.W. and Q.W.; formal analysis, W.H. and S.Z.; investigation, X.X.; resources, J.W.; data curation, L.J.; writing—original draft preparation, W.Y.; writing—review and editing, S.Z.; supervision, S.Z.; project administration, H.J.; funding acquisition, S.Z.

1. Using Artificial Intelligent Chatbots

No artificial intelligent chatbots have been used in sections of this manuscript.

1. Ethical Consideration

The subjects of this study are egg white lysozyme and Staphylococcus aureus. As no animal or human trials are involved, no ethical review documentation is required.

Background and Objective: Escherichia coli O157:H7, a major foodborne bacterial pathogen, is commonly detected in the intestines of healthy cattle, deer, goats and sheep. It can readily contaminate veal, ground beef and hamburgers, causing a range of human diseases, including mild diarrhea, hemorrhagic colitis, hemolytic-uremic syndrome and thrombotic thrombocytopenic purpura. Various treatment strategies are used to reduce E. coli O157:H7 contamination, ranging from washing to chemical and physical food decontamination. The increase of antibiotic-resistant bacteria has hence increased interest in lytic bacteriophages as biocontrol agents. This study investigated the use of bacteriophages to control E. coli O157:H7 growth in experimentally contaminated meat.

Material and Methods: Six E. coli O157:H7 specific bacteriophages were isolated from sewage, purified via plaque assay and assessed individually and in cocktails for lytic activity at various multiplicity of infection rates. Temperature-dependent effects on adhesion and lytic efficiency were assessed at refrigeration and ambient conditions. Phages were further used in commercial polyethylene food wraps to assess realistic surface delivery performance on inoculated meat samples and long-term phage stability was monitored during storage at 4 °C.

Results and Conclusion: Single-phage treatments achieved up to 100% bacterial elimination at higher multiplicities of infection, while optimized phage cocktails completely eradicated bacteria. Temperature affected bacterial recovery, with refrigeration decreasing bacterial adhesion and slightly slowing phage-mediated lysis. Importantly, this study has introduced an innovative practical approach using phages onto commonly used commercial polyethylene plastic wraps, demonstrating 99 and 93% reductions when phage‑treated wraps were used after 20 min and 24 h storages, respectively. Long-term storage of phage tf1 at 4 °C for seven months demonstrated stable plaque-forming ability, verifying durability for practical uses. These results highlight bacteriophages as effective, safe and scalable interventions to control E. coli O157:H7 in meat products, providing a practical strategy to enhance food safety, particularly for pathogens with extremely low infectious doses.

Keywords: Bacteriophage biocontrol, Escherichia coli O157:H7, Food safety intervention, Meat contamination, Phage cocktail therapy, Polyethylene wrap application

1. Introduction

Escherichia coli O157:H7 was identified in 1983 and has since been addressed as a major foodborne bacterial pathogen affecting humans [1, 2]. It is commonly detected in the intestines of healthy cattle, deer, goats and sheep and can therefore readily contaminate veal, ground beef and hamburgers [3]. The E. coli O157:H7 produces Shiga toxins and is classified as an enteric bacterium responsible for diarrhea and dysentery, also known as enterohemorrhagic E. coli (EHEC) [4]. It causes a range of human diseases, including mild diarrhea, hemorrhagic colitis (HC), hemolytic-uremic syndrome (HUS) and thrombotic thrombocytopenic purpura (TTP) [5]. Although other foodborne pathogens such as Listeria, Salmonella enterica Serovar Enteritidis and Shigella species have been isolated from meats, E. coli O157:H7 is particularly concerning due to its extremely low infectious dose, as few as 10 cells, and its ease of transmission during slaughter and meat processing [6]. The microorganism can spread directly from person to person, especially in child day-care settings and from animals to humans [7].

Globally, E. coli infections are estimated to cause 2.8 million cases annually, with approximately 63,000 cases of HC reported each year in the United States alone. Children are particularly susceptible to severe complications such as HUS, a leading cause of acute renal failure (ARF) [8]. Numerous studies have investigated the prevalence of E. coli O157:H7 in food products and the bacterium has been isolated from a variety of plant and animal-derived foods within several countries. Contamination is frequently associated with meat products due to exposure to intestinal contents during slaughter and processing. However, contamination has been reported in raw fruits and vegetables, often linked to the use of untreated animal waste or contaminated irrigation water. Reported prevalence rates vary by country and food type, but the detection of the pathogen in ready-to-eat foods highlights the risk of foodborne illness outbreaks [9]. From 2003 to 2012, 390 outbreaks of Shiga toxin-producing E. coli O157 in the United States led to 4,928 diseases, 1,272 hospitalizations and 33 deaths, with cattle serving as the primary reservoir and ground beef as the most common route of transmission. Undercooked ground beef and other non-intact cuts pose a high risk, making it critical to decrease contamination during beef production, processing and storage and to protect consumers through appropriate handling and cooking of raw beef products [10].

Despite the urgent need to control E. coli O157:H7 contamination, current treatment methods vary in effectiveness and present challenges. To effectively decrease or eradicate E. coli O157:H7 contamination, a variety of treatment methods are currently used, ranging from simple washing of foods to chemical (peroxyacetic acid, chlorine dioxide, sodium hypochlorite and organic acids) or physical (irradiation, pasteurization, high-pressure processing and pulsed electric fields) decontamination techniques. These methods vary in their efficacy, costs and effects on the sensory quality and visual appeal of foods. However, their practical use is limited because several of these treatments can modify texture, flavor and nutritional attributes and some also increase concerns associated to residue safety and toxicity. Furthermore, they are not universally usable within all food types, which limits their feasibility and consistency in real processing environments [11]. Although antibiotics are used as effective chemical agents for removing these bacteria, increasing reports of antibiotic-resistant bacteria suggest that an alternative effective method is needed [12]. The growing threat of multidrug-resistant bacterial pathogens necessitates urgent investigation of alternatives to traditional antibiotics [13].

Regarding the limitations of current treatments, bacteriophage therapy, first developed by Felix d’Herelle and historically overshadowed by antibiotics, has re-emerged as a promising alternative to traditional antibiotics. This is due to its specificity, ability to target biofilms, minimal side effects and potentials to decrease antibiotic resistance development [14]. In food safety, bacteriophages offer unique advantages because of their narrow host range, which enables them to eliminate pathogenic E. coli strains without harming beneficial microbiota [15]. Coliphages, in particular, are abundant in environments such as sewages, animal wastes and water sources, making them accessible for isolation and use against foodborne pathogens [16]. One challenge to their use is the possible emergence of phage-resistant bacteria; however, this can be addressed through the use of phage cocktails composed of multiple phages with complementary host specificities. Such formulations not only broaden the range of susceptible bacterial targets but also decrease the likelihood of resistance development. Moreover, phages include the ability to evolve in bacteria, allowing them to overcome resistance mechanisms that limit the effectiveness of antibiotics. These characteristics have provided a way for regulatory acceptance of several phage-based products in the food industry, particularly for the control of E. coli O157:H7 in beef and fresh products, highlighting their potential as sustainable and innovative tools to enhance food safety [17].

Bacteriophages have been recognized as safe for use in food production and are classified as generally recognized as safe (GRAS) by the U.S. Food and Drug Administration (FDA). While phages targeting E. coli O157:H7 such as EcoShield and SecureShield E1 have received GRAS approval for use in beef and other food products, similar approvals exist for phages against other foodborne pathogens. The GRAS phages are valued for their specificity, safety for human consumption, minimal effects on the food sensory characteristics and ability to decrease pathogen levels without contributing to antimicrobial resistance [18]. This research investigated the use of bacteriophages as biocontrol agents to control the propagation of artificially contaminated E. coli O157:H7 in meat samples.

1. Materials and Methods

2.1. Bacteria cultivation

Lyophilized E. coli O157:H7 PTCC 1860 was provided by the Persian Type Culture Collection (PTCC), Iran. The strain was cultured on tryptic soy agar plates at 37 °C for 16–18 h. For storage, bacterial cultures were suspended in tryptic soy broth (TSB) with 15% glycerol and stored at -80 °C. Bacterial counts were determined as colony-forming units (CFU) defined as viable bacterial cell capable of forming a visible colony on solid media.

2.2. Isolation and purification of bacteriophages

Bacteriophages were isolated from environmental water, sewage and slaughterhouse sewage samples. Briefly, 100 mL of sewage samples were collected in sterile bottles and transported to the laboratory. To enrich bacteriophages, samples were mixed with 200 mL of TSB media and inoculated with the E. coli host and then incubated overnight at 37 °C. Chloroform was added to the mixture for 10 min to lyse the bacterial cells, followed by centrifugation to remove debris. The resulting supernatants, containing potential bacteriophages, were assessed against host bacteria using double-layer agar method [19]. After overnight incubation at 37 °C, plates were investigated for plaques. Single plaques were isolated and subjected to repeated culture using double-layer agar technique to purify individual bacteriophages. Isolated bacteriophages were stored in SM buffer at 4 °C. Phage titers were expressed as plaque-forming units (PFU), where one PFU corresponded to an infectious phage particle that produced a visible plaque on a bacterial lawn.

2.3. Contaminated-meat experiments

Fresh beef was purchased from a local butchery and cut into similar pieces. Their sterility was assessed before use. For experimental contamination, E. coli O157:H7 was cultured overnight and a bacterial suspension was prepared at a concentration of 10⁷ CFU.mL^-1. Meat samples were immersed in the suspension and agitated for 10 min to allow bacterial adhesion. The multiplicity of infection (MOI) was reported as the ratio of phage particles (PFU) to bacterial cells (CFU).

2.3.1. Single Phage Treatment

Meat samples were divided into three groups and suspended in sterile water, including untreated meat (control), meat treated with E. coli O157:H7 at 10⁷ CFU.mL^-1 (bacteria only), meat contaminated with E. coli O157:H7 (10⁷ CFU.mL^-1) and subsequently treated with a specific bacteriophage at 10⁸ PFU.mL^-1 (phage-treated). All samples were agitated for 20 min; after which, aliquots were serially diluted (10-¹, 10-² and 10-³) for plating on eosin methylene blue (EMB) agar. Sampling and culturing were repeated after 60 min to assess bacterial decrease over time. The entire procedure was carried out at room temperature (RT) and 4 °C to assess the effect of temperature on phage efficacy.

2.3.2. Phage Cocktail Treatment

A mixture of three bacteriophages (TF1, TF2 and SC) was used in meat samples inoculated with E. coli O157:H7 (10⁷ CFU.mL^-1) at a total phage concentration of 10⁸ PFU.mL^-1. To further assess the cocktail performance, an additional trial was carried out; in which, each phage was used individually at 10⁸ PFU.mL^-1.

2.4. Contaminated ground beef experiments

Ground beef was prepared from the meat under sterile laboratory conditions and divided into three portions of equal weight. One portion was untreated as a control. The second portion was inoculated with E. coli O157:H7 at a concentration of 5 × 10⁸ CFU.mL^-1 without further treatment. The third portion was similarly contaminated with the bacteria, agitated for 20 min to ensure homogenized distribution and then treated with a bacteriophage suspension at 10⁸ PFU.mL^-1. These experiments were carried out at RT and 4 °C.

2.5. Plastic wrap treatment with bacteriophages

Beef samples were covered with plastic wrap under various conditions. The wrap included commercial food-grade low-density polyethylene available in the market. In the control setup, uncontaminated beef was wrapped with untreated plastic. A second setup involved contaminated beef wrapped with plastic that did not contain bacteriophages. In the phage-treatment setup, plastic wrap was pretreated with a suspension containing 10⁹ PFU.mL^-1 bacteriophages by spraying and storing at 4 °C for 20 min and then used in contaminated beef for 30 min at 4 °C. To further assess the protective effect of wraps over time, additional plastic sheets were incubated at 4 °C for 24 h after sprayed with the phage suspension followed by covering beef samples for 30 min.

2.6. Statistical analysis

All experiments were carried out in three replicates (n = 3). Quantitative data were presented as mean ± SD (standard deviation). Error bars in figures represented SD of the replicates.

1. Results and Discussion

Four bacteriophages were isolated from municipal sewage and two others from slaughterhouse wastewater. These were designated as tf1-tf4 and sc-sg, respectively. Of the six bacteriophages isolated, tf1 was chosen for further individual assessments based on its distinct plaque morphology. This produced large, clear and well-defined plaques, suggesting high lytic efficiency and strong bactericidal potential against E. coli O157:H7. In contaminated beef experiments, when 10⁶ PFU.mL^-1 of tf1 bacteriophage was used at RT against 10⁷ CFU.mL^-1 of E. coli O157:H7, 65% ±0.5  of bacteria were removed after 20 min (MOI = 0.1) at RT. Increasing phage dose up to 10⁹ PFU.mL^-1 led to 99% ±0.4  bacterial elimination (MOI = 100), as shown in Figure 1.

Exposure with phage for 60 min at RT, addition of 10⁷ PFU.mL^-1 of phage to 10⁶ CFU.mL^-1 of bacteria resulted in nearly 93.5% ±0.4 elimination (MOI = 10). However, bacteria were more effectively removed, when treated by 10⁸ PFU.mL^-1 phage even up to 99% ±0.2 (MOI = 100). Elimination rate was 100%, when 10⁹ PFU.mL^-1 of phage was added to 10⁷ CFU.mL^-1 of bacteria (MOI = 1000), as shown in Figure 2. Bacterial decreases are shown at the second y axis (right y axis).

For assessing the effect of phage cocktail, including tf1, tf2 and sc, cocktails with two various formulas were assessed. Meat containing 10⁷ CFU.mL^-1 bacteria was exposed to a dose of 10⁸ PFU.mL^-1 cocktail phages and showed nearly 99% ±0.1 of bacteria being lysed after 20 min and 99.5% ±0.1 after 60 min. When a cocktail included 10⁸ PFU.mL^-1 separately, bacterial removal was 100%; as shown in Figure 3. Comparison of performance between single-phage (Figure 1-2) and cocktail treatments (Figure 3) indicated that while tf1 alone achieved strong bacterial decrease at higher MOIs, the cocktails demonstrated enhanced efficacy and reliability in reaching levels below infectious (< 100 CFU) and ingestion (< 10 CFU) thresholds. The optimized cocktail not only maintained 99% bacterial elimination but achieved complete eradication when used at equivalent individual phage titers, demonstrating superior robustness in ensuring bacterial counts fall below clinically relevant exposure limits.

In ground beef experiments, meat samples were inoculated with E. coli O157:H7 at an initial concentration of 10⁷ CFU.mL^-1 and then treated with 10⁸ PFU.mL^-1 of bacteriophages for either 20 or 60 min at 25 and 4 °C. Bacterial recovery from the meat was affected by temperature, as adhesion decreased at refrigeration temperature, compared to ambient conditions, consistent with the lower metabolic activity and decreased surface interactions of bacteria at colder environments. After 20 min of phage exposure, bacterial counts decreased by 96.4% ±0.1 at 25 °C and 82.0% ±0.4 at 4 °C (Figure 4). When exposure time was extended to 60 min, decrease rates increased to 99.95% ±0.05 and 87.0% ±0.5 at 25 and 4 °C, respectively, indicating that phage-host interactions were kinetically dependent, with higher temperatures favoring further efficient adsorption and subsequent lysis. These findings suggested that rapid refrigeration of meat could decrease infection risk by limiting bacterial attachment, even though phage-mediated lysis occurred further slowly at lower temperatures

Contaminated meat samples were covered for 30 min with plastic wraps coated with 10⁹ PFU.mL^-1 phage for 20 min and 24 h. significant viable cell decreases, at the two conditions were observed with the greatest decrease in the 20 min-treated plastic wrap, compared to controls without phages. This included 99.0% ±0.9, compared to 93% ±2 in 24 h-treated group; as shown in Figure 5.

To assess the long-term stability of phage tf1, its plaque-forming ability was assessed over a 7 m storage at 4 °C. At a 10⁻⁷ dilution, plaque counts were consistent within all measured time points, showing no appreciable decreases in infectivity (Figure 6). The average plaque count calculated from all sampling intervals was stable, further verifying the persistence of active phage particles during prolonged storage. Minor deviations observed between the months were included within the expected range of experimental variability, which might be seen due to the counting precision or inherent assay fluctuations. Overall, these results indicated that phage tf1 preserved its infective capacity and structural integrity for at least 7 m under refrigerated conditions, highlighting its strong potential for use in long-term biocontrol uses.

The EHEC are pathogenic E. coli strains that produce Shiga toxins and cause HC and its life-threatening sequelae of HUS [20]. Since identification of serotype O157:H7 in 1983, many outbreaks associated with EHEC have been reported in the United States and E. coli O157:H7 has become one of the most important foodborne pathogens [21]. It causes severe human disease worldwide. It seems important to eliminate this pathogen in appropriate way [5]. Antibiotics were always an option; however, with increasing antibiotic-resistant strains due to use of antibiotics in cow feed, an alternative treatment is needed [22].

In a study, the overall prevalence of Shiga toxin-producing E. coli in raw beef samples collected from an industrial slaughterhouse in Hamadan, Iran, was 10.4%, with one isolate (12.5%) belonged to the O157 serogroup and the rest isolates of 87.5% classified as non-O157. The prevalence of O157 (1.3%) was similar to that reported in Iran (e.g., 1.35% in Mashhad), greater than those reported from the EU (0.3%), the U.S. (0.8%) and Brazil (1%) and less than those reported in Nigeria (2.2%) and Australia (1.7%). The study verified cattle as a reservoir for Shiga toxin-producing E. coli and highlighted seasonal and animal-associated factors such as spring sampling and the use of feeder cattle, which might contribute to the prevalence [23]. Additionally, they reported that 5.3% of raw poultry meat samples were contaminated with E. coli O157 strains. Out of 257 samples, 36% of the samples contained E. coli isolates, with 38.7% of the isolates identified as Shiga toxin-producing E. coli carrying virulence genes such as stx1, stx2 and eae. The isolates showed significant genetic diversity and high resistance to antibiotics, including nalidixic acid, tetracycline, ampicillin and trimethoprim-sulfamethoxazole [24].

Since discovery of bacteriophages, they included therapeutic potential as natural enemies of bacteria and phase-based elimination of bacteria in agriculture and food industry was started [25]. Therefore, bacteriophages have been reported as good antibiotic alternatives. They provide a highly specific, environmental-friendly effective means of controlling bacterial pathogens in agriculture; thereby, contributing to efforts against the increase of antibiotic resistance in the food sector [26].

This study aimed isolation of effective bacteriophages for biocontrol of E. coli O157:H7 in beef and ground beef. Six phages were isolated and assessed in suspension and spray methods on beef and beef ground samples at RT and 4 °C. The use of the submersion method to assess bacterial attachment to meat showed less bacterial recovery at 4 °C, compared to 25 °C, verifying that adhesion and phage inactivation efficiency were modulated by temperature, likely due to differences in bacterial physiology and phage adsorption kinetics. These findings highlighted the importance of environmental conditions and exposure time in optimizing phage-based biocontrol strategies in meat systems. Furthermore, the effect of various MOIs were assessed after incubation times of 20 and 60 min. The results showed that the phage treatment eliminated 82 to 100% of E. coli O157:H7. Phage cocktails included a better elimination rate. In a similar study, a cocktail of all three phages was assessed for their ability to lyse the bacterium in vivo and in vitro. The phage cocktail totally eradicated E. coli O157:H7 from beef surfaces in seven out of nine experiments. They have isolated phages from bovine farmyard slurry samples. The study achieved superior results when using phage cocktails similar to those used in this study, suggesting that a combination of multiple phages could enhance the effectiveness against the target bacteria. Another bacteriophage cocktail of three lytic phages for E. coli O157:H7 was investigated for its ability to decrease experimental contamination of hard surfaces, tomato, spinach, broccoli and ground beef by three bacteria. The observed decreases ranged from 94 to 100%. This suggested that naturally occurring bacteriophages might be useful for decreasing contamination of various hard surfaces, fruits, vegetables and ground beef by E. coli O157:H7 [10]. The E. coli O157:H7 is addressed for its extremely low ingestion (number of viable microorganisms that must be consumed via contaminated food or water orally to establish infection in a susceptible host) and infectious dose (the smallest number of viable microorganisms needed to cause infection in a host). Epidemiological evidence indicates that ingestion of fewer than nine viable cells can cause infection, with most reports estimating an infectious dose of fewer than 100 CFU [27]. This exceptionally low threshold highlights the importance of complete bacterial elimination in any antimicrobial intervention [28]. In the present study, the initial bacteriophage cocktail achieved a 99.95% decrease in E. coli O157:H7 counts, creating approximately 730 CFU. In contrast, the optimized cocktail achieved complete eradication (no detectable CFU). Regarding that single-digit cell numbers can lead to infection, the ability to decrease bacterial levels to less than the detection limit represents a critical improvement in ensuring microbiological safety.

Extending previous research on phage-based biocontrol strategies, the current study introduced a novel approach by spraying bacteriophage preparations onto plastic wrap used in cover meat containers, demonstrating that phage could effectively decrease E. coli O157:H7 contamination. This study assessed the stability of bacteriophage-coated plastic wrap under refrigeration. Wraps stored for 20 min or 24 h before contact with E. coli O157:H7 significantly decreased bacterial counts, with 99 and 93% decreases, respectively. These results indicated that phages were active on the wrap for at least 24 h, though shorter storage preserved slightly higher lytic activity. While phage uses in food safety such as sprays, edible films and integration into packaging have been reported in the literature [29, 30], the use of phage-coated plastic wrap directly covering meat surfaces is novel. This method offers a practical potentially-scalable strategy for controlling pathogenic contamination, particularly for pathogens such as E. coli O157:H7 with very low infectious doses, using a simple, economic readily available packaging material. The results showed that brief exposure to phage-coated surfaces could substantially decrease bacterial load, highlighting its relevance for enhancing food safety in processing and retail settings.

Previous studies have investigated the use of bacteriophages as natural antimicrobial agents in active or edible packaging by incorporating phages into biopolymer films or coatings [31]. Several studies have used phages within chitosan and PVA matrices for suppressing E. coli on food surfaces, demonstrating decreases in bacterial load on beef, poultry and other substrates [32-34]. These film or coating-based approaches, while effective, often need specialized formulation and fabrication and may face challenges such as controlled phage release, stability during storage and ensured sufficient contact with the food surface. In contrast, the present approach, spraying phage preparations directly onto standard commercial plastic wrap covering meat containers, eliminates the need for film casting or coating, enabling immediate phage contact with contaminated surfaces. This method can be rapidly implemented within existing packaging workflows without modifying materials or supply chains and the large decreases observed in these experiments suggest that sprayed phages may achieve antimicrobial effects on phage-embedded films while offering greater operational simplicity and flexibility.

Monitoring the viability of the tf1 phage over 7 m demonstrated long-term stability, with plaque numbers showing a slight increase during storage. This modest increase might reflect the gradual disaggregation of phage particles over time and further verifies that the phage preserves its stability and plaque-forming ability under standard storage conditions.

Bacteriophages can be limited by their narrow host range, complexity of formulating effective cocktails and decreased diffusion and activity within dense food matrices such as meats. Their stability is affected by environmental conditions and bacterial resistance may increase through receptor modifications. Additionally, regulatory restrictions and consumer perception challenges are difficulties to widespread adoption in food systems. In contrast to FDA, the European Food Safety Authority (EFSA) provides a further restrictive stance and bacteriophages are not currently approved as food additives or antimicrobials within the EU due to unresolved concerns regarding genetic stability, host-range specificity and safety classification [35]. This regulatory disparity highlights the need of continued investigation into phage safety, efficacy under practical food-processing conditions and standardized assessment frameworks to support future approval.

1. Conclusion

Use of bacteriophages is an effective strategy for controlling pathogenic E. coli O157:H7 in the beef and ground beef industry. Suspension and spraying methods demonstrated promising results in decreasing bacterial contamination. The ability to target dangerous pathogens with phages that are relatively easy to isolate highlights their potential as a practical and safe biocontrol tool. Widespread use of bacteriophages could enhance food safety by decreasing pathogenic E. coli contamination and should further be investigated for integration into standard food processing protocols, including their incorporation into sanitation procedures, formulation optimization and development of practical delivery methods appropriate for industrial use. The findings of this study represent an early-stage investigation of phage use in beef packaging systems. While the decreases are highly encouraging, continued research under diverse storage conditions, product types and scaling environments is needed before full industrial adoption. These preliminary results support further development of phage-based interventions as a promising addition to meat safety strategies.

1. Acknowledgements

The authors gratefully acknowledge the Microbiology Laboratory staff at Tabriz University, Ms. Fatemeh Khodayi and laboratory colleagues for their technical support and assistance throughout this study.

1. Declaration of competing interest

The authors declare no conflict of interest associated to this study.

1. Authors’ Contributions

Investigation, methodology, data curation, HMG; original draft preparation, editing and review, analysis, visualization, RR; supervision, conceptualization, FSH, supervision, conceptualization, methodology, formal analysis, review, resources, GZ.

1. Using Artificial Intelligent Chatbots

No AI assistance was used.

1. Ethical Consideration

No ethical approval was required for the conduct of this study

Abstract

Background and Objective: Probiotic whey-apple juice fermented beverages are becoming popular for their health benefits and incorporating bioactive ingredients such as licorice extract and caffeine enhances their functionality and consumer appeal. This study aimed to assess the physicochemical, microbial and sensory characteristics of a whey-apple juice based probiotic fermented beverage enriched with licorice extract and caffeine during storage.

Material and Methods: A probiotic beverage was formulated with various licorice extract (0–0.6%) and caffeine (60–100 mg.100 ml^-1) levels and stored for 1, 15 and 30 d. This was analyzed for pH, total soluble solids, viscosity, color, phenolic content, antioxidant activity, probiotic viability and sensory traits. A three-factor factorial design with statistical analysis was used for all experiments.

Results and Conclusion: Higher licorice extract and caffeine levels significantly increased Lactobacillus plantarum viability (from 6.27 to 7.79 CFU.ml^-1), phenolic content (from 125.87 to 164.16 mg GAE·g⁻¹), antioxidant activity (IC₅₀: from 145.46 to 112.26 μg.mg^-1) and viscosity (from 142.62 to 168.36 mPa.s^-1) while decreasing pH (from 4.26 to 4.17) and lightness (from 53.74 to 46.1) (p<0.05). The two additives altered color values, with a significant increase in redness only for licorice extract. The sample with 0.3% licorice extract and 80 mg caffeine included the highest sensory acceptability. Licorice extract and caffeine can improve the functional, physicochemical and sensory qualities of probiotic beverages during storage.

Keywords: Antioxidants, Caffeine, Fermented beverage, Glycyrrhiza, Lactobacillus plantarum, Probiotics

1. Introduction

Whey, a by-product of global cheese and casein production, represents environmental pollutant and valuable nutritional resource due to its high production volume and organic load. Rich in essential amino acids, peptides and proteins, whey offers numerous health benefits, including antihypertensive, antioxidant, anti-inflammatory and antimicrobial effects. Efficient use of whey supports sustainable waste management and creation of value-added functional foods [1-3]. Apple juice is a nutrient-rich beverage containing natural sugars, vitamins particularly vitamin C, minerals such as potassium and antioxidants such as polyphenols. Combining whey protein with apple juice not only improves the beverage taste but also enhances its nutritional profile by adding natural sugars, vitamins and antioxidants. Additionally, whey serves as an excellent base for probiotic beverages, promoting gut health and overall wellness [4, 5].

Licorice (Glycyrrhiza uralensis) is a widely used medicinal herb characterized by a diverse array of bioactive compounds, including prominent flavonoids such as liquiritigenin and isoliquiritigenin. Liquiritigenin, a primary constituent, shows multiple pharmacological activities encompassing antioxidant, anti-inflammatory, antibacterial, antidepressant and anxiolytic effects. Caffeine, the principal active ingredient in energy beverages, acts as a central nervous system (CNS) stimulant, enhancing alertness and cognitive function. The incorporation of licorice extract and caffeine significantly augments the functional attributes of the beverage. Collectively, these ingredients exert a synergistic effect, enhancing health-promoting potential and sensory qualities of the formulation [6-8].

Previous studies have documented the development and quality assessment of apple-whey based herbal functional ready-to-serve beverages [4,9]. For example, Sharma et al. [4] formulated a low-calorie beverage by mixing 75% apple juice with 25% whey protein, achieving the desired protein content. Additionally, kombucha beverages incorporating licorice and ginger have demonstrated strong antioxidant capacities and increased phenolic content, enhancing their functional characteristics [6]. Moreover, a non-alcoholic mixed beverage, containing coconut water, cashew apple juice and caffeine, has successfully been developed [10]. Despite the growing interest in functional probiotic beverages, no study has investigated the integration of probiotic cultures into whey-apple juice systems enriched with licorice extract and caffeine. This gap creates unresolved questions regarding microbial stability, physicochemical characteristics, sensory acceptance and health-promoting potential. Conventional formulations often suffer from decreased probiotic viability due to acidic fruit components and lack of natural synergistic stabilizers. The present study addressed this challenge by developing a novel probiotic apple-whey beverage fortified with licorice extract (0, 0.3 and 0.6%) and caffeine (60, 80 and 100 mg). This innovative combination affected licorice polyphenols and caffeine stabilizing effects to enhance probiotic viability, while assessing physicochemical, microbial and sensory attributes during storage.

1. Materials and Methods

2.1. Materials

Caffeine powder and carboxymethylcellulose (CMC) were purchased from Merck, Germany and licorice root was purchased from Shirin Darou, Shiraz, Iran. Lactobacillus plantarum (L. plantarum) PTCC 1058 cultures were provided by the Iranian Research Organization for Science and Technology (IROST, Tehran, Iran). The lyophilized strain (≥ 10⁸ CFU.ml^-1) was propagated in MRS broth for 24 h at 37 °C using CO₂ incubator (Memmert, Munich, Germany). Man, Rogosa and Sharpe (MRS) agar and broth were supplied by Merck, Darmstadt, Germany. All reagents in this study included analytical grade and were purchased from Merck, Germany.

2.2. Licorice Extract Processing

Licorice root was washed, dried and ground into powder, which was then extracted with 70% ethanol through a 72-h shaking process. The extract was filtered to remove residues, concentrated using rotary evaporation and dried to achieve the final extract [11].

2.3. Apple Juice Processing and Conservation

The apple juice was prepared by thoroughly washing and sorting the fruits, followed by cutting and grating. Apple juice was achieved using electric juicer and then passed through a muslin filter. The extracted juice was pasteurized in glass containers via heating and stored at room temperature (RM) until further use [9].

2.4. Whey Preparation from Cow Milk

Fresh cow milk was processed using method of Sharma et al. [9]. The milk was heated to 82 °C for 10 min, then cooled to 70 °C and acidified with 2% citric acid under continuous stirring to induce casein coagulation. The clotted mixture was passed through a double layer of muslin cloth and the achieved whey was centrifuged at 5000 rpm for 10 min to eliminate residual fat prior to its use in product formulation.

2.5. Production of Probiotic Beverage

Apple juice and whey were heated to 70 °C for 15 min and filtered and then mixed at 60:40 ratios. Carboxymethyl cellulose (0.5 g per 100 ml), caffeine (60, 80 and 100 mg per 100 ml) and licorice extract powder (0, 0.3 and 0.6%) were added to the mixture. The mixture was cooled to 37 °C and inoculated with a fresh L. plantarum probiotic culture at an initial level of approximately 10⁹CFU.ml^-1. Fermentation was carried out at 37 °C until the beverage reached a target pH of 4.5. The final pH after fermentation was assessed at 4.3 ±0.1. After fermentation, samples were stored at 4 °C. Experiments were carried out on Days 1, 15 and 30 to monitor physicochemical characteristics, microbial viability and sensory attributes of the beverage [12].

2.6. Physicochemical Analysis of Probiotic Beverage

During the storage time, probiotic beverage samples were analyzed for pH, total soluble solids (TSS) and viscosity. The pH and TSS were assessed using digital pH meter (AZ 86502, Taiwan) and digital refractometer (RX-7000α, India), respectively. The apparent viscosity was assessed using Brookfield viscometer (DV II + LV, Brookfield, Middleboro, MA, USA) equipped with an LV2 spindle. Viscosity was expressed in mPa_·s^-1 and assessed at 20 °C within fourteen spindle speeds ranging 1.5-2000 rpm; the primary reported values corresponded to 60 rpm. Assessments were carried out in triplicate to enhance reproducibility [9].

2.7. Color Assessment of Probiotic Beverage

The color characteristics of the probiotic beverage were assessed using HunterLab UltraScanvis US-Vis 1,310 device, USA, which assessed L*, a* and b* color values. The lightness parameter (L*) ranged from 0, representing black, to 100, representing white, a* scale assessed color on a spectrum from +127 (red) to -128 (green), while the b* scale ranged from +127 (yellow) to -128 (blue) [13].

2.8. Assessment of Phenolic Levels and Antioxidant Potential

Probiotic beverage samples were centrifuged and filtered prior to analyzing total phenolic content (TPC) using Folin-Ciocalteu method. The TPC was assessed using Folin-Ciocalteu method. A calibration curve was prepared with Gallic acid standards (0–200 mg.l^-1) and results were expressed as Gallic acid equivalents (GAE) per liter. Each assessment was carried out in triplicate. Antioxidant activity was assessed using DPPH radical scavenging assay. Sample extracts at various concentrations were reacted with 0.1 mM DPPH solution and absorbance was assessed at 517 nm after 30 min incubation in dark. The inhibition percentage was calculated and IC₅₀ values were achieved from a dose-response curve generated by plotting inhibition percentages against sample concentrations. All assessments were carried out in triplicate to ensure reproducibility [9].

2.9. Assessment of Probiotic Bacterial Survival

For microbiological assessment, 10 ml of the sample was mixed with 90 ml of sterile saline to achieve a 10^-1 dilution, followed by preparation of successive serial dilutions. Enumeration of L. plantarum was carried out by spreading the diluted samples onto MRS agar supplemented with vancomycin and incubating at 37 °C for 72 h using CO₂ incubator. The results were reported as logarithmic colony-forming units (Log CFU) per milliliter [14-15].

2.10. Sensory Analysis

A sensory analysis was carried out with a panel of 12 trained participants (six men and six women, aged 20–30 y) using five‑point hedonic scale ranging from 1 (“extremely dislike”) to 5 (“extremely like”). The panelists assessed the probiotic beverage samples for color, flavor, texture and overall acceptability during a 30‑d storage time. For each session, 20-ml portions of the beverage were present on coded plates, set at 4°C ±1 and panel members rinsed their mouths with water between the assessments [16].

2.11. Statistical Analysis

All experimental trials were carried out using completely randomized factorial design with three variables of caffeine concentration, licorice powder level and storage time to assess their individual and combined effects on the physicochemical, microbiological and sensory characteris-tics of the probiotic drink. Data analysis was carried out using SPSS software v.22 and significant variations within the treatments were identified using Duncan’s multiple range test at a 95% confidence threshold (p < 0.05).

1. Results and Discussion

3.1. The pH and Probiotic Viability Assessment

Figure (1a, 1b) shows the pH levels and L. plantarum viability in probiotic beverages formulated with various quantities of licorice extract powder and caffeine through storage. The pH values of all samples ranged 4.13-4.32, which set within the acceptable range for apple juice (3.2–4.2) according to the Iranian National Standardization Organization (INSO) (ISIRI 14345, 2011), though the use of whey increased the highest pH. The ANOVA revealed significant major effects of licorice (p < 0.001) and caffeine (p < 0.001) on pH, as well as a significant interaction between the licorice and caffeine (p < 0.001). Similarly, L. plantarum viability was significantly affected by licorice (p < 0.001) and caffeine (p < 0.001), with a strong interaction effect (p < 0.001). The two parameters decreased significantly over the storage time (p < 0.05).

These results verified that licorice and caffeine acted synergistically to enhance probiotic viability while decreasing pH. Additionally, caffeine and associated phenolic compounds affected L. plantarum metabolic pathways through enzymatic activity, potentially increasing bacterial viability.

The observed decrease in pH with higher licorice and caffeine concentrations reflected intensified fermentation by L. plantarum, resulting in greater production of organic acids such as lactic acid that further promoted probiotic function. Caffeine positive effect might be linked to its potential degradation by these bacteria into methylxanthines, which could play an important role in cultivating beneficial gut microbiota such as Lactobacillus species. Methylxanthines might stimulate lactic acid bacteria (LAB) through nutritional enhancement, serving as possible carbon sources; thereby, creating a further favorable environment for bacterial proliferation. This combination of biochemical interactions suggested the potential beneficial effects of licorice extract and caffeine on probiotic viability and activity [14,17].Over the storage time, a significant decrease in L. plantarum viability and pH was typical for probiotic beverages. As nutrients decreased and metabolic waste accumulated, probiotic cell numbers decreased. Acid production might decrease or stabilize over time, affecting pH measurements. The initial pH decrease was linked to active bacterial metabolism during early storage, but extended time led to lower bacterial activity and viability; thereby, decreasing the two parameters significantly (p < 0.05). This decrease highlighted the importance of optimizing storage conditions and duration for maintaining probiotic efficacy in commercial products [18]. Similarly, a study on Iranian ultrafiltered white probiotic cheese containing caffeine detected that the addition of caffeine improved the survival of L. fermentum [19]. Tsirulnichenko and Kretova [14] reported that licorice root was a valuable source of probiotic-supporting substances, particularly fructans. They identified that a 1% concentration of licorice root extract represented the minimum effective dose needed to significantly stimulate the growth of probiotic microorganisms. In addition, Aleman et al. [20] studied the effects of functional ingredients on yogurt and detected that adding licorice root improved the viability of L. bulgaricus during storage.

3.2. Total soluble solids and Viscosity Assessment

Figure (2a, 2b) illustrate the TSS and viscosity of probiotic beverages formulated with varying concentrations of licorice extract powder and caffeine during storage. The TSS values of the samples ranged 10.3-11.85 °Brix, while viscosity values varied 141.21-171.43 mPa.s^-1. The ANOVA results showed that caffeine concentration included no significant effect on TSS (p = 0.20) or viscosity (p = 0.11). Licorice extract concentration included a non-significant effect on TSS (p = 0.42), but significantly increased viscosity (p=0.004). Storage time significantly affected TSS (p = 0.002) and viscosity (p = 0.001). Importantly, the interaction between licorice and caffeine was not significant for TSS (p = 0.292) but was significant for viscosity (p = 0.018), indicating that higher licorice levels amplified the thickening effect of caffeine. These findings suggested that while variations in caffeine concentration might affect TSS and viscosity, the absence of statistical significance was likely attributed to the relatively low caffeine levels used in the formulation. The statistically significant increase (p < 0.05) in the viscosity of the probiotic beverage with increasing licorice extract concentration could be linked to several factors inherent to the licorice composition. Licorice extract is rich in soluble solids and crude fibers, which act as natural thickeners, enhancing the beverage viscosity. Additionally, these components improve the water-holding capacity, contributing to a thicker further stable texture [20]. The significant increase in TSS and viscosity of samples during storage could be attributed to several factors. Over time, the evaporation of water leads to a concentration effect, increasing TSS, which can contribute to higher viscosity. Additionally, licorice extract contains carbohydrates that may interact with other components in the beverage matrix, enhancing thickening and water retention during storage.

The presence of caffeine and bioactive compounds from licorice affects microbial activity and biochemical reactions during fermentation, potentially increasing the production of organic acids and other metabolites that modify the beverage rheological characteristics [17]. De Carvalho et al. [7], in their study on developing a mixed beverage made from coconut water and cashew apple juice with caffeine, detected that the TSS content of their samples ranged 10.6-11.6, closely similar to the current findings. Similarly, Mohamed et al. [22] reported that formulations containing licorice extract showed significantly greater viscosity than the control samples, suggesting that licorice extract enhanced the thickness and texture of health drinks. Additionally, Khoshdouni et al. [23] reported a slight increase in TSS values over time, corresponding to the results achieved in the current study.

3.3. Total Phenolic Content and Antioxidant Activity (IC₅₀)

Fig. (3a, 3b) show the effect of licorice extract powder and caffeine on TPC (mg GAE·g⁻¹) and antioxidant activity (IC₅₀, µg·mL⁻¹) of probiotic beverages during storage. The TPC values ranged 116.12-171.53 mg GAE·g⁻¹, while IC₅₀ values varied 108.25-156.74 µg·ml⁻¹ within the treatments and storage times. Moreover, ANOVA results indicated that the licorice extract concentration significantly increased TPC (p < 0.001) and antioxidant activity (p< 0.001). Caffeine concentration included a non-significant effect on TPC (p = 0.27), but significantly improved antioxidant activity as assessed by IC₅₀ (p < 0.001). Importantly, the interaction between licorice and caffeine was significant for TPC (p= 0.031) and IC₅₀ (p = 0.024), indicating that higher licorice levels amplified the antioxidant-enhancing effect of caffeine. The TPC and antioxidant activity decreased significantly over the storage time (TPC, p < 0.001; IC₅₀, p < 0.001), reflecting the gradual degradation of phenolic compounds and decreased bioactivity during prolonged storage.

Licorice extract contains abundant phenolic compounds and flavonoids, which confer strong antioxidant potential. From its significant constituents are phenylflavonoids such as dehydroglyasperin C (DGC), recognized for their effectiveness in neutralizing free radicals and safeguarding cells against oxidative injuries. The antioxidant activity of licorice is primarily attributed to its ability to quench reactive oxygen species, limit lipid peroxidation and enhance the function of antioxidant enzymes. These mechanisms highlight licorice reported health-promoting characteristics, including anti-inflammatory and anti-aging effects and highlight its value as a natural ingredient for improving the antioxidant profile of foods and beverages. Collectively, the diverse bioactive molecules in licorice provide substantial protection against oxidative stress and may help decrease the risk of associated disorders [24]. Incorporation of caffeine into probiotic beverage formulations has been shown to progressively increase antioxidant activity. Caffeine shows antioxidant capacity through radical scavenging and moderation of oxidative stress. Within a probiotic system, caffeine can act synergistically with other bioactive compounds; thereby, enhancing the overall ability of the beverage to counteract reactive oxygen species. This gradual improvement is likely associated with interactions between caffeine, probiotic microorganisms and phenolic constituents, ultimately strengthening the functional attributes of the product over time [25]. A significant decrease in phenolic content and antioxidant activity during storage (p < 0.05) is commonly observed in probiotic beverages. Throughout storage, factors such as exposure to oxygen and light, fluctuations in temperature and changes in pH contribute to the breakdown of phenolic compounds. Additionally, biochemical reactions and interactions between phenolics and other components within the beverage can further destabilize these antioxidants, resulting in decreased antioxidant effectiveness over time [26]. Quintana et al. [27] reported that the TPC in licorice root extract ranged 48.47-180.06 mg GAE.g^-1 while assessing its antioxidant activity. Similarly, Vieira et al. [25] demonstrated that caffeine effectively scavenged hydroxyl radicals and investigated the mechanisms behind its antioxidant potential. Fatima et al. [28] observed that antioxidant activity and free radical scavenging capacity gradually decreased during the storage of functional fruit and vegetable-based drinks containing herbal and spice extracts. Although licorice and caffeine significantly enhanced the phenolic profile and antioxidant activity of probiotic beverages, limitations should be addressed. The storage time was relatively short (30 d), which restricted conclusions on long-term stability; future studies should extend to 60 d to provide further comprehensive shelf-life data. Quantitatively, the TPC values observed in this study (116–171 mg GAE·g⁻¹) were within or slightly higher than the values reported by Quintana et al. (48–180 mg GAE·g⁻¹), highlighting the strong contribution of licorice extract. Importantly, published studies on probiotic beverages containing caffeine were limited, underscoring the novelty of this study and the need of further investigations in this area.

3.4. Color Assessment of the Probiotic Beverage

Figure (4a, 4b, 4c) shows the changes in the L*, b* and a* color indices of the probiotic beverage containing licorice extract and caffeine over the storage time. The L* values ranged 44.61-56.45, the a* values ranged 14.42-25.76 and the b* values ranged 17.23-26.4 within the treatments and storage times. The ANOVA results indicated that increasing licorice extract and caffeine concentrations significantly decreased L* (p = 0.012) and significantly increased b* (p = 0.008). The two factors increased a* values, but this increase was significant only for licorice extract (p = 0.021). Storage time significantly affected all three indices (p < 0.001). Importantly, the interaction between licorice and caffeine was significant for L (p = 0.033,) and b* (p = 0.027), but not for a* (p = 0.29). These findings verified that higher licorice levels amplified the color‑modifying effect of caffeine, leading to darker and further saturated.hues during the storage.

This decrease in lightness could be linked to the inherent color compounds present in licorice extract and caffeine such as polyphenols and alkaloids, which created a deeper color. Additionally, interactions between these compounds and the beverage matrix during storage might enhance pigment formation or aggregation, further decreasing brightness. The significant change at p < 0.05 verified that this effect was statistically robust, suggesting a clear effect of these ingredients on the visual appearance and potentially consumer perception of the product [29-30]. Increasing the quantities of licorice extract and caffeine causing significant increase in the b* (yellow-blue) value could be explained by the yellowish pigments naturally present in licorice, primarily flavonoids such as liquiritin and glabridin, which contributed to its characteristic yellow color. Studies indicate that licorice flavonoid content is responsible for this yellow hue, which can intensify as concentration increases, thus increasing the b* value towards yellow tones [31]. The significant increase in a* (red-green) values for licorice extract might be attributed to the presence of reddish-brown compounds such as melanoidins, which were formed during processing or storage through Maillard reactions. These polymeric brown pigments accumulated with heat treatment, contributing to darkening and shift towards red color. Additionally, heat treatment could increase the TPC and antioxidant activity of licorice extracts, affecting their color characteristics [32-33]. Kang et al. [31] demonstrated that during storage, licorice extract created the formation of brown melanoidin pigments, resulting from Maillard reaction between sugars and amino acids. Similarly, Ardalanian and Fadaei [33] observed that adding ginseng extract to doogh samples caused increases in a* and b* color values. In an associated study, Jooyandeh and Alizadeh Behbahani [32] reported that the inclusion of pepper extract in functional yogurt led to a decrease in L* (lightness) with increases in a* (red-green) and b* (yellow-blue) values.

3.5. Sensory Assessment of Probiotic Beverage

Figure (5) presents the sensory evaluation outcomes of probiotic beverage samples assessed by the panel on Day 30 of storage, considering attributes such as taste, aroma, color, texture and overall acceptability. Statistical analysis indicated that variations in licorice extract and caffeine levels significantly affected taste scores (p < 0.05). The combination of 0.3% licorice extract with 80 mg of caffeine improved taste perception, whereas higher concentrations resulted in decreased ratings. Caffeine content showed no significant effect on aroma or color, while 0.3% licorice extract enhanced color scores (p < 0.05). Texture was favorably affected by 0.6% licorice extract, with no significant effect from caffeine. The sample containing 0.3% licorice extract and 80 mg of caffeine achieved the highest overall acceptability.

Similarly, Azami et al. [34] developed a dairy beverage with licorice extract and cocoa powder and detected that moderate licorice levels improved taste and aroma, but excessive quantities led to decreased acceptability due to bitterness and off-flavors. Mahmoudi et al. [35] produced a functional non-dairy probiotic beverage containing fermented jujube extract, reporting that addition of extract enhanced aroma and taste, contributing to higher acceptability scores. Similarly, Ghosi Hoojaghan et al. [36] studied the effect of fennel extract in doogh, finding that low concentrations included insignificant effects on flavor, but higher concentrations decreased flavor scores.

1. Conclusion

This research formulated a functional probiotic drink enriched with licorice extract and caffeine, which showed improvements in probiotic survival, antioxidant potential and sensory attributes. Licorice extract significantly enhanced phenolic content, viscosity and antioxidant capacity, whereas caffeine contributed to antioxidant activity and probiotic viability. Appropriate concentrations of these components optimized flavor, appearance, texture and overall consumer acceptance. Overall, the results indicate that licorice extract and caffeine act synergistically to strengthen the health-promoting characteristics and sensory quality of probiotic beverages, highlighting their promise as valuable functional ingredients.

1. Declaration

5.1. Acknowledgements

Contribution of Islamic Azad University, Iran, Tehran, is deeply appreciated.

5.2. Declaration of competing interest

The authors declare no conflict of interest or finance.

5.3. Authors’ Contributions

Aydin Ali Bigelow: Writing original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Funding acquisition. Marjaneh Sedaghati: Writing–review and editing, Supervision, Methodology, Resources, Projected administration, Conceptualization. Mozhgan Emtyazjoo: Supervision, Resources, Projected administration, Conceptualization.

5.4. Using Artificial Intelligent Chatbots

AI chatbot was used only for grammar correction.

5.5. Ethical Consideration

No ethical approval was required for the conduct of this study.

Background and Objective: Kombucha, a beverage made from sugared black tea fermented by symbiotic culture of bacteria and yeasts (SCOBY), has been attended  as a health-promoting drink and its important antimicrobial characteristics, as it has demonstrated antimicrobial activity against spoilage-causing microorganisms, including those linked to the dairy industry.

Material and Methods: Four yeast strains were isolated from Kombucha and tested against Clostridium sporogenes ATCC 19404, a gas-forming, proteolytic bacterium implicated in spoilage of dairy products. Two strains with the highest antimicrobial activity were selected and identified via 18S rRNA sequencing. Antimicrobial activity of Kombucha against C. sporogenes was assessed. Kombucha was used as a coagulant with citric acid and coagulant salts, using acid-heat coagulation method in production of Iranian acid-heat coagulated cheese. With these three coagulants, six cheese groups were produced, including three were contaminated with C. sporogenes spores, while the other three were designated as control groups. All samples were stored for 40 d and spoilage control was assessed on Days 0, 10, 20, 30 and 40 under refrigerated storage at 4 ℃. All data were statistically analyzed using ANOVA and Duncan’s multiple range test was used to assess significant differences when P < 0.05.

Results and Conclusion: Results showed that cheese samples made with Kombucha showed greater inhibitory effects, compared to samples coagulated with citric acid and coagulant salts. Furthermore, Kombucha-based cheeses showed significantly lower pH (P < 0.05) and higher acidity over time, indicating greater inhibition of C. sporogenes. These samples preserved firmer texture and better sensory acceptability, compared to cheeses made with citric acid or coagulant salts.

Keywords: Acid-heat coagulated cheese, Anti-microbial characteristics, Clostridium sporogenes, Coagulants, Fermentation, Food Spoilage, Kombucha, Yeast

. Introduction

Iranian acid-heat coagulated cheese is a type of fresh cheese that differs significantly from other commercially fresh cheese made with starter cultures or rennet and it is produced by coagulation through combination of heat and acidification of milk with citric, acetic, lactic acid or coagulant salts at 94 ℃. The near-neutral pH of Iranian acid-heat coagulated cheese makes it susceptible to microbial spoilage, leading to economic loss for producers. Microbiological spoilage is challenging to control, due to the involvement of undesirable microorganisms such as coliforms, yeasts, heterofermentative lactic acid bacteria (LAB) and spore-forming bacteria [1]. Gas-forming spoilage in cheeses can be made by several anaerobes, most particularly spore-forming Clostridium strains, which produce CO₂/H₂ and off-flavors during growth. In ripened hard and semi-hard cheeses, late blowing defect (LBD) during warm ripening has primarily been linked to C. tryobutyricum and its linked strains. However, LBD develops at temperatures greater than 10 ℃ during ripening and not expected under refrigerated storage [2-4].

Several methods have been investigated in controlling spoilage in cheese, for example physical methods such as bactofugation or microfiltration of milk and using nitrite or lysozyme have been recommended [5]. However, each method includes drawbacks. Bactofugation decreases the number of Clostridium spores but not enough to prevent spoilage in cheese. Microfiltration removes 99% of spores, but is only usable for skimmed milk, as milk fat globules are too large to pass through the filter [6]. The use of nitrite and nitrate as food additives to prevent spoilage includes limitations due to European Food Safety Authority (EFSA), emphasizing on the production of nitrosamine, which is a carcinogenic compound [7]. Lysozyme, derived from egg white, is an effective additive but is an allergen and can be harmful for people with egg allergy [8]. Recently biopreservatives such as bacteriocins have been investigated for controlling cheese spoilage, [9-13] or using Kombucha inoculum as coagulant agent to control cheese spoilage [14].

Kombucha contains a variety of acetic acid bacteria (AAB), LAB and yeasts (e.g., Pichia, Candida, Saccharomyces, Brettanomyces and Zygosaccharomyces) [15]. The activity of these microorganisms and their metabolites provides Kombucha potentials, including antimicrobial, antioxidant [16-18] and therapeutic activities [15, 19-22]. The aim of this study was to investigate the antimicrobial characteristics of yeasts isolated from Kombucha and Kombucha alone as a non-conventional biopreservative against C. sporogenes, using Kombucha inoculum as a coagulant to produce Iranian acid-heat coagulated cheese and through manually contaminating the cheese with spores of C. sporogenes and monitoring the product to investigate its microbiological profile and antimicrobial activity in growth inhibition of C. sporogenes during 40 d of refrigerated storage.

1. Materials and Methods

2.1. Kombucha and Fresh Cheese production

Kombucha was produced in Laboratory condition using back slopping method and black tea extract (Camellia sinensis, 0.25% w/v) containing 7.5% w/v sucrose, cooled to room temperature (RT) and inoculated with 10% Kombucha from a previous fermentation and 3% w/v of SCOBY. Fermentation was carried out at RT for 2 m to reach pH 2.5 [23]. Fresh cheese was manufactured in Iran Dairy Industries (Pegah), Gorgan, Iran, from pasteurized milk according to the Institute of Standards and Industrial Researches of Iran (ISIRI) guideline no. 13863, Lactic Cheese - Specifications and Test Methods. To produce fresh cheese according to the guideline, 5% w/v of Kombucha were added to pasteurized milk at 94 °C. The coagulum was cut into pieces, cooled to RT, pack sealed and stored at 4 °C. Two other groups of cheese were produced using citric acid and coagulating salts for comparison [14]

2.2. Isolation of yeasts from Kombucha

Briefly, 100 μl of Kombucha were cultured on yeast extract glucose chloramphenicol (YGC) agar using surface plate method in triplicate and plates were incubated at 28 °C for 72 h. Colonies with various appearances were selected for microscopic assay and isolated using streak plate method [24].

2.3. Preparation of clostridial spore suspension

Clostridium sporogenes ATCC 19404 spore suspension was prepared using the method described by Gamal-eldin et al. (2017). The C. sporogenes was incubated under anaerobic conditions at 37 °C for 24–48 h. The culture containing C. sporogenes was then heated at 63 °C for 40 min to eliminate vegetative cells and quickly stored at 4 °C to initiate thermal shock. Spore concentrations were assessed by plating stock suspensions on RCM agar (reinforced clostridial media). Plates were incubated anaerobically at 37 °C for 48 h. Duplicate plates were used for all dilutions [25].

2.4. Antimicrobial activity of yeasts

The isolated yeast strains from Kombucha were cultivated on sweetened tea (7.5% sucrose) and incubated at 28 °C for 5 d. Cultures were then centrifuged (Centurion, UK) at 5000 rpm for 10 min and cell-free supernatant (CFS) was sterilized in three various ways, using 0.45-μm MCE syringe filters (Biofil, Germany), boiling at 94 ℃ for 2 min and autoclave (121 °C, 15 min). The purpose of boiling was to simulate an environment similar for the creation of acid-heat coagulated cheese and using autoclave to sterilize the yeast strains as well. A diluted bacterial suspension of C. sporogenes ATCC 19404 (10⁴ spores.ml^-1) was prepared as described in Section 2.3. Using 96-well microplate and Mueller-Hinton (MH) broth, the antimicrobial characteristics of the sterilized supernatant of each yeast isolate were investigated through incubation of the microplates at 37 °C under anaerobic conditions using anaerobic gas packs (Anaerocult A, Merck, Germany). The antimicrobial activity was assessed using microplate spectrophotometer and Eq. 1:

                                                                                                                                           Eq. 1

Where,  was an average of the replicates of light absorption values at 600 nm of the positive control and  was an average of the replicates of light absorption values at 600 nm of the negative control. Each test was carried out in triplicate [26].

2.5. DNA extraction and molecular identification of the yeast isolates

Two yeast isolates with the highest antimicrobial activity were selected and their DNA were extracted using phenol-chloroform method. The PCR reaction was carried out by amplifying 18S rRNA gene using ITS-1 (5’-TCCGTAGGTGAACCTGCGG-3’) and NL-4 (5’-GGTCCGTGTTTCAAGACGG-3’) primers. The thermal protocol included an initial denaturation of 94 °C for 5 min, followed by 35 cycles including denaturation at 94 °C for 30 s, primer annealing at 55 °C for 40 s and extension at 72 °C for 60 s. The process was terminated at a final extension of 72 °C for 7 min [27].

2.6. Antimicrobial activity of Kombucha

Antimicrobial activity of Kombucha against C. sporogenes was investigated using a similar protocol in Section 2.3. Kombucha first was neutralized using filter sterilized 1 N NaOH and antimicrobial activity test was carried out twice [24].

2.7. Microbiological analysis of fresh cheese made with Kombucha

Fresh cheese was made using Kombucha as coagulant agent and for comparison, fresh cheeses were produced using coagulant salts and citric acid. Each sample was contaminated with 10³ spores.ml^-1 of C. sporogenes, control samples without clostridial spores were made. Samples were pack sealed in cups and stored at 4 °C for further analysis during 40 d of storage with 10-d intervals. At each sampling time, 10 g of cheese were mixed with sterile saline (0.85% NaCl) and homogenized for 5 min using stomacher. Dilutions were prepared and each was plated on RCM agar, containing 0.7% l-cysteine hydrochloride as reducing agent to help anaerobic growth. Triplicate plates were incubated at 37 °C for 72 h in anaerobic jar (Merck, Germany) [19,20,28-47].

2.8. Sensory analysis

Sensory analysis was carried out with 15 trained panelists selected from university staff and students. Moreover, 15 trained panelists were asked to rate appearance, color, flavor and consistency, based on 5-point hedonic scale (5- really good, 4- good, 3- normal, 2- bad, 1- really bad) [14].

2.9. Texture profile analysis (TPA)

Using the procedure described by MacFie (1990), a texture profile analysis (TPA) test using TA.XT Plus (Stable Micro System, UK) was assessed through two-cycle compression testing, where each cheese sample was cut into 2 × 2 cm² cubes and the texture profiler was calibrated using strain rate of 2 mm.s^-1 for velocity, 5-s compression delay time and 0.2 N contact force [29].

2.10. Statistical analysis

All analyses and assays were carried out in triplicate for all produced samples and values were expressed as average ±SD (standard deviation). Microbiological data, pH, acidity and texture analysis were carried out using one-way analysis of variance (ANOVA) with SPSS v.26 (IBM, USA). Duncan’s multiple range test was used to assess significant differences within the studied parameters. Differences were considered statistically significant when P < 0.05.

1. Results and Discussion

3.1. Antimicrobial activity of the isolated Yeasts against Clostridium sporogenes

Using centrifuge to collect CFS of the yeast isolates in addition to use boiling and syringe filters, the pH value of CFS of each isolate were acidic. The KSY2 and KSY4 isolates included the lowest pH value and overall, showed the highest antimicrobial activity against C. sporogenes. As shown in Table 1, the inhibition percentage of the selected yeast isolated were 99 and 88.5 when using syringe filters and the inhibition percentages were 75.17 and 67.7 when using heat to CFS of the isolates. The KSY2 and KSY4 samples in the two treatments included significant differences (P < 0.05). The CFS of yeast strains were sterilized using autoclave, showing no antimicrobial activity.

The ability of yeasts to ferment sugar might explain the decrease in pH value due to decomposition of sucrose into glucose and fructose under anaerobic conditions, resulting in the production of CO₂ or alcohol in the fermentation [30]. It was reported that non-saccharomyces yeasts might play a significant role at the beginning of anaerobic fermentation, increasing ethanol content or producing various metabolites such as aromatic esters, organic acids, fatty acids or higher alcohols. However, it appears that the antimicrobial effects were due to the possible heat-stable metabolites produced by the yeasts such as organic acids produced through sucrose decomposition, leading to decrease in pH and subsequently showing antimicrobial activity against C. sporogenes [31].

3.2. Molecular identification of yeast isolates

Yeast strains of KSY2 and KSY4, which showed the highest antimicrobial activity against C. sporogenes, were identified as Candida parapsilosis (Table 2). It was reported that C. parapsilosis might naturally be present in the human intestinal microbiota at 10² to 10⁴ CFU.g^-1. Candida strains, including C. parapsilosis from feta cheese and feces of healthy babies, were isolated and their probiotic characteristics were assessed, as C. parapsilosis strains had the highest adhesion to intestinal cells and the highest cholesterol decrease in other Candida strains. This yeast has been reported in foods such as fermented olives, cassava and fermented dairy products [32-33]. The C. parapsilosis is known for its significant ability and capacity to ferment carbon sources. It was reported that C. parapsilosis could produce mannitol from glucose during the fermentation, with 1.97 g.l^-1 of mannitol produced after 120 h of fermentation [34]. It is possible that C. parapsilosis in Kombucha uses glucose and fructose as carbon sources as similar results reported by Sievers et al. (1995), suggesting this strain was involved in the production of mannitol in Kombucha [35]. Mannitol is a substance produced by microorganisms in Kombucha and can be used as a carbon source by AAB [36-37].

3.3. Antimicrobial characteristics of Kombucha against Clostridium sporogenes

Heat-treated Kombucha showed a 97% inhibition of C. sporogenes, slightly higher than that Kombucha without heat-treatment did (94%). Neutralized Kombucha with NaOH (1 N) showed no antimicrobial activity against C. sporogenes. Velicanski et al. (2014) reported that acetic acid produced during fermentation by AAB and yeasts decreased the pH, resulting in antimicrobial activity [38]. Kaewkod et al. (2019) highlighted the importance of organic acids in Kombucha antimicrobial activity against pathogenic bacteria such as Escherichia coli, Salmonella Typhi and Shigella dysenteriae, while neutralized Kombucha did not show any antimicrobial activity against these microorganisms. Kombucha contains heat-resistant antimicrobial agents, which can be used in various food matrices to control thermophilic spore-forming bacteria [39].

Kombucha antimicrobial characteristics are attributed to the activity and production of metabolites by yeasts and AAB during fermentation. According to Al-Mohammadi et al., 2021, Kombucha contains nine groups of chemical components, including alcohols, acids, lactones, condensed heterocyclic compounds, antibiotics, esters, aldehydes, fatty acids and alkaloids. These metabolites act synergistically, contributing to Kombucha antimicrobial characteristics. While the low pH contributes to the antimicrobial effects, other heat-stable and pH-dependent compounds play roles in Kombucha as an antimicrobial agent [40-41].

3.4. Analysis of Kombucha Fresh Cheese

3.4.1. pH and Acidity

Although, fresh cheese made with Kombucha generally included low pH levels (Table 3) [37-42; 43]. By comparing the 10-d interval of each sample, pH values decreased over time with the highest decrease in Kombucha cheese contaminated with spores (KB). The acidity, as shown in Table 4 and through comparing each sample the 10-d intervals, increased significantly in the same cheese group. The C. sporogenes is proteolytic and gas-forming and its metabolism can increase pH and drive spoilage in dairy matrices. In this study, Kombucha-derived treatments

limited these changes, consistent with decreased growth and activity of C. sporogenes [44]. Therefore, it could be concluded that the use of Kombucha might inhibit the growth of C. sporogenes inoculated into cheeses. Changes in the acidity of cheeses produced with coagulant salt and citric acid and contaminated with C. sporogenes and their respectively control groups were not significant (P > 0.05). However, the average acidity of Kombucha cheeses showed a significant difference (P < 0.05), compared to the other cheese groups. Due to the possible growth of C. sporogenes in cheeses produced with coagulant salt and citric acid and since their respective changes were not significant, it might indicate the growth of C. sporogenes and the neutralization of the acids during storage.

3.4.2. Antimicrobial effect of Kombucha on fresh cheese against Clostridium sporogenes

Microbial count on artificially contaminated cheese groups of Kombucha, coagulating salts and citric acid was carried out during 40 d of storage. The results (Fig. 1) indicated an increase in microbial count of cheese made with coagulant salt from 4.15 to 5.82 CFU.g^-1 and citric acid from 3.95 to 5.81 CFU.g^-1. There were no significant changes (P > 0.05) in these groups, indicating that neither coagulant salts nor citric acid showed antimicrobial effect against C. sporogenes in fresh cheese over 40 d of storage. Cheese samples made with Kombucha showed an increasing trend from 4.19 to 5.27 CFU.g^-1; however, the increase in microbial count was less than that in other groups. Comparison of average microbial count in Kombucha cheese on Days 30 and 40 of storage showed a significant change (P < 0.05), compared to the other two groups of cheese.

It was assessed that Kombucha cheese showed 87.57% higher inhibition effect, compared to citric acid cheese, and 62.34% higher inhibition effect, compared to coagulant salts cheese. Vukic et al. (2021) detected that use of Kombucha in producing fresh cheese included inhibitory effects on E. coli and Listeria monocytogenes during 30 d of storage, decreasing microbial count by 98.35 and 98.98%, compared to the control group, respectively. This effect was likely due to the antimicrobial characteristics of Kombucha, which were attributed to its pH and bioactive compounds, including phenolic compounds. These compounds included bacteriostatic and anti-proliferative effects against bacterial growth [38-45].

3.4.3. Texture Analysis

During 40 d of storage, there were no changes in the texture of the cheeses in the control group. According to Fig. 2 and by comparing the average of each contaminated cheese group in each interval, Kombucha cheese contaminated with C. sporogenes demonstrated a further stable texture, compared to cheeses made with coagulant salt and citric acid with significant changes (P < 0.05), suggesting possible antimicrobial effects of Kombucha against C. sporogenes. Moreover, the changes in cheese made with citric acid included a harder texture, which could be attributed to the purity of the citric acid in the formulation. Pure citric acid promotes further aggregation of milk proteins in cheese [46]. According to Gomez-Torres et al. (2015), the softening of cheese texture is due to metabolic activity of Clostridium genus, leading to the hydrolysis of milk proteins and further disruption of the cheese matrix [47]. Therefore, the softening of cheeses made with coagulant salt and citric acid was more significant than that in Kombucha cheese. The inhibitory effect of Kombucha on the growth of C. sporogenes, contributed to the firmer further consistent texture of the Kombucha cheese.

3.4.4. Sensory analysis

After cheese production, sensory analysis was carried out, with the sensorial attributes are present in Fig. 3 (color, aroma, cuttability, taste, chewiness and mouthfeel). Iranian acid-heat coagulated cheese produced with Kombucha included a mild sour taste with a soft spreadable texture, distinguishing it from commercially available acid-heat coagulated cheese, which were made with coagulant salt. It seems that except the overall result, the sensorial attributes between cheese made with Kombucha and cheese made with salt included no significant differences (P > 0.05) only in the color characteristic. Although the Kombucha cheese received an overall acceptable score (4.14/5), based on information in Fig. 3, its lowest score was in the color characteristic, attributed to the slight browning from the use of black tea in Kombucha production. Similarly, in the sensory evaluation by Vukic et al. (2021), freshly produced Kombucha cheese scored the lowest for aroma and color. Cheese produced with citric acid included the lowest overall score with an average of 3.38 and its sensorial attributes and overall results showed significant differences (P < 0.05), comparing to cheeses made with Kombucha and coagulant salt [14,16-22,28-48].

1. Conclusion

The inherent nature of dairy products and the processing of Iranian acid-heat coagulated cheese has made this product susceptible to various types of microbial spoilage, including gas-forming spoilage by Clostridium spp. in dairy systems. Under refrigerated storage (4 °C), Kombucha and cell-free yeast metabolites moderated spoilage indicators and inhibited C. sporogenes, compared to acid or salt coagulants. Several methods have been suggested to prevent microbial spoilage such as using chemical and physical methods. Considering the preferences of the society and the popularity of using natural preservatives in the field of food industry and the preference of consumers to use foods produced with natural preservatives with beneficial health characteristics; Kombucha, as a novel product with unique compounds resulted from the activity of bacteria and yeasts, were investigated as a biological additive. For Kombucha and the metabolites from identified Kombucha yeasts, C. parapsilosis had high antimicrobial characteristics against C. sporogenes. It was  investigated that the inhibitory effect of Kombucha, as a coagulant used in cheese, was 87.57% higher, compared to cheese coagulated with citric acid. In comparison to cheese made with coagulant salts, Kombucha showed an inhibitory effect of 62.34%. Therefore, Kombucha can be used as a potential biopreservative. Further research should be carried out to understand precise capabilities of various types of Kombucha due to the presence of their specific compounds and their potential use in dairy industry products, especially Iranian acid-heat coagulated cheese.

Background and Objective: The type of substrate, fermentation conditions and microbial flora can affect the functional properties of fermented beverages. Therefore, this study aimed to compare the microbiological, chemical, antioxidant and antimicrobial properties of kombucha beverages produced from four herbal infusions (Green tea, Black tea, Lemon verbena and Shirazi thyme) and investigate the effects of fermentation on altering herbal infusions characteristics.

Material and Methods: Various herbal teas were fermented for 12 days after inoculation with the kombucha starter culture at 28°C. Total phenolic compounds and antioxidant activity were measured by Folin–Ciocalteu assay and DPPH free radical scavenging activity before and after fermentation of herbal infusions, respectively. Also, antibacterial activity against pathogenic gram-positive and gram-negative bacteria was evaluated by the well-diffusion method. Antifungal activity of various herbal infusions and Kombucha beverages against Aspergillus niger and Aspergillus flavus was analyzed by agar dilution and well-diffusion methods.

Results and Conclusion: The pH of different herbal infusions decreased significantly after fermentation. Total phenolic compounds increased markedly during fermentation, reaching 65.82–153.38 mg GAE.L^-1 in the Kombucha samples, and green tea Kombucha indicated the highest value. Microbial analysis revealed that the highest microbial population in all Kombucha samples belonged to acetic acid bacteria (10⁶–10⁷ CFU.mL^-1), followed by lactic acid bacteria (10⁴–10⁵ CFU.mL^-1), and yeasts (10⁵ CFU.mL^-1). Green tea infusion and its Kombucha beverage showed the strongest DPPH scavenging capacity. Unlike herbal infusions, all Kombucha beverages revealed antifungal and antibacterial activity. According to these findings, fermentation significantly enhanced the functional, antioxidant, and antimicrobial properties of herbal infusions and green tea Kombucha was introduced as a considerably promising functional food preservative.

Keywords: Herbal infusion, Functional beverage, Fermentation, Antioxidant, Antimicrobial activity

1. Introduction

Kombucha is a traditional fermented, acidic drink with a relatively sweet taste, which has been popular in countries such as China, Russia, Japan, and Germany in the past due to its health-promoting properties. Today, the desire to consume this functional drink has increased in many countries, due to its antioxidant, antimicrobial, anticancer, antidiabetic, liver detoxification, cholesterol and blood pressure-lowering properties, and in the treatment of many heart diseases and insufficiency. Kombucha is the result of the fermentation of sweetened herbal tea for 7-14 days at 28-30 °C by acetic acid bacteria (AAB) (Acetobacter, Gluconobacter), lactic acid bacteria (LAB) and yeasts (Candida, Pichia, Saccharomyces, Schizosaccharomyces, Zygosaccharomyces, Brettaniomyces, and Torula) that exist symbiotically in the SCOBY (symbiotic culture of bacteria and yeast). The type of tea, sweetener, microorganisms, and fermentation conditions can affect Kombucha drink characteristics, including its chemical composition, color, flavor, and microbial flora [1]. In a recent study, Li et al. comprehensively compared microbial community and quality of kombucha from four different regions of China. This study identified 197 indigenous yeast and bacterial strains and demonstrated that fermentation time strongly influenced color, phenolic composition, volatile profiles, and antioxidant activity [2].

Although black tea (Camellia sinensis) is the most common substrate, interest in using alternative herbal infusions is growing, driven by the potential to improve bioactivity and diversify sensory profiles [3]. In a study, the antibacterial and antioxidant activities, as well as the chemical composition of kombucha beverages prepared from linden (lime tree), chamomile, nettle, St. John’s wort, rockrose (Cistus spp.), and green tea were comparatively evaluated [4]. SCOBY mediated fermentation significantly alters the chemical composition of herbal Kombucha, leading to notable changes in sugars, organic acids, and an increase in essential minerals such as iron, magnesium, and calcium. However, the type of herbal or plant waste infusion, such as mint, nettle, blackcurrant leaf, and banana peel, was identified as a key factor influencing the antioxidant potential and sensory attributes of the resulting beverages [5, 6].

Plant infusions such as green tea (Camellia sinensis), lemon verbena (Aloysia citrodora), and thyme (Thymus vulgaris) contain high levels of polyphenolic compounds and essential oils with well-documented antibacterial, antifungal, and antioxidant effects [7, 8]. The integration of these herbal extracts into Kombucha fermentation may influence both the microbial dynamics of SCOBY and the biochemical composition of the final product [9]. Several studies have investigated the antimicrobial and antioxidant potential of Kombucha brewed from black and green tea [10, 11], and some have examined chemical and biological characteristics of Kombucha derived from herbal infusions such as peppermint, oregano, sage, winter savory, stinging nettle, elderberry, quince and yarrow [12-14]. All of these studies have emphasized the effect of substrate type on the formation of beneficial and health-promoting compounds during Kombucha fermentation. Also, various antimicrobial activities of Kombucha beverages were reported in some studies due to the presence of organic acids, phenolic compounds, microbial protein and enzymes [15, 16].

Kombucha fermentation is a combination of three processes: lactic, alcoholic, and acetic acid fermentation. During tea fermentation, yeasts in the medium first convert sucrose, which can be consumed as a sweetener, into glucose and fructose, and then, by consuming glucose, ethanol and carbon dioxide are produced in the drink. The concentration of ethanol produced is usually less than 1%, which is oxidized by acetobacteria and leads to the production of acetic acid and acetaldehyde. In addition, glucose is also converted to gluconic acid by acetobacteria. Gluconic acid has been of interest in various studies in recent years due to its detoxification properties. This acid can remove various types of toxic substances, such as pollutants, chemicals, excess steroid hormones, and bilirubin from the human body through the urinary system. Gluconic acid is considered a precursor to vitamin C. In addition, gluconic acid can be converted into glucosamine, which is a useful substance related to the synthesis of cartilage, collagen, and interarticular fluids that are effective in the treatment of osteoarthritis. Lactic acid is another metabolite produced by LAB during the fermentation of glucose and sucrose [17]. As mentioned before, Kombucha contains a variety of compounds, including organic acids, amino acids, simple sugars, caffeine, tannins, phenolic compounds, folic acid, minerals, and vitamins C and B. Therefore, despite the active and health-promoting compounds mentioned in Kombucha, the use of this fermented beverage and its starters in the formulation of various foods can have a positive effect on their nutritional, quality, and shelf life as a novel biopreservation [18-22].

To produce Kombucha traditionally, black or green tea waste is usually used. Recently, Kombucha drinks based on various medicinal plants such as Thyme, Peppermint, and Lemon verbena, etc., are also been produced and marketed in Iran. However, there is limited information about functional properties such as phenolic components, antioxidants, antibacterial, and antifungal activities of various types of Iranian Kombucha beverages. Shirazi thyme (Zataria multiflora Boiss.) is one of the best-known medicinal plants in Iran from the mint family; most of its therapeutic, antimicrobial, and antioxidant properties are attributed to the presence of phenolic compounds such as thymol and carvacrol in it [23]. Lemon verbena (Aloysia citrodora) is also an aromatic plant from the Verbenaceae family that can be cultivated in different regions of Iran. The leaves and flowering tops of this plant have medicinal properties, and its essential oils are rich in flavonoid compounds [24].

According to our current knowledge, there is limited information to compare the microbiology, antioxidant, and antimicrobial activity of herbal infusions before and after fermentation with Kombucha starters. Given the importance of the type of herbal tea on the properties of Kombucha, research in this area is necessary and continues. Therefore, this study aims to compare the microbiology, antioxidant, antibacterial, and antifungal properties of different types of Kombucha fermented beverage produced from four herbal infusions (green tea, black tea, lemon verbena, and Shirazi thyme). In addition, the total phenolic content, antioxidant, and antimicrobial activity of four unfermented herbal infusions are compared before the production of kombucha beverages.

1. Materials and Methods

2.1. Materials

Black Tea, green tea, lemon verbena, and thyme were purchased from Mashhad local market. Chemical material and microbial culture medium were obtained from Merck company.

2.2.Various herbal infusions and Kombucha production

In this study, four different types of Kombucha were prepared based on the formulation proposed by Cardoso et al., with slight modifications [25]. First, 80 g of sugar was dissolved in 1 L of water (95 °C) and 5 g of each dried herbal leaf, including black tea, green tea, Lemon verbena, or Shirazi thyme, was added, and the infusion was filtered through a strainer filter after 10 min. After cooling the infusion to 25 °C, the SCOBY Kombucha (4% w.v^-1) and 40 mL of previously produced Kombucha drink as a starter were inoculated and placed in a dark place at ambient temperature (28°C) for 12 days to ferment. Finally, a newly formed scoby was removed from the surface, and all Kombucha beverages were stored at 4 °C after passing through a strainer until the tests were performed. It should be noted that unfermented herbal infusions without sugar were used as controls in various tests in this study.

2.3. Microbiological characterization

To determine the microbial flora of the produced Kombucha samples, after preparing serial dilutions, enumeration of LAB, AAB, and yeasts were performed on MRS agar (Man Rogosa Sharpe, Liofilchem, Italy) (37°C for 48 h), glucose yeast extract agar (glucose 50 g.L^-1, yeast extract 10 g.L^-1, and agar 20 g.L^-1) (30°C for 48 h), and yeast glucose chloramphenicol agar (YGC agar)(Condalab, Spain) (25°C for 72 h), respectively. The population of each group of microorganisms in the Kombucha beverage was reported as Log CFU.mL^-1.

2.4. pH and total acidity

The total acidity of each Kombucha beverage was determined by titration with standardized 0.01N NaOH and phenolphthalein as an indicator. The titrant was added dropwise until the pH reached 8.0, which was selected as the endpoint. The total acidity was calculated and expressed as g of acid / per L of sample (g.L^-1). All titrations were performed at room temperature (~25 °C). The pH was determined by a previously calibrated pH meter and in triplicate to ensure reproducibility.

2.5. Total phenolic content

The Folin-Ciocalteu method was used to measure the total phenolic content of the produced Kombucha and herbal infusions samples. For this purpose, 0.5 mL of each sample was added to 2.5 mL of Folin-Ciocalteu reaction solution (0.2 N) in a test tube. After 5 min at room temperature, 2 mL of sodium carbonate (7.5% w.v^-1) was mixed with the previous solution. After incubation for 30 min and the color change of the solution, the absorbance was measured at 760 nm. Total phenolic content was determined in terms of gallic acid equivalents using a gallic acid standard curve. The results were expressed as mg of gallic acid equivalents per mL of each Kombucha beverage (mg GAE.mL^-1) [25].

2.6. Antioxidant activity

First, 0.2 ml of filtered Kombucha sample or each herbal infusion was added to 2.8 ml of 0.1 mM ethanolic DPPH solution and after mixing with a shaker at room temperature, it was kept in a dark place for 30 minutes. Distilled water was used as a control in this test. Then, the absorbance of each of the tested solutions was measured at 517 nm and the percentage of DPPH radical scavenging activity was calculated with Eq. 1 [26].

DPPH scavenging activity (%) = [A control- A sample/ A control] ×100                                                                  Eq. 1

2.7. Antimicrobial activity

2.7.1. Antifungal activity

The antifungal properties of the Kombucha samples and herbal infusions against two common spoilage molds, Aspergillus niger (PTCC 5010( and Aspergillus flavus (PTCC 5004(, were investigated by well diffusion and agar dilution methods. In the well diffusion method, after inoculating the spore suspension (100 μL) of each of the molds (10⁴ spores.mL^-1) on the surface of the YGC agar medium, wells were created on its surface with a sterile cork borer and 100 μL of each sample was inoculated into them. After the plates were incubated for 72 h at 25 °C, the diameter of the inhibition zone was measured with a ruler and reported in mm [27].

 Also, in the agar dilution method, various concentrations of Kombucha or herbal infusion samples (10 – 40 % (v.v^-1)) were added to molten YGC agar and poured in plates to solidify. Then, a blank disk was placed in the center of the plate and 10 μl of spore suspension of each mold (10⁴ spores.mL^-1) was inoculated into it. In this test, a YGC agar plate without Kombucha was used as a control. After 96 h of incubation at 25°C, the diameter of mold growth in the plates was measured and the percentage of growth inhibition in the presence of Kombucha or herbal infusion samples was calculated with Eq. 2 [28].

Inhibition of mold growth (%) = 1- (mold growth diameter in treatment plate /mold growth diameter in control plate) ×1                                                                                                                                                                                                          Eq. 2

2.7.2. Antibacterial activity

The antibacterial activity of various Kombucha and herbal infusion samples against gram-negative (Escherichia coli (PTCC 1399), Salmonella entritidis (PTCC 1709)) and gram-positive (Listeria monocytogenes (PTCC 1298), Staphylococcus aureus (PTCC 1431)) foodborne pathogenic bacteria was evaluated by the well diffusion method. First, each stock culture of the pathogenic indicator bacteria was activated in Muller-Hinton broth (MHB) medium for 24 h at 37 °C. Then, 100 μL of overnight bacterial culture at a concentration 1.5 × 10⁸ CFU.mL^-1 (0.5 McFarland) was inoculated onto the surface of Muller-Hinton agar (MHA) medium and spread well using a sterile swab. Then, wells with a diameter of 6 mm were created in the culture medium using a cork borer and 100 μL of filtered Kombucha was injected into them. All plates were transferred to an incubator at 37 °C. After 24 h, the inhibition zone around the wells was measured using a ruler and reported in mm [16].

2.8. Statistical analysis

In the present study, all tests were performed with three replications. Data analysis was performed using a completely randomized design using SPSS version 23. Analysis of variance (ANOVA) was used to determine significant differences at a 95% confidence level, and Duncan's test was used to compare means. Graphs and tables were drawn using Microsoft Excel version 2020. Moreover, Pearson’s correlation coefficients (r) were calculated to assess the relationships among chemical parameters, bioactive compounds, functional properties, and microbial populations in herbal infusions before fermentation and kombucha beverages after fermentation. Two-tailed tests were applied, and statistical significance was set at p < 0.05 and p < 0.01.

1. Results and Discussion

3.1. Microbiological characterization

The microbial evaluation of four types of Kombucha demonstrated that the population of LABs varied between approximately 10⁴-10⁵ CFU.mL^-1, and the highest number of LABs was observed in BTK (4.89 ± 0.24 Log CFU.mL^-1) and GTK (4.85 ± 0.31 Log CFU.mL^-1), with no significant difference (P > 0.05), respectively. Moreover, there was no significant difference among the yeast population (5.29 ± 0.36 – 5.43 ± 0.35 Log CFU.mL^-1) in the four types of Kombucha (P > 0.05). The highest frequency belonged to the AAB and varied between 6.59 ± 0.47-7.35 ± 0.35 Log CFU.mL^-1 in the different studied Kombucha samples (Fig. 1). In the study by Cardoso et al., the population of LAB, AAB, and yeasts in two types of Kombucha (black tea and green tea) after 10 days of fermentation at 25 °C was reported to be about 10⁵-10⁶ CFU.mL^-1 [25]. Also, Zhao et al. calculated the number of mesophilic bacteria as 10⁶ CFU.mL^-1 and yeasts as 10⁵ CFU.mL^-1 in Kombucha after 10 days of fermentation [29]. This is while higher populations of AAB and LABs, as well as yeasts (10⁷ CFU.mL^-1) were obtained in the study by Neffe-Skocińska et al. [30] under similar fermentation conditions. This difference observed in different studies is likely due to differences in the amount of sweetener, type of herbal infusion, initial inoculum (amount of SCOBY and previously produced Kombucha as starter), and fermentation conditions and time. The study of Jafari et al. confirmed the presence of 3 species of yeasts, 2 species of AABs and 3 species of LABs in black tea kombucha after 14 days of fermentation [31].

3.2. pH and acidity

According to Table 1, the herbal tea infusion samples exhibited pH values ranging from 6.54 to 7.60. As shown in Table 2, the pH of the Kombucha samples varied between 2.63-2.82, and there was no significant difference in this regard among the different treatments (P > 0.05). The lowest and highest pH belonged to the GTK and STK Kombucha samples, respectively. Also, the pH of all Kombucha samples was suitable for human consumption of fermented beverages (pH > 2.5). At pH values < 2.5, the health of consumers will be at risk due to the high concentration of acetic acid. Similarly, at pH > 4.4, the microbiological safety of the beverage will be at risk [32]. The highest and lowest acidity were also related to GTK and STK Kombucha, respectively. Overall, there was a significant difference between the acidity of GTK and other Kombucha samples (P < 0.05). One of the factors that determines the end of the fermentation time of a fermented Kombucha beverage is the titratable acid level reaching about 4 to 6 g.L^-1 [33]. During the preparation of Kombucha, sugars are consumed by microorganisms such as AAB and LAB as well as yeasts and are converted into organic acids such as acetic acid, lactic acid, gluconic acid, etc. Therefore, following the fermentation of sweet tea, a decrease in pH and an increase in acidity occur. Instead of the yeast population, the largest population of AAB and LABs was observed in GTK and BTK samples, which could be the reason for producing more acid. Similarly, in the study by Cardoso et al. [25], green tea Kombucha had a lower pH and higher acidity compared to black tea Kombucha. They attributed the difference in the acidity of the Kombuchas to the difference in the abundance of AAB and LAB populations in these two types of beverages and their ability to produce organic acids. According to their report, the populations of AAB, LAB, and yeasts in green tea Kombucha were higher than those of black tea, but no significant difference was observed between them. In fact, the microbial population of the Kombucha sample at the end of fermentation depends on the type and content of sugar, herbal tea, and the amount of inoculation of SCOBY and abundance of various strains of AAB, LAB, and yeasts in it, as well as the fermentation conditions (time and temperature).

3.3. Total phenol content

Chemical compounds consisting of at least one aromatic ring with at least one hydroxyl group constitute phenolic compounds. They belong to secondary metabolites synthesized in plants and can be divided into different subgroups, including simple phenols, phenolic acids, flavonoids, coumarins, lignins, and tannins. The total phenol content (TPC) of herbal tea infusions and the four Kombucha samples was determined by the Folin-Ciocalteu method and a gallic acid standard curve. There was a significant difference between the TPC of herbal infusion samples and also Kombucha samples (P < 0.05). The TPC of herbal tea infusion samples varied considerably, with the lowest value observed in LV (27.71 ± 0.53 mg GAE. L^-1) and the highest in GT (70.53 ± 1.07 mg GAE. L^-1) (Table 1). Following fermentation and Kombucha production, TPC increased markedly. The TPC results revealed a significant difference among Kombucha beverages that varied between 65.82±0.93 to 153.38±1.07 mg GAE. L^-1 (Table 2). The highest and lowest TPC were assigned to GTK and LVK, respectively. The TPC of GTK was 18, 57.8, and 32.87 % higher than that of BTK, LVK, and STK, respectively (Table 2). In addition to increasing acidity, the fermentation led to an increase in the TPC of herbal tea infusions. In this study, TPC of GTK, BTK, LVK, and STK was 2.47, 2.17, 2.37, and 2.14 higher than GT, BT, LV and ZM infusions, respectively. Similarly, Kim et al. stated that the TPC of Kombucha samples (70.87- 250.52 µg GAE. mL^-1) prepared by inoculation of various AAB and yeasts was 1.27‒3.53-fold higher compared with black tea infusion [33]. The TPC of studied herbal infusions is related to the inherent presence of phenolic and flavonoid compounds in these medicinal plants.

The increase in TPC during Kombucha fermentation could be attributed to several microbial activities. Microbial enzymes hydrolyze complex polyphenols into simpler and more soluble phenolic components, while microbial metabolism generates additional phenolic compounds such as gallic and caffeic acids. The results of all these changes during fermentation are an enhancement in TPC [35-37].

3.4.Antioxidant activity

According to Fig. 2, the antioxidant activity of the four types of Kombucha samples was varied (47.13 – 92.11 %). The GTK had the highest DPPH free radical scavenging capacity (92.11%). This feature can be attributed to the presence of more phenolic compounds in green tea compared to other herbal teas, as confirmed in the previous section. There was no significant difference between the antioxidant properties of BTK and GTK, unlike LVK and STK. Similar results were observed for herbal diffusions (Fig. 3). The highest and lowest DPPH scavenging activity belonged to GT (81.01 %) and LV (38.91 %), respectively. Due to the fermentation of sweetened herbal infusions during Kombucha preparation by fermenting microorganisms such as AABs, LABs and yeasts, production of various organic acids and phenolic compounds had occurred, which could increase the antioxidant capacity of the beverage. Generally, these components are active as electron donors and could scavenge the DPPH free radicals. In this regard, Jakubczyk et al. also reported various DPPH radical scavenging capacity of different Kombuchas, including green tea (88.23%), black tea (61.04%), white tea (70.42%), and red tea (74.78%) after 14 days of fermentation [38]. Malbaša et al. observed that green tea and black tea Kombucha may have different antioxidant activity depending on the type of initial culture of yeast and AABs used in Kombucha production at 28°C for 10 days [39]. The highest DPPH radical scavenging capacity of black tea and green tea Kombucha in the presence of different starter cultures was determined to be about 50 %. In addition to the concentration of phenolic compounds in Kombucha, other metabolites produced during fermentation, such as ascorbic acid and other organic acids, may also modify the antioxidant capacity of Kombucha. In addition to starter cultures and herbal infusion types, the antioxidant capacity of Kombucha is also affected by the temperature and fermentation time. In Jafari et al. research, the invertase activity was enhanced in black tea kombucha by increasing time, then the content of polyphenolic compounds and subsequently the antioxidant capacity rose to 285.44 mg GAE. L^-1  and 89.32%, respectively [40].

3.5. Antimicrobial properties

3.5.1. Antifungal activity

In this study, the antifungal activity of Kombucha samples against two spoilage molds was evaluated by the well diffusion method. According to Fig. 4, A. flavus was more sensitive to the compounds of fermented Kombucha beverages compared to A. niger. The largest inhibition zone against A. flavus (20.33 mm) and A. niger (10.33 mm) belonged to GTK sample. Also, no significant difference was observed among other Kombucha samples against A. niger (P>0.05). Moreover, no inhibition zone was observed for herbal infusions against both studied molds in the well diffusion test. These results are likely due to the low concentrations of phenolic compounds in the herbal infusions, indicating the need to use a greater amount of dry leaves when preparing these infusions in order to enhance their antifungal activity. In another antifungal activity evaluation method, unfermented herbal tea infusions also showed no antifungal activity. A. niger exhibited greater resistance to Kombucha compounds in the agar dilution method (Fig. 5).

By increasing the content of the four types of Kombucha in the solid medium from 10% to 40%, the percentage of growth inhibition was enhanced. The highest growth inhibition of A. niger (10.63%) occurred in the presence of 40% BTK or GTK, which is probably related to the higher organic acids and phenolic content of these types of Kombucha, as confirmed in the previous tests. In contrast, the highest growth inhibition (%) of A. flavus (57.14 %) occurred in the presence of 40 % GTK in the solid medium. Studies on the antifungal properties of Kombucha beverages against food-spoilage molds are limited. For example, black tea Kombucha produced in the study of Cetojevic-Simin et al. did not show any antifungal activity against A. niger and A. Flavus [41]. While in another study, the growth of A. flavus was inhibited in the presence of 5696 μg/mL black tea Kombucha [42]. Also, Kombucha produced from supercritical water and yarrow extract in the microdilution method demonstrated a minimum inhibitory concentration of 1.39 μg/mL against A. niger [13]. According to the report of Al-Mohammadi et al., fermented infusion, unfermented infusion, neutralized Kombucha, and heat-denatured Kombucha of black tea illustrated various antifungal activities. Black tea Kombucha showed higher antifungal activity against A. flavus and A.niger than heat-denatured Kombucha and neutralized Kombucha. While unfermented tea had no antifungal activity [43]. Generally, the antifungal mechanism of fermented Kombucha beverage is primarily attributed to its bioactive compounds, including organic acids (such as acetic acid), polyphenols, and enzymes produced during fermentation. These components collectively lower the pH, disrupt fungal cell membranes, and inhibit fungal growth through oxidative stress and metabolic interference. The synergistic effect of these metabolites creates an inhospitable environment for fungal proliferation, thereby conferring antifungal properties to Kombucha [1].

3.5.2. Antibacterial activity

According to Fig 6, in contrast to Gram-negative bacteria, no significant differences were observed in the antibacterial activity of the Kombucha samples against pathogenic Gram-positive bacteria (p>0.05). The highest and lowest antibacterial activities against E. coli and S. entritidis in the well diffusion assay were associated with the GTK and LVK Kombucha samples, respectively. The Gram-positive S. aureus, as the most sensitive bacterium, exhibited an inhibition zone of around 22 mm in the presence of GKT components. This observation is related to the highest total phenolic content and acidity and the lowest pH of GTK sample. Moreover, no antimicrobial activity was observed against Gram-negative bacteria for non-fermented herbal tea infusions (Fig. 7).

The highest antibacterial activity of non-fermented herbal tea infusions against both Gram-positive bacteria (L. monocytogenes and S. aureus) was obtained in the presence of GT and BT, with an inhibition zone ranging from 10.33 to 11.66 mm. According to these findings, fermentation of herbal infusions with Kombucha starters could improve antimicrobial characteristics. Similarly, Kombucha based on Black and green tea, lemon verbena (Lippia citriodora), and peppermint (Mentha piperita), demonstrated different antibacterial activity against Gram-positive and Gram-negative food-borne bacteria [16]. In another study, Green tea Kombucha showed higher antibacterial activity than black tea Kombucha in a microdilution assay due to the various contents of phenolic compounds and organic acids [25]. In contrast, no antibacterial activity was observed for Kombucha beverages prepared using banana peel, nettles, and black tea infusions in Ebrahimi Pure research [6]. The comparison of antimicrobial activity of fermented and non-fermented varieties of various herbal infusions (thyme, lemon verbena, rosemary, fennel, and peppermint) using the well-diffusion method against different bacterial pathogens, as well as the neutralization effect of Kombucha samples, revealed that the antibacterial activity varied depending on the type of Kombucha, and thyme Kombucha exhibited the strongest activity. Furthermore, neutralization resulted in a significant reduction in antimicrobial activity of the samples, indicating the role of organic acids in the antimicrobial properties. Also, a smaller inhibition zone was reported for unfermented herbal infusions compared to fermented Kombucha [44]. In this regard, unfermented tea and neutralized fermented tea derived from black, green, oolong, and mulberry teas showed no antibacterial activity [45]. In general, during the fermentation process of Kombucha, the levels of chemical components change rapidly, resulting in a rich concentration of organic acids, polyphenols, microbial enzymes, and bacteriocins, all of which contribute to its antimicrobial properties. Since an aqueous extract of medicinal plants was used in the present study, the comparatively weaker antioxidant and antimicrobial performance of ST and LV infusions may be primarily attributed to the limited extraction efficiency of their key bioactive constituents under water-based conditions and also the short extraction time. Several studies have established that the major antimicrobial compounds of Thyme, particularly thymol and carvacrol, are lipophilic phenolic terpenoids that are poorly soluble in water but are extracted more effectively with organic or alcoholic solvents [46]. Similarly, for Lemon verbena, although water infusions contain some phenolic compounds, the extraction of key polyphenols such as verbascoside and more hydrophobic constituents like citral is enhanced with moderate polarity solvents (e.g., ethanol-water mixtures), yielding significantly higher total phenolic content and antioxidant capacity than water alone [47].

3.6. Correlation Analysis

Pearson’s correlation analysis (N = 4) (Fig. 8) revealed distinct but complementary relationships among chemical, microbiological, and functional parameters in herbal infusions before fermentation and in the corresponding kombucha beverages. In herbal infusions, pH exhibited a very strong and statistically significant negative correlation with antioxidant activity (r = −0.993, p = 0.007). Strong negative correlations were also observed between pH and total phenolic content (TPC) (r = −0.932) as well as antibacterial activity (r = −0.903). These findings indicated that lower pH values were associated with higher antioxidant and antimicrobial capacity.  In this regard, TPC showed positive correlation with both antioxidant (r = 0.885) and antibacterial activities (r = 0.869), suggesting that phenolic compounds are major contributors to the bioactive properties of the infusions. Following fermentation, the correlation structure became more complex due to microbial activity. In kombucha samples, TPC showed a strong and statistically significant positive correlation with antioxidant activity (r = 0.953, p < 0.05) and an exceptionally strong correlation with AAB counts (r = 0.991, p < 0.01), indicating a close link between microbial metabolism and phenolic compound availability. Total acidity was negatively correlated with pH (r = −0.813) and positively associated with both AAB (r = 0.800) and LAB (r = 0.633), highlighting the role of fermentation-driven acid production. Although several correlations involving antioxidant, antibacterial activity, and microbial populations were not statistically significant, their high correlation coefficients suggest biologically meaningful trends, likely constrained by the limited sample size. Overall, these results demonstrate that while the bioactivity of herbal infusions is primarily governed by inherent chemical composition, fermentation introduces microbial-driven transformations that strengthen and diversify the relationships among chemical composition, acidity, and functional properties in kombucha beverages.

1. Conclusion

The findings of this study demonstrate that the type of herbal infusion used in Kombucha production plays a decisive role in determining its microbiological profile as well as its functional and antimicrobial properties. Across all formulations, AAB constituted the dominant microbial population, followed by LAB and yeasts, reflecting a stable and well-balanced fermentation ecosystem. The fermentation process markedly enhanced the biochemical characteristics of the beverages, resulting in increased total phenolic content, antioxidant activity, and acidity compared to the corresponding non-fermented herbal infusions. Among the evaluated samples, green tea Kombucha exhibited the most pronounced functional performance, particularly with respect to antioxidant capacity and antimicrobial activity against pathogenic bacteria and spoilage molds. In contrast, lemon verbena and thyme Kombuchas displayed relatively lower bioactive potential, although fermentation still led to notable improvements over their original infusions. These findings highlight Kombucha fermentation as an effective strategy for enhancing the nutritional and antimicrobial properties of herbal beverages. However, further studies are required to support their safe consumption and commercial development. Future research should prioritize cytotoxicity and safety evaluations, and minimum inhibition concentration (MIC) and minimum bactericidal concentration (MBC), as well as chemical and sensory analysis studies to determine palatability and market potential of these functional beverages.

Background and Objective: Regarding the extensive industrial use of proteases, the discovery and detailed characterization of novel protease enzymes derived from native isolated microbial strains are scientifically and industrially important.

Material and Methods: In this study, a novel recombinant alkaline serine protease was heterologously expressed in Escherichia coli BL21. Following the expression, the enzyme was purified and subjected to biochemical characterization and in silico analysis.

Results and Conclusion: The purification process yielded an efficiency of 79.5% and the specific activity of the purified enzyme was assessed as 0.12 U mg^-1. The enzyme showed significantly thermal stability, preserving approximately 50% of its activity after 3 h of incubation at temperatures ranging 60-70 °C. Furthermore, pH stability assays demonstrated that the enzyme was still functionally stable over a wide pH range, preserving 60-80% of its residual activity after 48 h, particularly under neutral to alkaline conditions. Exposure to sodium dodecyl sulfate, hydrogen peroxide and ethylenediaminetetraacetic acid included partial effects, with almost all enzyme activity was seen. However, the presence of β-mercaptoethanol, phenylmethyl sulphonyl fluoride and NaCl significantly decreased enzymatic activity, with decreases of approximately 50%. In silico analysis identified a conserved peptidase S8 and thermitase-like domain spanning amino acid residues of 40 to 270. Catalytically essential residues Asp76, His109 and Ser263 were predicted within the active site. Regarding its robust stability under increased temperatures and alkaline conditions with favorable catalytic characteristics, the recombinant serine protease promises its uses in various industrial processes, compared to the commercial counterpart enzymes.

Keywords: Bioinformatics, Expression, Purification, Recombinant alkaline protease

1. Introduction

Proteases are one of the most important groups of industrial enzymes that are used in various industries [1,2]. Based on the Enzyme Commission classification (EC3.4.), proteases belong to Class 3 (hydrolases) and Subgroup 4 (hydrolyzing peptide bonds). The global market is expanding annually, accounting for approximately 60% of the total enzyme industry market [3]. Alkaline protease enzymes include several uses in fermentation, textile, leather, detergent, pharmaceutical, livestock and poultry food supplement production, lens and teeth cleaners, agriculture and food industry [1–3]. Alkaline serine proteases are the most important group of proteases used commercially [3]. Alkaline serine proteases include a molecular weight (MW) of 18–42 kDa and are active in a pH range of 6–11 at a temperature of 50–70 °C [3]. Proteases are used to produce cheese because proteases prevent the coagulation of milk by hydrolyzing casein [4]. In bakeries, proteases are used for faster dough preparation and gluten hydrolysis. The production of protein products with high nutritional value is carried out by adding microbial alkaline proteases. These protein products are vital in preparing baby foods and enriching soft drinks and fruit juices [1,5]. Researchers have reported that the use of protease in poultry feed increases the digestibility of food, improves the absorption of nutrients and digestibility of amino acids in the intestine, improves the digestibility of raw protein and amino acids, improves the growth performance of chickens and increases the food yield [6–8]. In the detergent and food industries, proteases are used, which include high thermal stability and their optimal activity is in alkaline conditions. Bacillus species are the major producers of extracellular proteases as microbial sources [1,2]. Since the production and extraction of commercial enzymes from wild-type microbial species include challenges, production of recombinant enzymes uses natural enzyme genes [9] and it seems necessary to identify these enzyme genes to expand the gene database for their recombinant production on an industrial scale. In addition to experimental studies, bioinformatics studies are useful for characterizing the novel enzyme and identifying its industrial use and production [4]. This study continued a previous study; in which, an alkaline serine protease was identified and its gene was cloned into an expression vector. Further investigations were carried out in this study to express and characterize the purified enzyme [9,10]. In 2018, Hadjidj et al. purified and biochemically and molecularly characterized a novel alkaline serine protease from Bacillus (B.) licheniformis K7A [11]. In 2018, Singh et al. isolated and characterized a serine protease enzyme from B. subtilis K-1 [12]. In 2013, Joshi et al. investigated the characteristics of an alkaline serine protease isolated from B. lehensis Strain MTCC7633 and identified its gene (BLAP), which was then expressed in Escherichia (E.) coli [13]. In 2012, Kaur et al. identified and isolated an alkaline protease gene from B. circulans MTCC 7906 and transferred and expressed the gene into E. coli DH5-α [14].  In 2010, Sadeghi et al. isolated an alkaline protease gene from B. subtilis, cloned it into the pTZ57R cloning vector and transferred it to E. coli Hb101 [15]. In 2021, Gurunathan et al. cloned a protease gene from a B. cereus strain from shallow marine hydrothermal vents located in the East China Sea and expressed the gene in the host E. coli BL21 [16].

In the present study, the expression parameters of the recombinant alkaline protease were optimized, followed by purification through affinity chromatography. Then, biochemical characteristics of the purified enzyme were characterized to assess its potential applicability in industrial processes. Based on experimental data and analyses carried out using ExPASy serverwith comparative modelling approaches, the three-dimensional (3-D) structure of the protein and amino acids of its active site were predicted. In this study, a novel recombinant alkaline protease was expressed and purified and its characteristics were described. The innovation of this study included its strong stability at high temperatures and alkaline conditions, as well as favorable catalytic characteristics, which are useful for use in various industrial processes and can be an appropriate alternative to its commercial types.

1. Materials and Methods

2.1. Identifying and isolating the alkaline protease gene

In a previous study, Bacillus sp. RAM53 was isolated by Amol Rice Research Institute and identified using biochemical and microbiological assays and molecular analysis [14]. Then, 1194 bp of the alkaline protease gene extracted from Bacillus sp. RAM53 was cloned in pET28a ⁺ [9] and used in this study for further studies. The gene analysis in Blastn showed 94% similarity with alkaline protease genes of Bacillus spp. Phylogenetic tree associated to the 16S rRNA sequence of Bacillus sp. RAM53 showed the proximity to B. cereus and B. thuringiensis [9].

2.2. Expression of the recombinant enzyme and its purification

First, a colony of recombinant bacteria was cultured overnight in 10 ml of Luria Bertani (LB) media containing kanamycin and then inoculated into 500 ml of fresh media [10]. When the bacterial growth at OD600 nm reached 0.4–0.6 in the bacterial culture media, isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to a final concentration of 1 mM for induction; then, culture was incubated at 20 °C for 20 h. After lysis of the cell wall of bacteria using sonication, the cell extract was precipitated with 80% ammonium sulfate salt. This was centrifuged at 11,000× g for 20 min at 4 °C to separate the precipitate of proteins. After centrifugation, the protein precipitate was dissolved in 20 mM Tris buffer with pH 9. To remove excess salt from the proteins, this was dialyzed in buffer containing 10-mM imidazole using 15-kDa cut-off dialysis bag. To purify the recombinant enzyme due to the presence of a histidine sequence at the carboxyl end of the recombinant protein, cobalt sepharose-resin affinity chromatography (Arg Biotech, Iran) was used. Then, 4 ml of the protein sample were poured onto the column (1.5 × 10 cm) at a flow rate of 1 ml min^-1. Imidazole concentration gradient was used to elute the protein samples from the chromatography column. The elution buffer consisted of two buffers of A and B; where, Buffer A contained 50 mM Tris, pH 7 and 1% Triton X-100; and Buffer B contained 150-mM imidazole and the concentration gradient was chosen 50–100% of Buffer B. After purification, the samples containing the recombinant enzyme were filtered to separate the additional proteins from the recombinant protein using 50-kDa cut-off filtration column. Then, 10% SDS-PAGE gel electrophoresis was used to verify the presence of recombinant protein in the samples.

2.3. Western Blot Analysis

Proteins separated by SDS-PAGE were transferred onto a nitrocellulose membrane using transfer buffer containing 39-mM glycine, 48-mM Tris Base, 0.037% SDS and 20% methanol, and Bio-Rad Mini-Protean Tetra Cell System (BioRad, USA). Following the transfer, membranes were blocked with 5% BSA in PBS (pH 7.0) for 2 h at room temperature (RT) to prevent nonspecific binding. The membrane was then incubated with monoclonal anti-6xHis tag antibody (1:3000 dilution in PBS containing 1% Tween 20) for 2 h at RT with gentle agitation. After three washes with PBS/T, the membrane was incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG secondary antibody (Sigma-Aldrich, USA), diluted 1:5000 in PBS/T containing BSA, at RT for 1 h with gentle shaking. Following another set of three PBS/T washes, protein detection was carried out using HRP chromogenic substrate. The enzymatic reaction was terminated by rinsing the membrane twice with distilled water (DW) [17].

2.4. Enzyme activity assay

For the standard assessment of total enzyme activity, 150 µl of the pure enzyme were mixed with 150 µl of 1% casein as substrate and 300 µl of 50 mM Tris Buffer with pH 9 and then incubated at 40 °C 30 min [10]. The absorbance after enzyme reaction at 280 nm included the absorbance of caseins broken by protease activity, which included a direct relationship with enzyme activity. Based on the definition, a unit is the quantity of enzyme that releases the fragments of peptides produced under standard conditions (0.001 A₂₈₀ nm) within 30 min.

2.5. The effects of temperature and pH on the stability of the purified enzyme

To investigate the effect of temperature on the activity and stability of the recombinant enzyme, the enzyme cocktail was incubated at temperatures of 0, 30, 40, 50, 60 and 70 °C for 3 h and then transferred into ice for 10 min and the quantity of residual enzyme activity was assessed using the standard method [18]. To investigate the effect of various pH levels on enzyme stability, an enzyme cocktail was prepared in various buffers. Sodium citrate buffer for pH 5 and 6, Tris Buffer for pH 7, 8 and 9 and glycine buffer for pH 10, 11, 12 and 13 were used. The enzyme cocktail was incubated for 0 and 48 h at RT and then the rest of enzyme activity was assessed using the standard method [18].

2.6. The effects of inhibitors, salt and detergent on the stability of the purified enzyme

First, the pure enzyme was treated with 10-mM PMSF, 10-mM EDTA, 0.5% SDS, 0.5-M NaCl and 5% H₂O₂ for 1 h at RT. Then, residual enzyme activity was assessed using standard method.

2.7. Statistical analysis

Statistical analysis was carried out to compare the differences between the treatments and control using Duncan's method and SPSS software v.22 (IBM, USA) and a significance level was recorded at p < 0.05.

2.8. Bioinformatics studies

Briefly, Blastn (NCBI) was used for the gene sequence alignment of the recombinant enzyme. The protein sequence information was used to identify the homologous protein sequences in Blastp (protein-protein BLAST). Protein sequences that were more than 80% identical to the recombinant enzyme sequence and originated from the Bacillus genus were used for further analysis. The position and types of amino acids present in the active site of the recombinant enzyme and its active domains were assessed using Merops-the Peptidase Database (https://www.-ebi.ac.uk/merops/), ProScan (http://www.ebi.ac.uk/-interpro/result/InterProScan/) and BLAST (https://blast.-ncbi.nlm.nih.gov/Blast.cgi) databases. The number of partial amino acids, MW and theoretical pI, instability index (II) and aliphatic index were computed using ProtParam and the protein sequence (https://web.expasy.org/protparam/).

The 3D structure of the protein was constructed using Swiss-model and the comparative modelling, corresponding to its amino acid sequences and a template selected from PDB with further protein sequence identity. The amino acids present in the active site of the enzyme were identified in the structure. Structure assessment of the model in Swiss-Model was carried out using alignment, Ramachandran plot and QMEAN (qualitative model energy analysis) local score (https://swissmodel.expasy.org/qmean/RzEH8V).

1. Results and Discussion

The recombinant enzyme was purified through cobalt affinity chromatography and the purification efficiency was assessed as 79.5%. The graph associated to purification is seen in Figure 1. The single long peak was associated to the release of the recombinant enzyme, which was eluted from the column at a concentration gradient of 90–100% elution buffer. The purified recombinant enzyme concentration included 130 µg ml^-1. The specific activity of the purified recombinant enzyme was 0.12 U ml^-1 mg^-1, and the specific activity of the recombinant enzyme before purification was 0.066 U ml^-1 mg^-1. The enzyme samples extracted from the cell extract after induction, as well as the same sample after concentration with ammonium sulfate salt and the purified recombinant enzyme sample, were electrophoresed on 10% SDS-PAGE gels. Western blot analysis with anti-His tag antibodies revealed a distinct band at approximately 41 kDa, corresponding to the expected MW of the recombinant peptide. The six-His-tag, engineered at the C-terminal end of the recombinant protein for affinity purification, was still intact and was not cleaved from the peptide, indicating successful expression and detection of the full-length His-tagged recombinant protein (Figure 2). The three major elements in the successful production of recombinant protein included host condition, culture media condition and vector type. One of the appropriate microbial strains for exogenous gene expression was the E. coli strain [19].

The activity of the recombinant enzyme was assessed before and after purification. The quantity of absorbance after enzyme reaction at 280 nm included the absorbance of caseins broken by protease activity, which included a direct relationship with enzyme activity. Based on the definition, a unit is the quantity of enzyme that releases the fragments of peptides produced under standard conditions (0.001 A₂₈₀ nm) within 30 min. In the present study, a novel recombinant serine alkaline protease derived from the native strain of Bacillus sp. RAM53 was expressed in a laboratory environment. The culture conditions were optimized for optimal production of this recombinant enzyme. The recombinant enzyme gene and protein with the scientific name of subtilase-like protease (MARF) were deposited in GenBank with accession numbers of OP922123.1 (https://www.ncbi.nlm.nih.gov/nuccore/OP922123.1) and WCO70948.1, respectively. The pure enzyme did not decrease its activity; instead, it increased its specific activity.

The specific activity of the purified recombinant enzyme was 0.12 U ml^-1 mg^-1, and the specific activity of the recombinant enzyme before purification was 0.066 U ml^-1 mg^-1. The concentration of the recombinant enzyme was assessed based on Bradford's standard method using a standard curve. The concentration of the recombinant enzyme was 0.3 mg ml^-1 before purification and 0.13 mg ml^-1 after purification.

Results of the thermal stability of the pure enzyme at various temperatures showed that the pure recombinant enzyme included relatively good stability for 3 h at 30 and 40 °C; hence, more than 70% of its activity were included. At 70 °C, 46% of the enzyme activity remained (Fig. 3). At all temperatures, a significant difference from the control was observed (p<0.05). The recombinant enzyme demonst-rated relatively good stability at high temperatures such as 60 and 70 °C without addition of stabilizing ions, compared to commercial enzymes. The subtilisin family of serine proteases represent an important class of industrial bio-catalysts widely used in the food industry [3]. Their extensive use is largely attributed to their favorable bio-chemical and catalytic characteristics. In addition, their high thermostability allows them to preserve activity under increased temperatures associated to industrial heat treat-ments [3,5,8,20]. The recombinant enzyme of the present study included good temperature resistance, compared to commercial enzymes.

Results of the stability to various pHs showed that the enzyme was stable at various pHs after 48 h of treatment, and its residual activity included 60–80% at neutral to alkaline pHs (Figure 4).

At all pHs, a significant difference with the control was observed (p < 0.05), and it showed similar stability at neutral to alkaline pHs, except for pH 9, and the highest stability was reported at pH 9. Subtilisin typically demonstrated significant stability under alkaline pH conditions, making it particularly appropriate for protein hydrolysis processes commonly used in food manufacturing [3,21].

Aryaei et al. identified and isolated an alkaline protease gene from a native strain of Bacillus sp. RAM and cloned the gene into pET28a+. The enzyme showed maximum activity at pH 9 and 40 °C [10]. Moradian et al. purified an alkaline serine protease from Bacillus sp. HR-08 and the protease activity was at pH 10 and 60 °C [18].

A thermostable serine-protease enzyme from Aeribacillus pallidus c10 strain was purified by Yildirim et al. and the optimal activity of this enzyme was at 60 °C and pH 9 [22]. Bervibacillus brostelensis isolated from a hot spring in Qeinerjeh produced an alkaline protease and showed maximum activity at 60 °C and pH 9 [23]. Alkaline proteases produced by Bacillus species are greatly important due to their thermal stability and stability at various pH levels, which are important in industries [1,24].

Results of pure enzyme treatment with 10-mM PMSF, 10-mM EDTA, 0.5% SDS, 0.5-M NaCl and 5% H₂O₂ after 1 h are reported in Figure 5. The EDTA did not affect the activity, and beta-mercaptoethanol decreased the activity of the enzyme a little with no significance (p < 0.05). The PMSF and NaCl decreased the enzyme activity by nearly 50% with significance (p < 0.05). In the presence of SDS and H₂O₂, almost all enzyme activity was preserved.

Several members of the S8 family (e.g. subtilisin) bind calcium for stability; inhibition can be seen with EDTA and EGTA, which are often reported as specific inhibitors of metallopeptidases. Calcium binding stabilizes these proteins in natural extracellular environments but poses a problem in industrial environments containing high concentrations of metal chelators. Subtilisin Carlsberg possesses three Ca²⁺ binding sites that stabilize its 3-D structure. The important industrial enzymes, subtilisin Carlsberg and subtilisin BPN, respectively included thermostability of 3.4 and 2.4 min at 50 °C and their thermal stability enhanced in presence of calcium [25,18]. These proteases are addressed as important commercial proteases and, therefore, the recombinant enzyme of the present study includes good temperature resistance, compared to commercial enzymes. In addition, the gene of this enzyme originates from B. mesophilus. A serine protease gene of B. megaterium, ispK, was cloned into E. coli. The thermostability increased 32.9-fold from 3.3 to 108.5 min at 60 °C in the presence of 2 mM of Ca²⁺ [26]. In the present study, the recombinant enzyme activity did not decrease in the presence of a chelating agent, resulting in calcium ions not neccessary for activity and stability in the structure. This characteristic is important for use in industry. Bacillus proteases have been shown stable in a wide range of pH and temperature [27]. A serine alkaline protease purified from Bacillus sp. HR08 was stable to hydrogen peroxide and its activity was complete [18].

The recombinant enzyme protein with the scientific name of subtilase-like protease was deposited in GenBank with the accession no. of WCO70948.1

(https://www.ncbi.nlm.nih.gov/protein/WCO70948.1). Based on the sequence alignment of S8 family peptidase (Bacillus) using Blast, NCBI, a sequence with reference number WP_000790938.1 was selected, which had nearly 86% identity to the recombinant enzyme sequence (Table 1). Then, the information on the amino acids of its active site was used to find amino acids in the active site of the recombinant alkaline serine protease. Based on the alignment and sequence homology, which included 86% identity with the subtilase family and available information on the active amino acids in the active site of the enzymes of this family, aspartic acid, histidine and serine (MEROPS - Peptidase Database) were assessed for the active site of the recombinant enzyme. The S8 termitase-like peptidases region is located at amino acids 40 to 270 in the recombinant enzyme.

    In sequence analysis using ProScan (http://www.ebi.ac.uk/interpro/result/InterProScan/), the type of protease from the S8 subtilase family and the presence of a domain similar to thermitase were verified (Figure 6a). In addition, presence of Asp76, His109 and Ser263 in the active site of the enzyme was verified using ProScan and PROSITE (Figures 6b and 6c). Based on ProtParam prediction, the instability index (II) was computed as 18.38. This classified the recombinant protein as stable for more than 10 h in E.coli expression system. The aliphatic index of the recombinant enzyme was assessed as 71.52 as a thermostable protein. Several important subtilisins have been produced using various bacilli, including subtilisin Carlsberg from B. licheniformis; subtilisin BPN and Dj-4 from B. amyloliquefaciens; subtilisin E, NAT and J from B. subtilis; and subtilisin amylosacchariticus from B. amylosacchariticus [27]. The alkaline protease gene (Apr) from B. licheniformis 2709 was cloned into an expression vector of pET28b (+) and expressed in a high-expression strain of E. coli BL21 [29]. The recombinant enzyme included identities with S8 family peptidase from B. cereus, S8 family peptidase from Bacillus sp., S8 family peptidase from B. firmicutes and S8 family serine peptidase from B. thuringiensis as 86, 86, 86 and 85%, respectively. The subtilisin family is one of the largest serine peptidase families characterized to date. Structures have been assessed for several members of the subtilisin family that contain similar catalytic triad residues in an order of DHS (Asp, His, Ser)

(https://www.ebi.ac.uk/interpro/entry/InterPro/IPR034202/). Subtilisin possess broad substrate specificity, enabling efficient degradation of diverse protein substrates and demonstrates strong catalytic efficiency, which contributes to reduce processing times and improved production yields. Collectively, these characteristics highlight the significant potential of S8/thermitase-like proteases in various food processing uses [3,21]. The tertiary structure of the present alkaline serine protease was made based on a template selected by the Swiss model building. The template included thermitase with PDB code 1thm.1A, as well as sequence identity of 85% (Figure 7a). In Ramachandran plot, distribution of the backbone dihedral angles (ϕ and ψ) showed that a majority of residues were located in the most favored regions (dark green areas), primarily corresponding to α-helical and β-sheet conformations. A smaller fraction of residues was set within additionally allowed regions (light green areas), which was acceptable for certain amino acids in specific conformations. Overall, the plot suggests that the protein model includes good stereochemical quality, with most residues adopting energetically favorable conformations and a minimal number of outliers that are unlikely to affect the structural integrity significantly (Figure 7b). The local quality assessment in QMEAN, expressed as the predicted local similarity to the target, showed that the majority of residues included high similarity scores (> 0.8), indicating reliable modelling in most structural regions. Peaks close to 0.9–1.0 suggested well-defined secondary structures with high confidence in backbone conformation. However, a few distinct regions particularly near residues of 25–30 and 235–245 showed significantly lower quality scores (< 0.6), with the lowest approaching 0.3. These decreases often corresponded to loop regions, flexible termini or poorly conserved areas in the template alignment. Overall, the QMEAN profile supported a generally reliable global structure, with local weaknesses restricted to specific flexible or unresolved segments (Figure 7c). Using experimental data and ExPASY analysis as well as comparative modelling, this study predicted the protein 3-D structure and active site amino acids. Assessmentn of the predicted 3-D structure using Ramachandran plot revealed that a majority of residues were located in the most favored regions, with only a few residues in additionally allowed areas and a minimal number in disallowed regions. This distribution indicated that the backbone dihedral angles were largely in stereochemically favorable conformations, supporting the overall correctness of the model [29]. Collectively, these results suggest that the predicted protein model possesses good overall stereochemical quality and global reliability.

SWISS-MODEL is a fully automated protein structure homology-modelling server, accessible via the Expasy web server or DeepView program (Swiss Pdb-Viewer). Homo-logy modelling allows the protein engineer to design modifications of a protein before the determination of the 3-D structure using X-ray technique. Of all current theoretical approaches, comparative modelling is the best method that can reliably generate a 3-D model of a protein from its amino acid sequence [30,31]. The alkaline protease from thermophilic bacteria included an aliphatic index of 69.4 [32], while that of the recombinant enzyme was assessed as 71.52, showing that the recombinant enzyme included higher aliphatic index than the thermophilic type and thermal stability despite its mesophilic origin, compared to other thermophilic enzymes.  This index should be addressed as a positive factor for increasing the thermal stability of globular proteins [33].

More than one million proteolytic enzyme sequences have been recorded in the Merops Database [34]. Therefore, identifying novel strains with the ability to produce enzymes appropriate for industrial uses and genes of these enzymes for the production of recombinant enzymes has always been useful. Especially in countries that buy industrial enzymes, this decreases the purchase costs. In addition to laboratory studies, bioinformatics studies for characterizing alkaline proteases help identify their industrial use and large-scale production. Therefore, protein-structure prediction tools are appropriate for the study of protein structures and functions [35].

1. Conclusion

The current study focused on the biochemical characteristics and tertiary structure, physicochemical characteristics and catalytic potential of a novel recombinant alkaline serine protease. Despite lack of sufficient funding to sequence the N-terminal region of the purified protein that was a limitation of the present study, structural engineering methods can be used to improve enzyme production for the production of a novel high-quality recombinant enzyme on an industrial scale by predicting the tertiary structure of this enzyme using bioinformatics. Given its favorable biochemical characteristics, including pH stability over the range of 6–13, thermal stability at 60 °C, independent of Ca²⁺ and resistance to SDS and oxidizing agent of H₂O₂ with its classification as an alkaline protease, the enzyme demonstrates strong potential for use in various industrial sectors such as detergent, food and animal and poultry feed productions.

Background and Objective: Foodborne infections represent a significant global health concern, affecting numerous countries. From these, Salmonella Typhimurium ATCC 13311 is a prevalent foodborne pathogen, contaminating poultry products, especially chicken meats. This study investigated the synergistic effect of nisin in combination with allicin against Salmonella Typhimurium ATCC 13311, aiming to enhance the preservation of chicken meats from contamination.

Material and Methods: The inhibitory effect of nisin and allicin against Gram-negative bacterial strains was assessed using microbroth dilution method. Their synergistic effect was investigated through a checkerboard assay. Kinetics of inactivation, membrane permeability using β-galactosidase analysis, hemolytic effect, preservative potentials against Salmonella Typhimurium and calorimetric analysis were assessed.

Results and Conclusion: Nisin and allicin inactivated the proliferation of the test bacteria at minimum inhibitory concentrations of 0.063 and 0.016 mg ml^-1, respectively. A combination of the two preservatives was used to treat meat contaminated with Salmonella Typhimurium ATCC 13311. At 35°C, nisin and allicin minimum inhibitory concentrations significantly decreased Salmonella Typhimurium ATCC 13311 colony counts to 2.24 ±0.1 and 2.29 ±0.09 log CFU g^-1 after 30 min (p = 0.0193). Moreover, nisin and allicin resulted in 0.08 ±0.02 and 0.09 ±0.02 log CFU g^-1, respectively, at 4 °C with no statistical difference compared to the control (p = 0.1064). However, a significant decrease in the population of Salmonella Typhimurium ATCC 13311 at 35 and 4 °C (p = 0.001) was observed when treated with the combination of the two compounds at 2× fractional inhibitory concentration (3.44 ±0.04 and 1.09 ±0.04 log CFU g^-1, respectively). Single and combined concentrations were non-hemolytic at minimum inhibitory concentrations and 2× fractional inhibitory concentration. The study suggested the use of nisin and allicin concoctions as a novel approach to control Salmonella Typhimurium ATCC 13311 induced spoilage of chicken meats.

Keywords: Allicin, Chicken meat, Nisin, Synergistic effects, Salmonella Typhimurium

1. Introduction

The Food and Agriculture Organization of the United Nations (FAO UN) [1] reported the output of nearly 73 million tons of eggs and poultry meat, exceeding 100 million, 2016. These figures are consistently accelerating due to an increase in the global population, which contributes to rapid economic growth and urbanisation [2]. However, projections for chicken production in 2024 were reported as 146 million tons, an increase of 0.8% over previous years [3]. This increase was due to high consumer demands for chickens, which are affordable despite global inflation. Additionally, developing and developed countries demand poultries and their products [4]. Salmonella foodborne infections account for approximately 85% of the annual global estimates (200 million incidences), with 93% resulting in gastroenteritis and 155,000 fatalities [5]. Moreover, Salmonella infections from chicken and turkey meat consumption cause 23% of the total foodborne infections (1.35 million) in the US, as reported by the Interagency Food Safety Analytics Collaboration [6].

In most food industrial settings, chlorine and organic acids are used in disinfecting materials used in meat processing; however, Salmonella spp. can evade the effect of such chemical interventions, especially when the bacteria produce biofilm [7]. However, biopreservation uses metabolic byproducts and compounds from microorganisms, plants and animals to preserve the sterility of foods while ensuring its safety [8]. Bacteriocins and plant-derived compounds such as nisin and allicin include biopreservative potential against various species of foodborne pathogens. Therefore, they can be used to minimize the challenges faced by food industries when preserving chicken meat against Salmonella spp.

Nisin is one of the purified bacteriocins and accepted as a food additive by the European Union (EU) [9]. A study detected that 0.5% concentration of nisin decreased the number of Listeria monocytogenes to 2.0 log colony-forming units per gram (CFU g^-1) in processed turkey ham upon storing at 4 °C for 63 d [10]. Therefore, nisin has been proven as an additive for meat preservation, contributing to the quality and integrity of food in response to consumer demand. Despite the limitation of nisin towards Gram-negative bacteria, its synergy with other natural preservatives can enhance its effectiveness and preservative potential against Gram-negative bacteria, specifically Salmonella Typhimurium, causing chicken meat spoilage [11]. Allicin is a prominent compound of garlic (Allium sativum), with a broad spectrum of activity on various microbial cells using multiple targets, which makes it difficult for bacteria to resist its effect [12]. Allicin can bind to bacterial membrane proteins and phospholipids, enhancing bacterial outer membrane permeability [13]. However, use of allicin alone as a biopreservative may experience unavoidable limitations, especially at high doses, which can cause cytotoxic effects and disrupt the normal gut microbiota [14].

Several studies have reported on the antimicrobial efficacy of nisin on Gram-negative bacteria, including Salmonella Typhimurium [15][16]. Moreover, studies were reported by various scientists on the efficacy of allicin on Gram-negative and Gram-positive bacteria, which were detected effective [17]. However, there is insufficient data on the synergy between the two bioactive agents (nisin and allicin) in inhibition of Salmonella Typhimurium, particularly in fresh chicken meats. These deficiencies created significant gaps for investigating their combined actions in the real food system (chicken meats) and practical context regarding its quality and safety. Although the two compounds have been investigated individually, by verifying their synergistic inhibition against Salmonella Typhimurium under real-life meat storage conditions, the study has provided a better alternative for effective poultry preservation; thereby, decreasing use of synthetic and chemical preservatives in the food industry. Therefore, this study assessed the synergistic antibacterial efficiency of nisin and allicin against Salmonella Typhimurium ATCC 13311 associated with chicken meat contamination under various storage settings.

1. Materials and Methods

2.1 Collection and Preparation of the Bacterial Strains

For this research, bacterial strains (Pseudomonas aeruginosa DMST15501, Salmonella Typhimurium ATCC 13311, Escherichia coli ATCC 8239 and Vibrio spp.) were provided from the Microbiology Laboratory, King Mongkut’s University of Technology Thonburi, Thailand. Ten microliters (10 μl) of refrigerated bacterial strains were inoculated onto sterile Luria-Bertani (LB) agar plates and incubated at 37 °C overnight. To achieve a pure culture of the strains, the incubated isolates were re-subcultured and incubated at 37 °C for 24 h.

2.2 Preparation of Nisin and Allicin Stock Solutions

Nisin used in this study was procured from Bangkok Chemical, Thailand, while allicin was purchased from Wang Zhao, China. Nisin and allicin were prepared in mg ml^-1. For nisin (2.5% w/w), 10 mg were dissolved in 1 ml  of 0.02-N hydrochloric acid (HCl) and set in boiling water (100 °C) for 4 min to enhance its solubility and achieve a uniform solution [15]. However, 2 mg of allicin powder (98% w w^-1) were dissolved in 2 ml of 5% dimethyl sulfoxide (DMSO) [18]. The stock solution of nisin and allicin was set at 4 °C before use, with allicin stock wrapped with aluminum foil. The stocks were filtered using 0.22-µm membrane filter (Millipore, USA) to achieve sterility.

2.3 Assessment of Minimum Inhibitory Concentration and Minimum Bactericidal Concentration

To assess the least concentration of the preservatives capable of inhibiting the growth and/or killing the bacterial cells as minimum inhibitory concentration (MIC), a few colonies of the pure isolates were inoculated into tubes containing 1 ml of LB broth and incubated for 2 h. The absorbance of the bacterial suspension was adjusted to 0.5 McFarland level (10⁵ CFU ml^-1). After adjustment, 50 μl of the standardized suspension were transferred to the 96 microwells and gently mixed with 50 μl of the prepared concentrations of nisin and allicin, separately. Wells containing LB broth without bacterial suspension were used as negative control, whereas wells containing LB broth and bacterial suspension without preservatives were set as positive control. The analysis was carried out in triplicate, followed by incubation at 37 °C for 18 h. The incubated agar plates were interpreted using microplate reader at 600 nm. The lowest concentration with the same optical density (OD) as that of the positive control was reported as MIC. Ten microliters from the MICs and subsequent concentrations were used to enumerate the MBC by plating onto a labelled sterile nutrient agar and incubating at 37 °C overnight. The concentrations that showed no growth were reported as MBC.

2.4 Inhibitory Synergistic Effects of Nisin and Allicin

The response of the bacterial strains to the combination of nisin and allicin was assessed in the ratio of one-to-one (1:1) as reported by Field et al. [19] with minor modifications. Shortly, fractions of MIC results (1/4) of each preservative were used against the bacterial strains by adopting the checkerboard method. Briefly, 50 μl of the MIC of each preservative were dispensed into a 96-well microplate in vertical and horizontal arrangements. The MICs were further diluted accordingly into fractions as previously stated. The other wells contained the mixture of their corresponding dilutions (25 μl from each preservative). A positive control well was set containing LB and preservatives. The standardized bacterial suspension (10⁵ CFU ml^-1) was dispensed into every well, mixed gently and incubated at 37 °C overnight. After incubation, the OD of the combinations was observed using microplate reader, compared to the positive control. The inhibitory synergistic effect of the combination was assessed by calculating their fractional inhibitory concentration index (FICI), as demonstrated by Hossain et al. [20].

FICI = MICN / MICN’ + MICA / MICA’

Where, MICN was the MIC of nisin combined, MICN' was the MIC of nisin alone, MICA was the MIC of allicin combined, and MICA' was the MIC of allicin alone. FICI ≤ 0.5 was synergistic, 0.5-1.0 was additive or partially synergistic, 1.0-4.0 was indifferent and > 4.0 was antagonistic.

2.5 Kinetics of Inactivation

The rate; at which, the cells of Salmonella Typhimurium ATCC 13311 decreased over time upon treatment with nisin and allicin, separately and in combination, was assessed by measuring their OD. A 96-well microplate was used; in which, MICs of nisin, allicin and FICs were dispensed into separate wells in triplicate. A standardized bacterial suspension (10⁵ CFU ml^-1) was added to each well. Luria-Bertani broth with bacterial suspension was set as control. The decrease and the increase in their OD were measured using microplate reader at 600 nm every 1 h for 5 h, adopting the procedure of Ajingi et al. [15] with adjustments.

2.6 Beta-galactosidase Analysis as a Model of Membrane Permeability

To assess the major contributing factor toward the synergism of nisin and allicin on E. coli ATCC 8239, its ability to disrupt the bacterial cell membrane (CM) needs assessment. Therefore, this method measured the ability of the β-galactosidase enzyme to leak out of the bacterial cytoplasm into the extracellular space by converting o-nitrophenyl-β-D-galactopyranoside (ONPG) into o-nitrophenol (ONP) [21]. To induce the expression of β-galactosidase, E. coli ATCC 8239 was grown in lactose broth and incubated at 37 °C with shaking (200 rpm) until the OD reached 0.4 at 600 nm. The cells were concentrated by spinning at 5000 rpm for 5 min at 4 °C. The media residue was washed out of the sediment by washing with phosphate-buffered saline (PBS) (pH 7.4). The sediment was diluted in PBS and adjusted to 0.1 OD (10⁷ CFU ml^-1) at 600 nm. Fifty microliters (50 µl) of the bacterial suspension were added to 50 µL of MICs of nisin and allicin and their FICs in 96-well plates, followed by 30 µl of ONPG (20 mM). Then, 50 µl of PBS mixed with 50 µl of the bacterial suspension were set as control. The 96-well plate was incubated at 37 °C for 30 min and the OD was read at 405 nm using microplate reader.

2.7 Hemolysis Assay

Red blood cell (RBC) lytic assay was carried out following the procedure explained by Jiddah et al. [21] with minor adjustments. The RBCs were washed in 1× PBS three times and centrifuged at 14,530× g for 10 min. The sediment was dissolved in 1× PBS to achieve a 4% concentration. In general, 500 μl of the RBC dilution were mixed with 500 μl of nisin and allicin, separately and in various combinations (1× FIC and 2× FIC), using tubes. The positive control consisted of a solution with 0.1% Triton X-100, whereas the negative control comprised 1× PBS. The tubes were incubated in a microtube heating block at 37 °C for 1 h and then centrifuged at 14,530× g for 5 min. Then, 100 μl of the supernatant were pipetted from each tube and dispensed into the corresponding wells of the 96-well plate. The release of hemoglobin was assessed by reading the absorbance at 540 nm.

2.8 Inoculation of Chicken Meat Samples

Fresh chicken meat was purchased from the market and washed thoroughly with distilled water (DW) to remove any contaminants. Then, this was cut into pieces (10 g each) using sterilized knife and laminar flow hood. A loopful of fresh overnight culture was transferred to a tube containing 1 ml of LB broth and incubated for 2 h with constant agitation (200 rpm). After incubation, the absorbance of the bacterial suspension was compared with that of the McFarland standard. Nearly 100 μl of the adjusted bacterial suspension (10⁵ CFU ml^-1) were distributed on each 10 g of the washed chicken meat samples and allowed bacterial attachment for 30 min before subjecting them to treatment.

2.9 Treatment

The meat samples inoculated with bacterial culture were soaked in 200 ml of the MICs of nisin and allicin and their 2× FIC (0.016 and 0.008 mg ml^-1) for 1 h followed by DW as control. Use of 2× FIC was because antimicrobial agents were relatively less active in complex food matrices such as meat due to protein binding, fat content and limited diffusion. This is adopted from the method of Ajingi et al. [15]. After the removal of the meat sample from the solutions, this was drained for 1 h using laminar flow hood. The samples were transferred into sterile polyethylene bags after being treated in two sets. One set was stored at 4 °C using refrigerator and the other set was stored at 35 °C using incubator for 6 d. Samples were analyzed on Days 0, 2, 4 and 6.

2.10 Sampling and Analysis

From the two sets of incubated treated samples (10 g), 90 ml of peptone water were poured into each sample and vigorously smashed for 60 sec to suspend bacteria in the solution using laboratory stomacher. After smashing, the solution was diluted following a tenfold serial dilution in buffered peptone water. After serial dilution, 50 μl of the diluted solution were dispensed into sterile plates, followed by pouring a warm sterilized LB agar solution and incubating at 37 °C for 24 h. After incubation, colonies were counted and recorded in log CFU g^-1. In addition to common storage conditions, chicken samples were exposed to 35 °C to mimic severe temperature abuse scenarios that can occur during inappropriate handling, defective equipment and cold chain interruptions. While 35 °C is not appropriate for retail or household storage temperatures, increased temperature was similar to that of recent research, assessing microbial behavior and spoilage under abusive conditions [22].

2.11. Colorimetry Analysis

The sample colors were analyzed using ColorQuest colorimeter (Hunter Associates Laboratory, Inc.,USA). The color values of Hunter (L, a, b), showing lightness, redness and yellowness, were recorded at three sample sites each [15].

2.12 Data Analysis

GraphPad Prism statistical software v.10.4.1 (GraphPad Software, San Diego, CA, USA) was used to analyze data. Statistical differences between the inactivation and membrane permeability were compared using paired t-test. While differences between log decrease and calorimetric parameters were compared using one-way and two-way ANOVA, respectively. The differences were defined as statistically significant (95% confidence interval, p ≤ 0.05).

1. Results and Discussion

3.1 Assessment of Minimum Inhibitory Concentration and Minimum Bactericidal Concentration

In this study, a promising alternative for inhibiting the proliferation of Gram-negative bacterial strains, as well as combating the effect of Salmonella Typhimurium ATCC 13311 associated with chicken meat spoilage, was investigated from a globally accepted food preservative and organosulfur compound originating from a natural source. The mean values of the activity of bioactive agents against the Gram-negative bacterial strains are present in Table 1. The results show that nisin and allicin inhibited the proliferation of bacterial strains at various concentrations, with Salmonella Typhimurium ATCC 13311 was the most susceptible, followed by P. aeruginosa DMST15501 and E. coli ATCC 8239, with Vibrio spp. as the least susceptible. The minimum concentrations of nisin and allicin that killed the bacterial strains were higher than their respective MICs. However, Salmonella Typhimurium ATCC 13311 was killed at a lower concentration, compared to the other three bacterial strains. These results suggested that nisin could still inhibit and deactivate Gram-negative bacteria at higher concentrations than Gram-positive bacteria. These findings were supported by studies of Ajingi et al. [15] and Charest et al. [23]. However, the inhibitory and bactericidal effects of allicin in this study was similar to those in another study of Tao et al. [24].

3.2 Inhibitory Synergism of Nisin and Allicin

In this section, FICI was used to assess the synergism of nisin and allicin in all bacterial strains, as shown in Table 2. The FICI value against Salmonella Typhimurium ATCC 13311 was 0.5 when subjected to the combination of 1/4 fraction of nisin and allicin MIC values. In contrast, the concentration used against the other bacterial strains was investigated by how well each preservative acted against that strain. The result revealed that the combination was synergistically effective against all four bacterial strains at various FICs, with FICI less than or equal to 0.5. Moreover, Salmonella Typhimurium was shown as further susceptible to the combination (FIC), including FICI of 0.5.

Nisin, a globally recognized food additive with no allergic reactions or side effects when consumed by humans, includes limitations for activity on Gram-negative bacteria such as Salmonella Typhimurium because of the presence of an outer membrane that delays it from penetrating and attacking CM of the bacteria [25]. Allicin, an active organo-sulfur compound originating from garlic, was reported to improve antimicrobial agents, including peptides and common antibiotics, on Gram-positive and Gram-negative bacteria used for food preservation [26]. Despite the inhibitory potentials of allicin, it can be toxic to eukaryotic cells at a high concentration, which reacts with thiol groups [27]. The synergistic activity of nisin and allicin against Salmonella Typhimurium in this study was similar to that of reports by Ajingi et al. [15] and Duscaron et al. [17]. Additionally, Yang et al. [28] reported synergistic inhibitory effect of nisin and cellulose nanofibrils (CNFs) on various strains of Gram-positive and Gram-negative bacteria with Salmonella Typhimurium.

3.3 Kinetics of Inactivation

The pattern; by which, Salmonella Typhimurium ATCC 13311 inactivated over time when subjected to nisin and allicin at their MICs and FICs was investigated by the fluctuations in their OD at 600 nm for 5 h. The result present in Figure 1 shows that the MICs of nisin and allicin included a moderate inhibitory effect against Salmonella Typhimurium ATCC 13311 through 5 h with a slight increase in their OD, showing incomplete bacterial growth inhibition within the time frame. Moreover, 1× FIC and 2× FIC were more potent, compared to MICs of individual preservatives. Overall, these results suggested that the 2× FIC combination was associated to increased inhibitory activity against Salmonella with a significant difference, compared to other concentrations and the control (p = 0.008), as predicted by synergistic interactions based on the common definition of FICI (≤ 0.5). Patterns of suppression of such a behavior were similar to those of expectations for outcomes that were usually verified through time-kill kinetics, which is still the gold standard to assess antimicrobial interactions [29].

Regarding standardized Clinical Laboratory Standard Institute (CLSI) guidelines for MIC assessment [30], this integration of OD-based evidence established a strong theoretical basis for the efficacy of the nisin and allicin combination (1× and 2× FICs) and highlighted further potential synergistic therapy development. The synergistic inactivation pattern by 1× and 2× FICs in this study was supported by the report of Ajingi et al. [15], when nisin was synergized with organic acid on Gram-positive and Gram-negative bacterial strains.

3.4 β-Galactosidase Analysis

This study investigated the activity of nisin and allicin at their MIC and FIC levels on E. coli ATCC 8239 and linked Gram-negative bacteria such as Salmonella Typhimurium ATCC 13311 by changes in the OD of the treatment after the appearance of yellow coloration, indicating release of the beta-galactosidase enzyme capable of converting ONPG into ONP (Figures 2A and B). The results showed that nisin at its MIC did not cause release of beta-galactosidase, presenting almost a similar OD to that of the control. However, allicin has demonstrated appreciable effects on the bacterial membrane during 60 min after the color change and before the decrease in the OD level. Furthermore, 2× FIC and 1× FIC caused a significant release of beta-galactosidase with a relatively rapid increase in OD level. Additionally, the permeability effect of the FICs was different, with 2× FIC more effective than 1× FIC. Differences in the permeability of FICs (1× and 2× FICs), MICs (nisin and allicin) and control were statistically significant (p = 0.012). This verified synergism between allicin and nisin.

Ideally, nisin inhibits the Gram-positive bacterial cell wall biosynthesis via interruption of trans-glycosylation and mislocalization of lipid II, which is the building block of the bacterial cell wall, leading to the bacterial CM disruption through pore formation [31]. However, nisin alone at lower concentrations faces difficulties to compromise the outer membrane layer of Gram-negative bacteria, as lipopolysaccharide prevents it from contacting with the peptidoglycan layers. Therefore, synergizing nisin with allicin has facilitated its entry into the inner membrane of the bacteria through a gradual process. The combined action of the previously stated mechanisms for nisin and allicin is believed to make their synergy successful against Salmonella Typhimurium ATCC 13311 in this study.

Salmonella Typhimurium is the target pathogen, while E. coli is used as a proxy for β-galactosidase-based permeability assays due to their similarities as Gram-negative Enterobacteriaceae. The two include a double-membrane system with an outer membrane rich in lipopolysaccharides (LPS) [5]. Thus, changes affecting the outer membrane such as increased permeability or antimicrobial disruption similarly affect the two species. Future studies can incorporate Salmonella-specific reporter constructs or stress response markers (e.g., PhoP/PhoQ regulators, oxidative stress indicators and fluorescence-based viability reporters) to provide a pathogen-relevant assay similar to the β-galactosidase assay. These approaches allow mechanistic assessment directly in Salmonella without relying on lacZ-dependent systems.

3.5 Hemolysis Assay

To ascertain the safety of the nisin and allicin at the combined fractional inhibitory concentration used for chicken meat treatment, their hemolytic effect against RBC was studied. Results of the analysis demonstrated that nisin and allicin at their MICs coupled with their double FIC (2× FIC) included no hemolytic activity by showing a small OD at OD540 (similar to the negative control of 1× PBS) with no visible red coloration (Figure 3). However, the positive control (Triton X-100) included a high OD, suggesting that it damaged the RBCs by showing red color after centrifugation. Even though nisin and allicin are bioactive agents originating from natural sources, it was reported that high concentrations and exposure to allicin could begin cytotoxic and erythrocytic oxidative damages to the host cells [31]. Another study reported that allicin might cause dose-dependent hemolysis, supported by phosphatidylserine externalization, calcium influx and oxidative stress [32]. In contrast to allicin, nisin is generally regarded as safe, with limited or no reported cytotoxicity or hemolysis to mammalian cells and RBCs at an antibacterial concentration. Hence, this study has verified the safety of nisin and allicin at the concentration used for the chicken meat treatment; similar to the study of Sarkar and Jayanta [33].

3.6 Inhibitory Effects of Nisin and Allicin on Salmonella Typhimurium ATCC 13311 Contaminating Chicken Meats at 35 and 4 °C Storage Conditions

The synergistic inhibitory effects between the MIC of individual preservatives and their 2× FIC on chicken meats spoilage by Salmonella Typhimurium ATCC 13311 at abusive temperature (35 °C) and refrigerated condition (4 °C) is present in the logarithmic form of their mean values (Log CFU g^-1) and standard deviation (Table 3). Colonies of Salmonella Typhimurium ATCC 13311 after 30 min of treatment with nisin, allicin and their FIC combination are present in the table as Day 0. However, the individual preservatives significantly decreased the population of Salmonella Typhimurium ATCC 13311 on Days 2 (0.04 ±0.02), 3 (1.04 ±0.02) and 6 (1.07 ±0.03 and 1.05 ±0.03) with a p-value of 0.0193, compared to the control. Moreover, the 2× FIC caused a more significant difference in log reduction (p = 0.001) of Salmonella Typhimurium ATCC 13311 on the treated meat samples, compared to the control, showing the highest decrease on Day 2 (2.23 ±0.2), 4 (3.21 ±0.1) and 6 (3.15 ±0.03), respectively. Figure 4A illustrates the appearance of the meat sample after treatment with DW and 2× FIC, stored at 35 and 4°C, respectively.

However, allicin showed a significant difference in log decrease of Salmonella Typhimurium ATCC 13311 under refrigerated storage conditions (4 °C) (p = 0.0144). In contrast, no significant difference was seen in the decrease of Salmonella Typhimurium by the MIC of nisin, compared to the control (p = 0.1064) under similar conditions. However, the 2× FIC established a certain preservative effect against Salmonella Typhimurium ATCC 13311 better than that of individual MICs and control, with a significant logarithmic decrease on Days 2 (2.23 ±0.03), 4 (1.24 ±0.04) and 6 (1.18 ±0.04), respectively (p = 0.001).

Despite the ability of Salmonella Typhimurium to adhere to food substances, mostly meats such as chicken and other poultry meats, this study revealed a promising achievement in the decrease of the bacterial growth in fresh chicken meats by the combination of nisin and allicin at their 2× FIC level after 6 d of storage conditions. This effect could be associated with the inhibitory effect of allicin on biofilm formation by Salmonella Typhimurium, which is the major mechanism of the bacteria used to firmly adhere to the meat surface [34]. Mechanisms exerted by allicin on Salmonella Typhimurium are further effective when synergized with other antimicrobials; as reported by Choo et al. [26]. This verified synergism of nisin and allicin against Salmonella Typhimurium ATCC 13311 in the fresh chicken meat samples. The multiplication of Salmonella Typhimurium ATCC 13311 in this study was higher at an abusive temperature (35 °C) than refrigerated conditions (4 °C). The low-temperature effect was possibly further justified because such a low temperature significantly decreased bacterial metabolic processes, which was a factor in minimizing the growth and activity of Salmonella Typhimurium ATCC 13311 during incubation [35]. Additionally, activity of nisin depended on the temperature condition and was reported more active at an increased temperature due to the active state of the target organism, which enhanced the interactions of the nisin and the organism membrane components. However, the log decrease observed after treatment could be associated to allicin augmentation. Although the present study demonstrated antimicrobial effectiveness over the storage, concentrations of residual allicin were not quantified. Further studies incorporating chromatographic techniques such as high-performance liquid chromatography (HPLC) can allow direct assessment of compound stability and degradation kinetics, strengthening long-term efficacy claims.

The synergistic antibacterial effect of nisin and allicin in this study was supported by the findings of Ajingi et al. [15], who reported that nisin in combination with organic acid demonstrated effective preservative potential against Bacillus subtilis in chicken meats and potatoes. Additionally, a study by Singh et al. [36] reported a synergistic effect of nisin and aqueous garlic extract against four strains of L. monocytogenes in a food system stored at 4 °C, similar to the findings of this study. Moreover, the garlic allyl isothiocyanate bioactive compound was reported to enhance the antibacterial activity of nisin against Gram-negative foodborne pathogens, including Salmonella Typhimurium [37]. This verified ability of garlic biocompounds in augmenting other antimicrobials against various species of foodborne pathogens. One limitation of this study included use of 35 °C, as the sample temperature was inconsistent with that commonly used to avoid extreme temperature abuse in chicken meats, which is typically set below 5 °C.

3.7 Calorimetric Analysis

The microbial spoilage of food generally produces undesirable changes in the actual color of the food sample and can be attributed to the secretion of metabolic byproducts, most especially at the stationary phase of bacterial growth within the food sample. In this study, the calorimetric analysis was carried out on the beginning day of the treatment (Day 0) and the final day (Day 6) after storage at 35 and 4 °C, respectively. The results revealed that the chicken meat treated with MICs of nisin and allicin under an abusive temperature demonstrated an increase in the degrees of lightness (L*), redness (a*) and yellowness (b*) over 6 d, with no significant difference with the control treatment (p = 0.2424, 0.2413 and 0.9417, respectively). This suggested higher pigment oxidation and spoilage discoloration at abusive temperatures. The combination of nisin and allicin (2× FIC) helped preserve the natural color of the treated chicken meat samples, leading to a slight increase in lightness (L*) and redness (a*) with significant differences with red color of the control meat sample after 6 d (p = 0.0001). However, the 2× FIC treatment showed no significant difference in the degree of yellowness (b*) with that of the control and the two MIC treatments on the final day of treatment (p = 0.7816).

At the 4 °C refrigeration temperature, the findings for lightness (L*) and redness (a*) were similar to those observed at the higher temperature. Result showed that the lightness of meat did not change significantly between the MIC of nisin and allicin treatments and the control treatments at Day 6 (p = 0.5114). In contrast, the degree of lightness (L*) for their combination (2× FIC) showed a significant difference between the control and MIC treatments (p = 0.001). Similarly, the degree of redness in chicken meat was not significantly affected using either MIC allicin, MIC nisin or the 2× FIC combination at Day 0 (p = 0.168), indicating that these treatments did not interfere with the characteristic red color of the meat during the beginning refrigerated storage (Day 0). However, results demonstrated a significant difference between the combination treatment (2× FIC) and the control at Day 6 (p = 0.0001).

Moreover, MIC treatments showed no significant difference in the yellowness (b*) of the treated meat samples at Days 0 and 6, with a high p-value of 0.114. Meat samples treated with 2× FIC were significantly different from those treated with control (DW) after 6 d of storage (p = 0.0082). This was similar to a study by Charest et al. [23], who stated that nisin contributed to the delay of biochemical processes used by bacterial cells to deteriorate the color of meat samples. Additionally, stability in the color of chicken meat samples stored at 4 °C was linked to the synergism of the combination of nisin and allicin (2× FIC) at chilled conditions; supported by a study of Dacheng Pharma [38]. Results of the calorimetry analysis in this study was correlated to a study by Ajingi et al. (15). Physical appearance of the treated meat samples is provided in Figure 4A. Regarding acceptability and safety of the highlighted synergistic antimicrobials, the current study provided a safer further cost-effective substitute for controlling chicken meats and associated food spoilage caused by Salmonella Typhimurium.

1. Conclusion

It was elaborated that nisin included an inhibitory effect against Gram-negative bacteria, when synergized with allicin at lower concentrations. It was also observed that the combination included preservative potentials against Salmonella Typhimurium ATCC 13311, responsible for chicken meat spoilage within 6 d at various temperature conditions. The study indicated that nisin and allicin could synergistically disrupt the CM of Gram-negative bacteria. Furthermore, it was revealed that nisin and allicin included no hemolytic effect on RBCs at their MICs and 2× FIC. The study indicated that the combination of nisin and allicin (2× FIC) used in the treatment of chicken meat samples included no detrimental effect on the color of the treated meats. However, studies are needed to unveil other possible mechanisms investigated by the combination against Salmonella Typhimurium.

Background and Objective: Potato pulp, a major agro-industrial by-product, contains valuable nutrients, which are still underused due to their limited bioavailability and functional characteristics. This study aimed to enhance the nutritional and functional qualities of potato pulp through solid-state fermentation (SSF) using Lactobacillus fermentum MT-ZH893 (LF) in combination with Aspergillus oryzae (AOR) and Rhizopus oryzae (ROR).

Material and Methods: Potato pulp was utilized as the substrate for solid-state fermentation. Four strategies combining Lactobacillus fermentum with fungi were investigated, including: simultaneous fermentation (LF+AOR and LF+ROR) and two-stage fermentation (fungi inoculated first, followed by the bacteria of; AORLF and RORLF). Key parameters assessed included crude protein content, probiotic viability, total phenolic content, antioxidant capacity (2,2′-azinobis-(3-ethylbenzothiazoline-6-sulphonate and 2,2-diphenyl-1-picrylhydrazyl assays), organic matter digestibility and metabolizable energy.

Results and Conclusion: Simultaneous fermentation with Aspergillus oryzae increased crude protein content, whereas treatments involving Rhizopus oryzae resulted in a significant decrease in crude protein. Simultaneous fermentation significantly enhanced probiotic viability. In contrast, the two-stage process using Rhizopus oryzae produced superior outcomes in total phenolic content (1011 mg Gallic acid equivalents per gram of dry material, antioxidant capacity (2,2′-azinobis-(3-ethylbenzothiazoline-6-sulphonate, 934.9 μmol Trolox; 2,2-diphenyl-1-picrylhydrazyl, 97.01%), organic matter digestibility and metabolizable energy. These results indicated that while simultaneous fermentation optimized probiotic viability and protein content (with Aspergillus oryzae), the two-stage approach with Rhizopus oryzae significantly enhanced the functional characteristics and digestibility of potato pulp, presenting promising strategies for value-added bioprocessing of agro-industrial residues.

Keywords: Agroindustrial byproducts, Aspergillus oryzae, Lactobacillus fermentum, Potato pulp, Rhizopus oryzae, Solid-state fermentation

1. Introduction

Potatoes rank as the world’s fourth most significant food crop, contributing profoundly to global food security [1]. With their extensive consumption and industrial processing, vast quantities of wastes, particularly potato peels and pulp, are generated, which are projected to reach over 8 million tons by 2030, emitting nearly 5 million tons of CO₂ equivalents [2]. While such wastes pose serious environmental challenges, these represent a valuable reservoir of bioactive compounds and fermentable substrates with great potential for sustainable bio-valorization [3]. Potato pulp, achieved at roughly 0.75 tons per ton of extracted starch [4], is a lignocellulose-rich byproducts consisting of water, starch, cellulose, hemicellulose, pectin, protein, fibrous carbohydrates, phenolic compounds and minerals. Despite its valuable composition, economic and technical barriers linked to preservation and transport currently limit its industrial use [5].

To overcome these limitations and safely use these agro‑industrial residues, various strategies have been suggested. From these suggestions, microbial fermentation has significantly been interested as a highly effective approach for increasing the nutritional values and functional characteristics of such byproducts. By converting complex macromolecules into simpler further digestible components, fermentation not only improves the overall nutritional quality of the residue but also facilitates its stabilization and further use.

Therefore, solid-state fermentation (SSF) has emerged as a compelling solution to increase the added value of potato pulp. Rather than fundamentally solving economic transport barriers, SSF represents an economically viable and environmentally friendly bioprocess characterized by low water and energy consumption. The efficiency of SSF is ruled by several key variables, including moisture content, temperature, fermentation time and inoculation strategy. Use of probiotics such as Lactobacillus fermentum with filamentous fungi such as Aspergillus oryzae and Rhizopus oryzae in SSF systems is particularly advantageous. The L. fermentum shows immunomodulatory, anti-inflammatory and microbiota‑balancing effects. Complementarily, these fungal species secrete hydrolytic enzymes that facilitate starch and protein degradation, improving substrate digestibility and supporting probiotic viability [6]. This integration promotes circular bioeconomy principles by transforming agri‑food wastes into high‑value health‑promoting products [7].

Furthermore, numerous studies have directly highlighted the efficacy of SSF in valorizing potato byproducts. For example, potato pulp has been used as a substrate for enzyme production using recombinant A. oryzae strain [8]. Moreover, SSF of potato peels uses R. oryzae enriched fungal biomass with proteins and essential amino acids (EAA) [9]. A two‑stage fermentation converted potato starch wastes into high-quality animal feed using mutant A. niger followed by Bacillus licheniformis [10]. More recently, enzyme-assisted fermentation of potato pulp promoted the growth of L. plantarum, increasing organic acid production and altering microstructural characteristics to facilitate rapid drying [11]. Co‑fermentation of Aspergillus spp. with B. subtilis improved physicochemical characteristics and decreased spoilage of potato pulp [12].

Regarding sequential strategies aimed at probiotic enrichment, fungal pretreatment with A. oryzae enhanced the viability of Pediococcus acidilactici by degrading the rigid lignocellulosic structure of the substrate [13]. Similarly, SSF of industrial potato wastes with L. plantarum, S. cerevisiae and A. oryzae increased phenolic compounds, antioxidant capacity and simultaneously decreased anti-nutritional factors such as phytic acid [14]. Despite these advancements, studies specifically assessing the SSF of potato pulp that use simultaneous or sequential cultures of L. fermentum with A. oryzae or R. oryzae are still limited.

Accordingly, the overall strategy of this study was to valorize potato pulp through solid-state fermentation by comparing simultaneous and two-stage fermentation systems of L. fermentum with either A. oryzae or R. oryzae. By focusing on interspecies metabolic interactions, the study aimed to systematically investigate the effects of various inoculation strategies on nutritional quality, antioxidant enrichment, probiotic stability and substrate digestibility. Ultimately, this study was carried out with the goal of developing a value-added bio-based feed additive from this starchy agro-industrial waste; thereby, identifying the optimal approach for its microbial upcycling.

1. Materials and Methods

2.1. Materials

2.1.2. Potato pulp and microorganisms

Potato pulp was sourced from Alvand Starch, Iran, dried at 50 °C for 48 h and stored at -18 °C. The L. fermentum MT.ZH893 was isolated from Mazandaran local cheese at the Food Science and Engineering Department, Tarbiat Modares University, Tehran, Iran [15]. The A. oryzae PTCC 5163 was provided from the microbial collection of the Faculty of Veterinary Medicine, University of Tehran, Tehran, Iran, and R. oryzae PTCC 5176 was provided by Persian Type Culture Collection (PTCC), Iranian Research Organization for Science and Technology (IROST), Tehran, Iran.

2.1.3. Materials and Reagents

The chemicals and culture media used in this study were as follows. Briefly, de Man, Rogosa and Sharpe (MRS) agar media were purchased from Ibresco, Iran; methanol (CH₃OH) was purchased from Majalli Industrial Chemicals, Iran; 2,2-diphenyl-1-picrylhydrazyl radical [2,2-diphenyl-1-picrylhydrazyl (DPPH⁺) ≥ 98% purity], 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt [2,2′-azinobis-(3-ethylbenzothiazoline-6-sulphonate (ABTS⁺) ≥ 98% purity] and Folin-Ciocalteu phenol reagent were purchased from Sigma-Aldrich, USA. Trichloroacetic acid (TCA) and sulfuric acid were purchased from Merck, Germany. All reagents included analytical grade or higher and were used without further purification unless otherwise stated.

2.2. Methods

2.2. Fermentation process

In the two-stage fermentation, 20 g of dried potato pulp were transferred into a 250-ml Erlenmeyer flask, brought to 60 % humidity and sterilized at 121^∘ C for 15 min using autoclave. Then, 1 ml of fungi [A. oryzae or R. oryzae (10⁸ cfu ml^-1)] was added to the media and incubated at 30 ^∘ C for 24 h. The mixture was autolyzed at 55 ^∘C for 24 h. After cooling down, the probiotic L. fermentum was added to the potato pulp at a concentration of 1.5 × 10⁸ cfu ml^-1 and incubated at 37 ^∘C for 72 h. In the simultaneous fermentation, fungi and L. fermentum were inoculated in a single step. The fermentation condition was microaerophilic [16,17]

2.3. Viability analysis

After the fermentation period, 1 g of the fermented substrate was weighed using conical tube (15 ml) with physiological serum under sterile conditions and diluted serially after complete homogenization. Then, 0.1 ml of each dilution was surface-cultured on MRS agar media, which was incubated at 37 °C for 72 h. With known percentage of moisture in the fermented substrate, the colony count was calculated based on dry matter (g) of the substrate as cfu per gdry weight substrate (cfu gdws^-1) [18]

2.4. Total protein and total nitrogen

The total nitrogen content of digested samples was assessed using micro-Kjeldahl apparatus. For the analysis, 0.2 g of dried sample and 1 g of Kjeldahl digestion catalyst (sodium sulfate with copper sulfate and/or selenium) were transferred into the designated digestion tube of the system. Then, 7.5 ml of concentrated sulfuric acid (H₂SO₄, 98%) were added and digestion was carried out at 360°C for 5–6 h. The total nitrogen value was multiplied by 6.25 to calculate the total protein percentage [19].

2.5. True protein

To assess true protein content, the fresh samples were first dried at 50 °C for 24 h until constant weight using oven. The dried samples were thoroughly mixed and homogenized to ensure uniformity; then, 0.5 g was accurately weighed using laboratory balance. The weighed sample was transferred into a standard laboratory glass beaker and mixed with 50 ml of distilled water (DW). The mixture was set at room temperature (RT) for 30 min. Then, 10 ml of 10% trichloroacetic acid (TCA) were added and the sample was set standing for 30 min at RT. The precipitated proteins were filtered through Whatman filter paper no. 1 and the residue was dried using oven. The nitrogen content of the dried residue was assessed using micro-Kjeldahl method and the non-protein nitrogen (NPN) content was calculated as difference between the total nitrogen and the protein nitrogen from the filtrate [20].

2.6. Assessment of gas production, organic matter digestibility and metabolizable energy

Gas production resulting from fermentation was assessed based on a method of Menke et al. (1979). First, samples were ground using a sieve with a 1 mm pore size. A 200 mg portion of the dried sample (60 °C) was transferred into each calibrated glass syringe. After sample loading, the plunger walls were lubricated with petroleum jelly and the syringes were pre-incubated at 39 °C to equilibrate with the temperature of the rumen fluid–artificial saliva mixture. Blank syringes were included at the beginning, midpoint and end of each incubation series. Additionally, a standard feed substrate with a known gas production value (e.g., standard hay or concentrate) was used for calibration purposes.

Rumen fluid was collected from at least two rumen-fistulated wethers and strained through layers of cheesecloth to remove large particles. The collected fluid, containing solid and liquid fractions, was homogenized thoroughly. Donor animals were maintained on a diet containing 30–50% concentrate to ensure forage quality. Artificial saliva was prepared by mixing macro-mineral, micro-mineral, buffer, resazurin and reducing solutions. The DW, buffer and mineral solutions were mixed together using flask, incubated at 39 °C using water bath and supplemented with resazurin to produce a blue solution. This mixture was stirred continuously while a gentle stream of CO₂ was passed through. A reducing solution was added to the mixture until the color changed from blue (oxidized) to purple and then cleared (reduced). The CO₂ flushing continued for nearly 10 min (occasionally up to 15–20 min) to achieve full reduction. Rumen fluid was added to the artificial saliva at a 2:1 ratio, followed by 10 min of CO₂ flushing. The CO₂ inlet was positioned above the fluid surface and flow was adjusted to maintain an anaerobic atmosphere.

Thirty milliliters of the inoculum mixture were injected into each syringe containing the feed substrate. Excess air was expelled by holding the syringe vertically, opening the clamp and adjusting the plunger. Syringes were incubated at 39 °C and cumulative gas volumes were recorded regularly for up to 120 h, until gas production plateaued (e.g., equal readings at three consecutive time points). Organic matter digestibility (OMD) was calculated from the 24 h gas volume (GP₍₂₄₎) from fermentation of 200 mg DM using Eq. 1:

                  (1)       

Where, GP₍₂₄₎ was gas volume after 24-h incubation (ml/200 mg DM), CP was crude protein (g kgDM^-1) and XA was ash content (g kgDM^-1). Metabolizable energy (ME) was calculated based on Eq. 2:

                       (2)

Where, ME was Metabolizable energy, expressed in megajoules per kilogram of dry matter; GP₍₂₄₎ was gas volume produced after 24 h; CP was crude protein concentration, expressed in grams per kilogram of dry matter; and EE was ether extract, expressed in grams per kilogram of dry matter [21].

2.7. Total phenolic content

For the quantification of total phenolic content (TPC) in the samples, Folin-Ciocalteu method was used. In this procedure, the fermented sample extracts were achieved by centrifuging the samples at 26,832 g for 15 min using refrigerated centrifuge (Kubota 6900, Japan). A volume of 200 µl of the prepared extracts was mixed with 500 µl of Folin’s reagent (diluted 10-fold with DW) in 2-ml vials, which were set in dark for 5 min. Then, 1 ml of 5.7% sodium carbonate (Na₂CO₃) solution was added to each vial. After vortexing, the samples were set in dark for 1 h. The control sample was prepared similarly with the exception that DW replaced the sample. After the incubation period, absorbance of the samples was measured at 765 nm using spectrophotometer. Gallic acid standard curve was plotted with concentrations ranging 0–250 ppm and the phenolic content was expressed as milligrams of Gallic acid equivalents per gram of dry matter (mg GAE gDM^-1) [22].

2.8. The DPPH radical scavenging assay

The antioxidant activity of the samples was investigated using DPPH free radical scavenging method. Briefly, 0.1-mM DPPH solution in methanol was prepared and thoroughly vortexed. The extracts of the fermented samples were achieved by centrifugation at 4000 rpm for 15 min. Then, 500 µl of the sample extract and 500 µl of the DPPH solution were mixed and vortexed using 2-ml vials. Furthermore, a control sample consisting of 500 µl of methanol and 500 µl of DPPH solution was prepared. The vials were set in dark for 30 min; then, their absorbance was measured at 517 nm using spectrophotometer. The radical scavenging activity (RSA) was calculated using Eq. 3 [23].

Radical scavenging activity (%) =                        (3)

Where, A_control was absorbance of the control and A_sample was absorbance of the sample.

2.9. The ABTS radical scavenging assay

The ABTS cation radical scavenging activity was calculated based on the reduction in the green-blue color of this radical by antioxidant compounds. To prepare the reagent solution, 7 mM ABTS solution and 2.45 mM potassium persulfate were mixed in a 1:1 ratio and set in dark for 12–16 h. Then, 600 μl of the resulting solution were diluted with 40 ml of DW to adjust the absorbance to approximately 0.7 at 734 nm. For the assay, 20 μl of the sample extract were mixed with 980 μl of the ABTS solution and set in dark for 10 min; the control was prepared by replacing the extract with DW. Absorbance of the samples was measured at 734 nm and ABTS radical scavenging ability of the samples was calculated in micromoles based on the Trolox standard curve [24].

2.10. Scanning electron microscopy

The fermented potato pulp samples were frozen at -74 °C for 24 h, followed by freeze-drying using laboratory freeze dryer. The dried samples were ground and sieved through a 40‑mesh sieve (0.45-mm aperture). Then, the powdered samples were coated with a thin layer of gold and studied using scanning electron microscope (SEM) operated at an accelerating voltage of 5.0 kV [11].

2.10. Statistical analysis

Data were analyzed using one-way ANOVA. To analyze the significance of the means, JMP Pro 18 software (SAS, USA) was used, providing robust statistical tools to handle the various data sets generated throughout the study. The Tukey test was used to compare the means and to assess the statistical significance of differences in all stages. All graphs were generated using Excel 365 v.2508 (Microsoft, USA). All assessments were carried out in triplicate.

1. Results and Discussion

3.1. Viability of Lactobacillus fermentum

As shown in Figure 1, fungal pretreatment, whether through simultaneous inoculation or two-stage fermentation, generally enhanced the survival of L. fermentum. In the simultaneous inoculation method, the fungus and probiotic bacteria were introduced into the culture substrate at the same time, initiating a multi-faceted synergistic interaction. This interaction is environmental and nutritional as A. oryzae consumed oxygen during its initial growth phase, creating microaerophilic conditions that stimulated growth of the facultative anaerobic L. fermentum. Simultaneously, the fungus secreted a range of hydrolytic enzymes such as cellulase, amylase and protease, which broke down complex substrates into simpler further absorbable monomers for the bacteria [25]. The ability of A. oryzae to produce cellulase is particularly effective for the hydrolysis of lignocellulosic wastes [26]. These effects provided conditions that significantly increased the survival of L. fermentum.

In the two-stage fermentation method, the fungus was first grown alone, allowing it sufficient time to produce and secrete a maximum level of macromolecule-degrading enzymes; hence, creating favorable conditions for bacterial growth and survival [27]. This fungal pretreatment improved nutrient availability by enzymatically degrading macromolecules while preventing the excessive consumption of fermentable sugars; thereby, minimizing the production of inhibitory compounds such as furfural [28]. Furthermore, as the fungus grew, its mycelia penetrated the substrate, which decreased particle size and created a further suitable physical environment for the bacteria. Then, fungal autolysis could release intracellular nutrients back into the media [17]. These processes collectively improved the growth substrate for the bacteria; however, the accumulation of fungal metabolites and the consequent pH decrease following pretreatment might shift the environment away from the optimal range for L. fermentum. This acidic stress could be detrimental, as it might force bacteria into a lag phase or impair growth by decreasing the uptake of essential elements such as iron and phosphorus, even if food sources were available [27]. Overall, the two methods, through the degradation of lignocellulosic compounds and the production of beneficial metabolites, enhanced probiotic survival. Nevertheless, the simultaneous inoculation method in this study demonstrated a statistically significant improvement in the survival of the probiotic bacteria.

3.2. Crude protein and true protein

Based on the results presented in Figure 2, the LF+AOR treatment showed the highest crude protein content (15.91%) in all treatments (p ≤ 0.05). The LF and AOR LF treatments contained 15.4 and 14.5% of crude protein, respectively. The control treatment (PP) included 13.96% of crude protein, while the lowest value (12.5%) was observed in the LF+ROR treatment (p ≤ 0.05). The fermentation process could cause significant changes in the protein content of potato pulp. Based on the results, simultaneous fermentation, especially with A. oryzae, led to a significant increase in crude protein, compared to the control and other treatments (p ≤ 0.05). The increase in protein during fermentation could be attributed to various factors such as the accumulation of fungal and bacterial biomass on the fermented potato pulp substrate [29,30] or secretion of extracellular enzymes by microorganisms during cellular metabolism, which were proteinaceous in nature. An increase in protein content played a critical role in improving the nutritional value of animal feed. The results indicated that solid-state and submerged fermentation of fruit and vegetable wastes could produce fermented products with high protein contents. Fermentation by fungi, whether in solid or liquid form, represented a successful strategy for valorizing waste materials to produce valuable and sustainable alternative protein sources for animal feed formulations [31]

The increase in protein content during fermentation could be attributed to several biological and biochemical mechanisms. These included the production of proteolytic enzymes by microorganisms during cellular metabolism, which facilitated breakdown of complex proteins, bioconversion of carbohydrates and lipids into proteinaceous compounds, and growth and proliferation of fungal mycelia that contributed to the overall protein biomass. Additionally, the decrease in pH during fermentation helped prevent nitrogen loss, while the conversion of non-protein nitrogen into microbial protein further enhanced total protein levels. Other contributing factors included the accumulation of fungal and bacterial biomass on the potato pulp substrate after fermentation [29,31,32]. Another study assessed the effect of fermentation on the composition of Irish and sweet potato peels, reporting increases in moisture, ash, fat and proteins with decreases in crude fibers and carbohydrate contents [34].

Based on the results in Figure 2, the highest true protein content (11.87%) was observed in the control treatment (PP), whereas the lowest value (5.97%) was recorded for the LF+ROR treatment (p ≤ 0.05). The LF and AOR LF treatments contained 8.36 and 8.72% true protein, respectively, the two showing significant differences compared to the control group (p ≤ 0.05). The LF+AOR treatment demonstrated 9.15% of true protein, while the two-stage ROR LF treatment showed 7.76%. In this experiment, fermentation particularly in combined treatments, resulted in a significant decrease in true protein content, compared to the control (p ≤ 0.05). The discrepancy between crude protein and true protein suggested that protein structure was modified during fermentation, potentially enhancing amino acid availability, but not all the nitrogen was retained in a form measurable as true protein. Previous studies have reported variable effects of fermentation on true protein content, depending on microbial strains and process conditions. For example, in the solid-state fermentation of sweet potato industry residues, the highest true protein yield was achieved by co-culturing of A. oryzae with B. subtilis that led to an overall increase in protein components after fermentation [35].

Similarly, Borras et al. (2020) detected that true protein in fermented potato wastes increased by 2.7 U within 24 h at 25 °C [36]. In contrast, Djukic-Vukovic et al. (2015) reported that in lactic acid and animal feed production from distillery stillage using L. rhamnosus, protein contents decreased after fermentation while total nitrogen increased, compared to unfermented samples. Similarly, Dadkhodazadeh et al. (2024) observed that solid-state fermentation of fish wastes with probiotic strains led to an increase in non-protein nitrogen accompanied by a decrease in true protein [37,38]. These findings collectively indicated that under certain conditions such as high proteolytic activity or imbalance between protein degradation and microbial protein synthesis, fermentation might decrease measurable true protein despite increasing crude protein, highlighting the need to optimize process parameters to maximize nitrogen retention in its most digestible form.

3.3. Organic matter digestibility and metabolizable energy

Based on the results in Figures 3 and 4, the effects of treatment type on organic matter digestibility (OMD) and metabolizable energy (ME) of potato pulp were clearly evident. Two-stage treatments involving fungal fermentation followed by L. fermentum (AOR LF and ROR LF) demonstrated the highest OMD and ME values, which were significantly different from other treatments (p < 0.05). The observed increase in two-stage treatments was likely attributed to the effective degradation of cell wall structures by fungal enzymes (e.g., cellulase and hemicellulase) during the first stage, followed by the production of beneficial metabolites (organic acids and proteolytic enzymes) by L. fermentum in the second stage, resulting in enhanced nutrient release and digestibility. In contrast, simultaneous co-cultures (LF+ROR and LF+AOR) showed lower performance, possibly due to microbial competition and limited specific enzymatic activities, compared to two-stage treatments. Single-stage fermentation with L. fermentum (LF) induced a moderate improvement in OMD and ME, compared to the control (PP). These findings emphasized that the two-stage fungus-bacteria fermentation strategy was more effective in enhancing digestibility and energy value than that simultaneous and single-stage fermentations were, which could be explained by synergistic enzymatic and biometabolic mechanisms.

The results demonstrated that two-stage fermentation, using R. oryzae and A. oryzae in the initial stage followed by L. fermentum in the subsequent stage, facilitated structural cell wall component breakdown and nutrient release through synergistic microbial activities, significantly improving the digestibility and nutritional value of plant-based substrates [38–40]. Fungi secrete enzymes such as cellulose, xylanase [41,42] and proteases for protein hydrolysis [43,44], effectively pretreating the substrate. In the second stage, L. fermentum produced lactic acid and organic acids that decreased pH and facilitated complex molecule degradation, improving flavor profile and nutritional indices [45,46]. The findings of this study, which included an increase in OMD and ME after fermenting the samples, were similar to those of previous studies [47–50]. These results indicated that microbial fermentation, improved organic matter digestibility and feed energy use through the degradation of lignocellulosic structures.

3.4. Total phenols and antioxidant capacity

In this study, it was observed that fermentation with L. fermentum increased TPC (Figure 5) and antioxidant capacity (Figures 6 and 7). The R. oryzae facilitated the enzymatic degradation of cell wall components during solid-state fermentation, releasing bound phenolic compounds. When coupled with L. fermentum, the liberated phenolics could be involved in further biotransformation, leading to increased antioxidant activity. This synergistic interaction was especially effective in two-stage fermentations, where initial fungal pre-treatment maximized phenolic release, followed by bacterial fermentation that stabilized and enhanced the antioxidant potential. Such improvements have been verified in various fermented food substrates such as wheat and corn kernels, with significant increases in TPC and antioxidant activity measured through DPPH, ABTS and FRAP assessments [51,52].

The combined fermentation of food matrices with R. oryzae or A. oryzae with L. fermentum significantly enhanced total phenolics, flavonoids and antioxidant indices. These effects were attributed to the enzymatic activity of R. oryzae (e.g., β-glucosidase and polysaccharidases) and the strong cell wall-degrading ability of A. oryzae, which promoted the release and stabilization of antioxidant compounds [53]. Similar to [54] who reported an increase in TPC and antioxidant activity of fermented malt cake as a result of solid-state fermentation with fungal strains, the present study demonstrated that fermentation of potato pulp samples led to the release of bound phenolic compounds and improvement in free radical scavenging capacity. This effect was likely attributed to the production of hydrolytic enzymes by the microorganisms and the cleavage of bonds between the phenolic compounds and the cell matrix during the fermentation process. Overall, the integration of these microbes in fermentation processes not only boosted antioxidant characteristics but also ensured greater bioactive functionality, as consistently demonstrated through DPPH and ABTS assessments.

3.5. Scanning electron microscopy analysis

The SEM analysis provided mechanistic insights into the structural degradation of potato pulp during fermentation (Figure 8). The unfermented control (Figure 8A; PP) showed intact, smooth starch granules encased within a cohesive matrix, representing a significant physical barrier to nutrient bioavailability, whereas monoculture fermentation with L. fermentum (Figure 8B; LF) initiated structural erosion via surface pitting and micro-cracks, indicating early-stage enzymatic and acid-driven disruption. Furthermore, simultaneous co-fermentation strategies (Figure 8C, LF+ROR; and Figure 8D, LF+AOR) revealed synergistic breakdown with profound starch granule collapse, extensive cell wall degradation and network-like microbial clusters, facilitating highly efficient substrate depolymerization. Sequential two-stage fermentations (Figure 8E, ROR-LF; and Figure 8F, AOR-LF) demonstrated the most severe microstructural dismantling, where the original matrix was completely replaced by fragmented debris, proving that fungal pretreatment successfully maximizes subsequent bacterial penetration. These microstructural shifts verified that complex fungal-bacterial fermentation strategies fundamentally dismantled the inherent physical barriers of potato pulp, significantly amplifying substrate accessibility and its ultimate nutritional value.

1. Conclusion

This study highlighted the novelty and efficiency of integrating L. fermentum with filamentous fungi (A. oryzae and R. oryzae) through solid‑state fermentation to valorize low‑value potato pulp into a functional livestock feed. Comparative analysis of inoculation strategies demonstrated that while simultaneous fermentation maximized probiotic viability, the two‑stage process (fungal pretreatment followed by bacterial fermentation) achieved the most comprehensive nutritional enhancement. This approach effectively used fungal enzymatic activity to degrade complex fibers, improving digestibility, energy value and releasing bound phenolics that enhanced antioxidant capacity. Overall, the synergistic and eco‑friendly bioprocessing model provided a sustainable framework for converting agro‑industrial residues into functionally enriched feed ingredients, supporting waste decrease and the circular bio‑economy in animal nutrition.

Background and Objective: Hyaluronic acid is a high‑value biopolymer widely used in the food, cosmetic, therapeutic and pharmaceutical industries. This study aimed to establish a biosafe “Generally Recognized as Safe” platform for hyaluronic acid biosynthesis via engineering Bacillus subtilis 168.

Material and Methods: The hyaluronan synthase gene from Streptococcus dysgalactiae and tuaD homolog NX02_04625 from Sphingomonas sanxanigenens were selected based on their reported catalytic performance and codon‑optimized for Bacillus subtilis. The complete pHT01 vector construct harboring the two genes was commercially synthesized and sequence‑verified. A single‑gene plasmid was generated and the two plasmids were introduced into Bacillus subtilis 168 via electroporation. Hyaluronic acid production was quantified using cetyltrimethylammonium bromide assay, structurally validated by Fourier transform infrared spectroscopy and characterized by size‑exclusion chromatography.

Results and Conclusion: The dual-gene strain produced 1.20 ± 0.03 gl^-¹ hyaluronic acid, achieving a 2.4‑fold increase relative to the single‑gene strain and showed a trimodal molecular‑weight profile dominated by high‑molecular-weight polymers (> 1.18 MDa, 43.45%), accompanied by medium‑range species (~0.99 MDa, 24.69%) and low‑molecular-weight-oligomers (~55 kDa, 31.86%). Collectively, these findings highlight the efficiency and biosafety of the engineered production platform and its potential for food‑grade uses, supported by the predominance of high‑molecular-weight hyaluronic acid with favorable gel‑like characteristics.

Keywords: Bacillus subtilis 168, Hyaluronan synthase gene, Hyaluronic acid production, Metabolic engineering, Strain development, UDP‑glucose dehydrogenase

1. Introduction

Hyaluronic acid (HA) is a high-value, linear glycosaminoglycan renowned for its exceptional biocompatibility, viscoelasticity and hygroscopic characteristics [1]. As established in recent literature [2], HA is structurally composed of repeating disaccharide units of D-glucuronic acid (GlcUA) and N-acetyl-D-glucosamine (GlcNAc). The HA is a ubiquitous component of the extracellular matrix in vertebrates; where, it regulates cell proliferation, tissue hydration, inflammation and wound repair [3]. Due to its shear-thinning behavior and high moisture retention, HA has become essential in biomedicine (e.g. dermal fillers and drug delivery), cosmetics (e.g. anti-aging formulations) and regenerative medicine [4, 5].

The unique physicochemical characteristics of HA, including its molecular weight-dependent biological functions, support its clinical versatility [1]. High-molecular-weight (HMW) HA is generally associated with tissue homeostasis and restraint of cell proliferation/migration, whereas HA fragmentation into low-molecular-weight (LMW) species is linked to inflammatory and promigratory signaling [6]. The LMW-HA can enhance endothelial angiogenic responses and migration in vitro [7], supporting the use of tailored HA preparations for various therapeutic uses.

 In comparative studies using mouse models of persistent inflammation, immunosuppression and catabolism syndrome, low (3 kDa), medium (100 kDa) and HMW (1600 kDa) HA were assessed at 30 mg kg⁻¹, revealing distinct biological responses as 1600 kDa HA significantly promoted recovery of intestinal structure and was associated with improved immune recovery while 100 kDa HA induced stronger inflammatory cell infiltration after prolonged treatment and 3 kDa HA showed comparatively milder effects [8].

The HA is a promising safe additive for functional foods. In yogurt, HA at 180, 350 and 1280 kDa showed that 350 kDa improved pH, texture and sensory quality most effectively; 1280 kDa increased stability but caused whey syneresis and 180 kDa decreased stability [9]. For sweetness at pH 4.0, HA at 100, 400 and 1090 kDa extended sweet taste duration, with 100 kDa showing the strongest longest-lasting effect via enhanced sucralose transport. Advances in sustainable production support broader uses, with HMW-HA critical for gel-like texture, stability and viscosity [10]. In meats, a 0.9% HA coating has preserved the freshness of crucian carp for 24 d [11]. The HA has formed stable complexes with fish gelatin, improving its emulsifying characteristics [12]. Products such as Cenovis HA beverage [13] and rooster comb extract in dairy products [14] demonstrate the increasing use of food-grade HA. The HA strength, biodegradability and biocompatibility include excellent potential for edible packaging. Its versatility in forming gels and complexes with neutral taste and color makes it a promising material for food coatings and edible packaging materials [15–17].

Conventional HA production methods face critical limitations. Fermentation using pathogenic Streptococcus strains (S. zooepidemicus) increases biosafety concerns and complicates the downstream processing [18, 19]. The relatively low yield of natural HA producers makes conventional production economically and industrially less feasible. This limitation has accelerated efforts to develop genetically engineered generally recognized as safe (GRAS) microbial platforms for sustainable HA synthesis [20, 21]. In recent years, Bacillus subtilis (B. subtilis) has attracted significant attentions as a microbial cell factory for the biosynthesis of valuable biomolecules and biopolymers [22-25]. Well-characterized genome, efficient protein secretion system and strong metabolic capacity has made B. subtilis a particularly attractive chassis for metabolic engineering and synthetic biology uses [22, 26].

Several studies have demonstrated the feasibility of engineering B. subtilis to produce heterologous biomolecules, highlighting its potential as a safe and scalable platform for industrial biotechnology [23, 27]. Regarding its non-pathogenic nature, endotoxin-free status and established use in industrial biotechnology, this bacterium has been engineered to produce enzymes, polysaccharides and vitamins [23, 24, 28, 29]. In addition, engineered B. subtilis has been used for the production of industrial biopolymers such as poly‑γ‑glutamic acid, further highlighting its potential as a microbial cell factory [30]. However, engineering B. subtilis for HA production presents specific genetic and metabolic challenges. Unlike natural HA producing organisms such as Streptococcus species, B. subtilis lacks a native hyaluronan synthase gene responsible for the polymerization of HA from UDP sugar precursors. Therefore, reconstruction of the HA biosynthetic pathway needs heterologous expression of a functional hyaluronan synthase and B. subtilis only carries precursor-related genes such as tuaD (UDP-glucose dehydrogenase) [21, 22, 31]. In addition, efficient HA synthesis depends on a balanced intracellular supply of key precursors. Limitations in precursor availability and metabolic flux distribution can restrict HA yield, making pathway optimization and targeted metabolic engineering essential for achieving efficient HA production in this host [22, 24]. In addition to heterologous expression of HA synthase, efficient HA production in B. subtilis needs precise metabolic rewiring. This involves redirecting the carbon flux by modulating the competing pathways such as glycolysis and the pentose phosphate pathway while maintaining cellular energy balance. For example, complete inactivation of the pfkA gene is lethal, indicating that these pathways must be carefully fine-tuned rather than fully disrupted [22]. Recent developments in strain engineering and bioprocess optimization with advances in enzyme/pathway engineering have enabled the identification of improved UDP‑sugar pathway enzymes and hyaluronan synthases that enhance HA biosynthesis in non‑pathogenic hosts [21, 22, 32–34].

In this study, a B. subtilis strain was developed that was capable of efficient HA production through heterologous expression of the hyaluronan synthase gene, which is absent in the native genome and NX02_04625, a functional homolog of tuaD that enhances the supply of UDP-GlcUA precursors [22, 24]. Single and dual-gene expression systems were also constructed and compared. By introducing a novel genetic combination of hyaluronan synthase gene with an additional tuaD homolog from another bacterial source with favorable functional traits, this approach establishes a novel expression platform that addresses biosafety concerns associated with pathogenic production strains and provides a promising foundation for scalable HA biosynthesis. Unlike previous studies that overexpressed the native tuaD or used various hyaluronan synthase sources, this study co‑expressed a codon‑optimized hyaluronan synthase from S. dysgalactiae with a high‑performance tuaD homolog (NX02_04625) from Sphingomonas sanxanigenens, a combination not previously reported in B. subtilis. This approach potentially improves critical metabolic difficulties identified in previous studies and contributes to the advancement of safe and efficient microbial platforms for HA production.

1. Materials and Methods
   • Bacterial strains and growth conditions

Escherichia coli (E. coli) DH5α ATCC 53868 used for plasmid propagation and B. subtilis 168 ATCC 23857 used as the expression host for HA production were provided by the National Institute of Genetic Engineering and Biotechnology. Cultures were grown in Luria-Bertani (LB) medium (10 g l⁻¹ peptone, Merck, Germany; 5 g l⁻¹ yeast extract, Merck, Germany; 10 g l⁻¹ NaCl, Merck, Germany). Precultures were prepared by inoculating a single colony into 5 ml of LB medium and incubating overnight at 37 °C with agitation at 180 rpm. For plasmid maintenance, appropriate antibiotics were added to the medium, including ampicillin (Sigma‑Aldrich, USA) at a final concentration of 100 μg ml⁻¹ for E. coli and chloramphenicol (Sigma‑Aldrich, USA) at a final concentration of 10 μg ml⁻¹ for B. subtilis. For production experiments, the overnight preculture was transferred into fresh medium at 1% (v/v). The selected production medium contained sucrose (30 g l⁻¹; Merck, Germany), yeast extract (10 g l⁻¹; Merck, Germany), soybean peptone (5 g l⁻¹; Merck, Germany), ammonium sulfate (1 g l⁻¹; Merck, Germany), K₂HPO₄ (9.15 g l⁻¹; Merck, Germany), KH₂PO₄ (3 g l⁻¹; Merck, Germany) and trisodium citrate (1 g l⁻¹; Merck, Germany) [22].

Gene expression was induced by adding isopropyl β‑D‑1‑thiogalactopyranoside (IPTG) to a final concentration of 1 mM (Thermo Fisher Scientific, USA) at the beginning of cultivation and the cultures were incubated at 30 °C with agitation at 180 rpm. Samples were collected at designated time points (24, 48 and 72 h after inoculation of the production culture, depending on the experiment) and centrifuged at 2400× g for 5 min to collect cell‑free supernatants, which were used for further analyses.

• Gene selection and vector design

To confer HA biosynthetic capability to B. subtilis 168, hyaluronan synthase gene from S. dysgalactiae ATCC12394 was selected based on its catalytic performance reported in the Braunschweig Enzyme Database. The gene encodes a hyaluronan synthase used for construct assembly and experimental procedures. To further enhance HA production, NX02_04625 from S. sanxanigenens DSM19645 was coexpressed, which was a tuaD homolog with favorable kinetic characteristics.

The two genes were codon-optimized for B. subtilis (designed using SnapGene v.3.2.1; GSL Biotech, USA). These were then inserted into the shuttle vector pHT01 under the IPTG-inducible promoter Pgrac01, forming a single transcriptional unit with an intergenic ribosome binding site to enable independent translation. A schematic representation of the dual-gene construct (pHT01-hyaluronan synthase gene-NX02_04625), designed for HA production in B. subtilis 168 (Figure 1), synthesized and sequence-verified by Gene Universal, USA, through its official representative of Pars Biotek, Iran. A single-gene control plasmid was constructed in house.

2.3. Cloning procedures and construct validation

The dual-gene plasmid, hereafter referred to as PST, was reconstituted and diluted to a working concentration. Chemically competent E. coli DH5α cells were transformed using the standard heat-shock method with minor modifications. Transformed cells were plated on LB agar containing ampicillin [35]. To excise the NX02_04625 region, the PST was digested with the XbaI restriction enzyme (Thermo Fisher Scientific, USA; lot no. 01258845), resolved on agarose gel and the PS was excised and purified using FavorPrep gel/PCR purification mini kit (Favorgen Biotech, Taiwan; cat. no. FAGCK001) according to the manufacturer’s instructions. The purified backbone-hyaluronan synthase gene, referred to as PS, was ligated using T4 DNA ligase (Thermo Fisher Scientific, USA; lot no. 00024630).

Ligation products were transformed into E. coli DH5α and plated on ampicillin LB agar. Plasmid DNA was extracted using plasmid DNA extraction mini kit (Favorgen Biotech, Taiwan; cat. no. FAPDE050) according to the manufacturer’s instructions. Construct sizes were verified using gel electrophoresis.

To prepare 0.7% (w/v) agarose gels, 0.7 g of agarose was dissolved in 100 ml of 0.7× tris-borate-EDTA (TBE) buffer. The mixture was heated using microwave. After cooling down, the gel was poured into a casting tray and a comb was inserted to form the wells. After solidification, the gel was transferred into an electrophoresis tank filled with 0.7× TBE buffer. The DNA samples were loaded and separated at a constant voltage of 5–8 V cm⁻¹ until adequate migration was achieved. Band sizes were determined by comparison with 1-kb plus DNA ladder (SinaClon, Iran; cat. no. SL7052).

• Transformation into Bacillus subtilis 168 and validation

The minimum inhibitory concentration (MIC) of chloramphenicol for B. subtilis 168 was assessed as 10 μg ml⁻¹. Electrocompetent B. subtilis cells were transformed with PS and PST constructs via electroporation [36] and selected on LB-chloramphenicol plates. Colonies were validated using colony PCR and gene-specific primers synthesized by Genfanavaran, Iran. Colony PCR was carried out on B. subtilis transformants to verify insertion of recombinant constructs.

• Hyaluronic acid purification

From the media [27, 37], only the modified Westbrook formulation [22] supported HA synthesis. Culture supernatants from engineered B. subtilis were clarified by centrifugation and subjected to sequential trichloroacetic acid (Merck, Germany) and ethanol (Merck, Germany) precipitation. For trichloroacetic acid precipitation, 100 µl of 100% (w/v) trichloroacetic acid were added to 1 ml of clarified supernatant, followed by incubation on ice for 30 min and centrifugation at 16,200× g for 15 min at 4 °C. The resulting supernatant was mixed with 2 ml of cold ethanol (absolute), transferred to -70 °C for 2 h and centrifuged at 4,400× g for 20 min at 4 °C. The pellet was washed with 1 ml of cold ethanol (absolute), incubated at -20 °C for 1 h and recentrifuged. After air‑drying, the HA pellet was dissolved in 1 ml of deionized water and incubated at 45 °C with agitation until fully solubilized. Samples were stored at 4 °C until further analysis.

• Cetyltrimethylammonium bromide turbidimetric method

The HA levels were quantified using cetyltrimethyl-ammonium bromide turbidimetric assay. Samples were mixed with acetate buffer (0.2 M, pH 6; Merck, Germany) and cetyltrimethylammonium bromide reagent (2.5% w/v in 0.5 M NaOH; Merck, Germany) and absorbance was measured at 540 nm. A standard curve prepared from commercial HA (Sigma-Aldrich, USA; lot no. BCBN4845V) was used for the quantification, ensuring a consistent linear relationship between absorbance and HA concentration within the assay working range [38].

• Fourier-transform infrared analysis

Fourier-transform infrared (FTIR) spectroscopy was used to verify the structural identity of the purified HA. Spectra were recorded in the range of 4000–400 cm⁻¹ using standard scanning parameters on the Bruker FTIR spectrometer (Bruker, USA). Characteristic peaks corresponding to hydroxyl, carboxyl and amide groups were assessed.

• Molecular weight determination by size‑exclusion chromatography

Size‑exclusion chromatography separates molecules based on their hydrodynamic size, which depends on molecular weight and molecular conformation in solution. This technique is historically known as gel filtration or gel permeation chromatography (GPC) [39]. The GPC was carried out to assess molecular weight distribution of the purified HA samples. The system (Waters, USA) was equipped with two PL aquagel-OH 8 µm mixed-H columns (300 × 7.5 mm each, separation range 6,000–10⁷ g mol⁻¹; Agilent Technologies, USA) and a refractive index detector (Waters 2414, Waters, USA). Deionized water was used as the mobile phase at a flow rate of 1.0 ml min⁻¹ and the column temperature was set at 35 °C. Calibration was carried out using polyethylene glycol standards (Agilent Technologies, USA), ranging from 1.03 × 10³ to 1.18 × 10⁶ g mol⁻¹. All samples were filtered through 0.45-µm membrane filters (Merck Millipore, USA) prior to injection.

• Statistical analysis

All experiments were carried out with at least two biological replicates and results were expressed as mean ±SD (standard deviation). Normality of data was assessed graphically. Statistical comparisons were carried out using Microsoft Excel v.16.79, USA, and GraphPad Prism v.9.5.0, USA. The HA concentrations, calculated based on optical density values and validated standard curve, were used for statistical analyses. For comparison between the wild-type and PST strains, Welch’s two‑tailed t‑test was used to account for unequal variances. One‑tailed t‑test was used to compare HA production between the PS and PST strains. Paired t‑test was used for temporal comparisons (24 against 48 h, 48 against 72 h) within each strain. A p‑value less than 0.05 was considered statistically significant.

1. Results and Discussion

3.1. Construction of the single-gene construct

Polymerase chain reaction analysis of the PST verified the expected amplicon sizes for the two target regions, validating construct integrity. Transformation into E. coli DH5α yielded robust colony formation on selective medium and plasmid extraction produced DNA of the expected size, verifying plasmid stability. Digestion with XbaI generated fragments of the anticipated sizes, enabling recovery and successful assembly of the PS construct.

3.2. Verification of single-gene construct via plasmid size comparison

Size analysis of the plasmid preparations showed a clear distinction between the two constructs. Figure 2 illustrates the agarose gel electrophoresis of the PS and PST constructs. The upward shift observed for the PST variant verified the expected size increase, consistent with the presence of NX02_04625 in PST and its successful excision during generation of the PS construct.

3.3. Introduction of single and dual-gene constructs into Bacillus subtilis 168

The PS and PST constructs were successfully established in B. subtilis 168, as evidenced by the formation of discrete and stable colonies on selective medium. Colony-level analysis verified the expected inserts in B. subtilis 168 transformants. The PS and PST transformants respectively produced amplicons, indicating the presence of PS and PST constructs and verifying successful uptake and stable maintenance in the host strain.

• Initial assessment of hyaluronic acid yield in wild-type against dual-gene recombinant Bacillus subtilis

The comparison between the wild‑type and PST‑harboring B. subtilis 168 strains in LB medium revealed clear evidence of HA production. Figure 3 illustrates the HA concentrations produced by these two strains at 24 and 48 h. The PST recombinant produced 0.22 ±0.03 and 0.31 ±0.02 g l⁻¹ HA at 24 and 48 h respectively, whereas only a negligible signal was observed in the wild-type strain, likely due to background turbidity in the assay. The differences between the two strains were statistically significant at the two sampling times, indicating the HA biosynthetic capacity conferred by the PST construct.

• Comparison of hyaluronic acid production in single and dual-gene-engineered Bacillus subtilis strains

Modified Westbrook fermentation medium was used for comparative evaluation of the PS and PST strains [22]. Figure 4 shows the HA concentrations produced by PS and PST at 24, 48 and 72 h post-induction. In the PST, HA levels reached 1.20 ±0.03 g l⁻¹ at 48 h and 1.14 ±0.07 g l⁻¹ at 72 h with no significant difference between the two time points (p > 0.05). The PS produced 0.49 ±0.07 g l⁻¹ at 48 h and 0.50 ±0.13 g l⁻¹ at 72 h, showing no significant temporal change (p > 0.05). At the two sampling times, HA production in the PST strain was higher than that in the PS strain (p = 0.064), suggesting improved biosynthetic performance of the dual‑gene construct.

The PST construct demonstrated an advantage over the PS construct, achieving a HA yield of 1.20 ±0.03 g l⁻¹ in shake-flasks compared to 0.50 ±0.13 g l⁻¹, a 2.4-fold increase. This clear difference highlighted critical difficulties in the precursor supply chain, specifically the conversion of UDP-glucose to UDP-GlcUA, which apparently improved by the expression of NX02_04625. This finding aligned with and reinforced the metabolic engineering principle established by Westbrook et al. [22], who identified the increase of the UDP-GlcUA pool via the native tuaD gene as the most impactful intervention, boosting the HA titer 5.6‑fold in their B. subtilis system.

A broader comparison of HA yield and molecular weight reported in previous studies provided useful context for interpreting the current results. The B. subtilis systems produce approximately 1–2 g l⁻¹ HA in shake-flask cultures. Under fed-batch fermentation conditions, Corynebacterium glutamicum achieved up to 28.7 g l⁻¹ (0.21 MDa) [40] and 21.6 g l⁻¹ (1.28 MDa) [41]. Studies on B. subtilis in shake-flask systems have reported yields nearly 1.2 g l⁻¹ [22, 24], with molecular weights ranging from 1.5–2.0 MDa [22]. Zhang et al. reported a yield of 1.8 g l⁻¹ under similar conditions [37]. It is noteworthy that the reported HA yields depended strongly on cultivation conditions; fed-batch bioreactor systems generally achieve higher titers than shake-flask cultures. Therefore, the values should be interpreted as approximate comparisons between various production platforms, rather than direct performance equivalents.

The core of our engineering strategy was the rational reconstruction and enhancement of the heterologous HA biosynthetic pathway, enabling a final HA yield of 1.20 g l⁻¹ in shake-flasks. Native B. subtilis possesses the genetic machinery for UDP-N-acetylglucosamine (UDP-GlcNAc) synthesis but lacks the hyaluronan synthase gene and shows limited flux toward the second precursor of UDP-GlcUA [22, 31]. The present approach involved the heterologous coexpression of hyaluronan synthase gene from S. dysgalactiae and a high-performance tuaD homolog (NX02_04625) from S. sanxanigenens, selected for its favorable enzymatic kinetics from the Braunschweig Enzyme Database.

• Fourier-transform infrared analysis of produced hyaluronic acid

Figure 5 shows the FTIR spectrum of purified HA produced by engineered B. subtilis 168 carrying the PST plasmid. This spectrum showed the characteristic absorption bands of HA, including O–H stretch (3415 cm⁻¹), C–H stretch (2934 cm⁻¹), asymmetric and symmetric COO⁻ stretches (1683 and 1459 cm⁻¹) and C–O–C vibration at 1097 cm⁻¹. The current characterization methods supported the biosynthesis of HA. The FTIR spectroscopy showed absorption bands attributable to key functional groups (O–H, COO⁻ and C–O–C) consistent with previously reported HA spectra [42]. This was consistent with functional heterologous expression of the hyaluronan synthase gene, a critical step toward biomedical applicability [43].

• Molecular weight distribution by size‑exclusion chromatography

Figure 6 illustrates results of size‑exclusion chromatography analysis of purified HA produced by engineered B. subtilis 168 carrying the PST plasmid. This revealed a trimodal distribution.

• Molecular weight parameters

Quantitative MW data for fractions within the calibration range are reported in Table 1. The HMW-HA (Peak 1) exceeded the column calibration range; its MW was estimated as > 1.18 × 10⁶ g mol^-¹ but needed absolute quantification. The GPC analysis revealed a heterogeneous molecular weight distribution, which was typical for microbial HA production due to the presence of endogenous hydrolases or shear stress [44, 45]. The trimodal distribution indicated significant heterogeneity. The dominant HMW fraction (43.45%) aligned with native HA intrinsic chain length, while the mid and LMW fractions (56.55% combined) suggested partial hydrolysis during processing. The narrow PDI (1.20) of fragmented HA suggested controlled degradation, whereas the broader PDI (1.34) of oligomers reflected heterogeneous chain cleavage. In the present study, the engineered B. subtilis strain produced HA with a dominant molecular weight exceeding 1.18 MDa. The HMW-HA (>1.18 MDa) is highly valued in medical and food uses [46, 47]. While this heterogeneity might need further purification for specific uses, it presented an opportunity as low and medium-MW-HA fractions include their own unique therapeutic and food grade uses [48–50]. The average molecular weight of the main fragmented fraction was calculated at 9.88 × 10⁵ g mol^-¹, which was within the highly desirable range for commercial uses [51].

A previous study reported that HMW-HA increased the hardness of ginkgo seed protein isolate gels and improved in vitro digestion by up to 70%. Gel texture and digestibility could be controlled by adjusting pH and HA molecular weight, likely due to physical entanglement between HA and protein molecules that restricted their movement and stabilized the gel network [52]. The HMW-HA enhanced texture (via stronger gel networks) and nutrition‑related attributed such as digestibility. Its mucoadhesive characteristics prolonged the contact time of substances with taste receptors, aiding in the perception of saltiness [50].

The escalating demand for HA in the food, pharmaceutical, biomedical and cosmetic industries necessitates a paradigm shift away from traditional, limited production methods. While extraction from animal tissues increases concerns about pathogen contamination and batch-to-batch variability and fermentation using pathogenic Streptococcus strains introduces significant biosafety and regulatory hurdles, the development of efficient GRAS microbial platforms has become a central focus of metabolic engineering. The advantage of the current B. subtilis-based system includes its combination of a GRAS status, high secretory capacity, well-known genetics and absence of endotoxins, which may facilitate safer development and regulatory consideration for food and biomedical uses, compared to E. coli [53].

In summary, this study provides a minimalistic, effective genetic engineering strategy for HA production in B. subtilis. By co-expressing hyaluronan synthase gene and a high-performance tuaD homolog, the study helped decrease precursor bottlenecks under the tested conditions, achieving shake-flask titers that are competitive with early non-pathogenic systems. The GRAS status, high secretory capacity and well-characterized genetics of B. subtilis suggest that this platform may represent a safer alternative to conventional animal-derived or pathogenic fermentation systems. The findings are similar to the authors’ recent complementary studies on HA production systems. In those studies, other HA-producing strains were assessed under various conditions and the regulatory mechanisms of HA biosynthesis were investigated using genome-scale approaches [54, 55]. Insights from those studies contributed to the design and development of the engineered strain in the present study.

Although the present study was carried out at the shake‑flask scale, scaling up HA production to bioreactor systems may introduce several engineering and physiological challenges. Accumulation of HA can increase broth viscosity; thereby, limiting mixing efficiency and oxygen transfer. The HA biosynthesis competes with cellular metabolism for key precursors such as UDP-GlcNAc and UDP-GlcUA, potentially imposing metabolic burden during high‑density cultivation. Furthermore, agitation and aeration affect HA production and molecular weight, while excessive shear stress may damage cells and decrease polymer size. Accumulation of byproducts such as lactic acid may inhibit cell growth and HA synthesis, highlighting the need of careful bioprocess design during scale‑up [19]. The high viscosity of HA-rich fermentation broths poses significant challenges for downstream processing, including filtration and recovery of HA. These challenges necessitate careful optimization of the purification steps, especially during scale-up to industrial production [56].

1. Conclusion

In this study, the dual-gene-engineered strain demonstrated enhanced HA yield, compared to the single-gene-engineered strain, highlighting the critical role of precursor supply in HA biosynthesis. Characterization of the purified HA by FTIR verified its chemical identity, showing all expected functional groups, while GPC analysis revealed a predominantly HMW polymer fraction, suitable for food, biomedical and cosmetic uses. The findings establish this engineered B. subtilis platform for HA biosynthesis, with the potential to meet growing market demand for safe and high quality HA. This study highlights B. subtilis as a promising chassis for the safe and efficient biomanufacturing of high-value biopolymers.

In addition to the scientific and technical contributions, this study includes broader industrial significance. The establishment of a GRAS, non‑pathogenic and secretion‑competent B. subtilis platform provides a further safe, regulatory‑friendly cost‑effective alternative to animal‑derived or pathogenic fermentation routes currently used in the biopharmaceutical and cosmetic industries. The ability to produce HMW-HA, a premium grade material for dermal fillers, wound healing products, moisturizers and functional foods, highlights the translational potential of this platform as a promising foundation for further industrial bioprocess development. However, translating this proof‑of‑concept into an industrially relevant platform needs further optimization of process scale-up, molecular weight control and plasmid stability.

Based on the limitations identified in this study, further research should focus on improving HA production in B. subtilis through metabolic engineering and process optimization. Although a titer of 1.20 g l^-¹ was achieved at shake-flask scale without optimization, additional studies are needed to enhance production efficiency and assess plasmid stability during scale-up. Further studies investigate hydrolase gene deletion to improve molecular weight, optimization of fed-batch fermentation and engineering strategies to decouple production from sporulation for further scalable and stable processes.

1. Declaration

5.1. Acknowledgements

This study was carried out as a part of the first author's PhD project.

5.2. Declaration of competing interest

The authors declare no conflict of interest.

5.3. Authors’ Contributions

Rouzbeh Almasi Ghale: Conceptualization; data curation; formal analysis; investigation; methodology; validation; visualization; writing – original draft; writing – review & editing.

Reza Faghihi: Investigation; formal analysis; writing – original draft; visualization.

Marjan Talebi: Formal analysis; writing – original draft; writing – review & editing; visualization.

Mehdi Shamsara: Conceptualization; formal analysis; investigation; methodology; supervision; validation; writing – original draft; writing – review & editing.

Fatemeh Tabandeh: Conceptualization; formal analysis; investigation; methodology; project administration; funding acquisition; supervision; validation; writing – original draft; writing – review & editing.

5.4. Using Artificial Intelligent Chatbots

The authors used ChatGPT (OpenAI) for text refinement and improving readability of the manuscript. All scientific content, data interpretation and final responsibility for the manuscript are linked to the authors.

5.5. Ethical Consideration

None.

Background and Objective: The metabolomic profile represents the totality of all low-molecular-weight metabolites. It reflects the physiological state of micro-organisms and their potential for biotechnological uses. The aim of this study was to analyze the metabolomic profile of Lactococcus strains, which allowed the researchers to study the characteristics of these microorganisms. As a result, principles for designing lactic acid bacteria consortia with pronounced production-significant chara-cteristics for the creation of next-generation fermented milk products can be developed.

Material and Methods: The study microbes were Lactococcus strains. The content of organic and amino acids in the experimental samples was assessed by capillary electrophoresis. The contents of vitamins, mono- and disaccharides were assessed using high-performance liquid chromatography.

Results and Conclusion: The metabolomic profiles of 16 industrially associated Lactococcus strains were analyzed. All strains produced lactic acid, with the highest content observed in Strain Ll1, at 8677.8 mg·kg^-1. The experimental samples showed a significant increase in methionine content, with the highest value of 311.2 mg·(100 g)^-1 in Strain Ll3 and cystine content, with the highest value of 45.6 mg·(100 g)^-1 in Strain Ld1. Strain Lc8 showed the ability to synthesize methionine and cystine, which might be promising in cheese production, since these sulfur-containing amino acids contribute to the aroma and flavor of fermented milk products. This strain produced formic, succinic and acetic acids and could catabolize citric acid and galactose. The Strains Ld1, Ld2 and Ld3 showed galactose accumulation, indicating a lack of enzymatic activity for the oxidation of this monosaccharide. Literature data on variability in vitamin synthesis levels between the strains have been verified. Thus, only Strain Ld2 synthesized vitamin B2, the content of which significantly exceeded the control and reached 103.18 ±2.06 μg·(100 g)^-1. Five strains were capable of synthesizing vitamin B6 up to 247.69 ±5.45 μg·(100 g)^-1. The vitamin B9 content increased, compared to the control, in eight samples, with maximum values of 69.59 ±1.46 μg·(100 g)^-1 for Strains Ll3. The dataset on lactococci metabolomic profiles can serve as a basis for the further development of methodological systems, mathematical models and algorithms for selecting strains for the creation of consortia and the production of products, including those with functional characteristics. The results can serve as a basis for the design of consortia of lactic acid bacteria in biotechnological approaches to produce novel fermented milk products with specific characteristics, including functional ones, which include positive effects on human organs when consumed regularly.

Keywords: Lactic acid bacteria, Lactococcus, Functional characteristics, Metabolomic profiles, Metabolites

1. Introduction

Fermented dairy products contain various metabolites produced by starter cultures that offer health benefits to the human body. These include organic acids, exopolysacchar-ides, vitamins and nitrogenous compounds such as peptides and amino acids. Metabolites are formed as a result of the vital activity of lactic acid bacteria (LAB), which ferment lactose into lactic acid, hydrolyze proteins and lipids to peptides, amino acids and free fatty acids. Some metabolites can regulate human metabolism, decrease inflammation or include antimicrobial effects [1]. Lactococcus bacteria (Streptococcaceae family) are among industrially asso-ciated bacteria in the dairy industry. They are Gram-positive, facultative anaerobic catalase-negative cocci [2]. Their primary role in fermentation processes includes efficient conversion of lactose into lactic acid, thereby increasing medium acidity, which enhances antimicrobial characteristics and extends the shelf life of the final product [3]. Lactose metabolism occurs via the pentose phosphate pathway and/or the Embden-Meyerhof-Parnas pathway (glycolysis) through the action of the enzyme β-galactosidase, which hydrolyzes lactose into glucose and galactose. Glucose enters glycolysis, forming pyruvate, which is subsequently reduced to L-lactic acid by lactate dehydrogenase. Another major route of lactose catabolism is the pentose phosphate pathway, in which lactose is phosphorylated to lactose-6-phosphate during transport and then cleaved into glucose and galactose-6-phosphate by phospho-β-galactosidase. Then, galactose-6-phosphate is converted into tagetose-6-phosphate by isomerase, which turns into glyceraldehyde-3-phosphate and dihydroxyace-tone phosphate. Glyceraldehyde-3-phosphate enters glycol-ysis and leads to pyruvate and L-lactic acid formation [4]. The accumulation of L-lactic acid leads to a decrease in the pH of the medium, which determines the intensity and pathway of lactic acid fermentation. Other key factors affecting this process are the acidification rate and the presence of oxygen and carbon dioxide in the medium. Within the Genus Lactococcus, L. lactis subsp. lactis stands out for its higher rate of lactose utilization and the ability to ferment maltose and trehalose [2].

Moreover, acid production is indirectly linked to exopolysaccharide (EPS) production. The EPS are high-molecular-weight carbohydrate polymers (either homo or heteropolysaccharides) secreted by microorganisms into the extracellular environment. Their precursors are intermediate metabolites of glycolysis and the pentose phosphate pathway glucose-6-phosphate, fructose-6-phosphate, uridine diphosphate glucose and other nucleotide sugars. Since these intermediates serve as precursors for lactic acid, a competitive relationship for carbohydrate substrates occurs [5]. The highest activity of the enzymatic complex of EPS synthesis is achieved at lower acidity levels during lactate biosynthesis (pH 5.0–5.5) [6, 7]. The production of EPS leads to the formation of a slimy layer surrounding the cell, which forms a colloidal aggregate in the form of an amorphous substance with little or no cell adhesion or a cohesive capsule [8]. Acid formation and then EPS production by Lactococcus spp. contribute to the texture of dairy fermented products, resulting in a smooth, creamy, homogeneous consistency with possibly slight ropiness.

Cow milk contains 0.4–0.5 mg·(100 g)^-1 of potassium and sodium citrates. Among Lactococcus species, only L. lactis subsp. lactis var. Diacetylactis is capable of citrate metabolism. End products of citrate metabolism include diacetyl, acetoin, 2,3-butanediol, acetaldehyde, ethanol and lactic acid. The induction of the citrate permease catalytic enzyme is caused by the presence of lactic acid rather than citrate, which is synthesized from glucose and citrate under acidic pH conditions [9]. Thus, acid formation affects citrate metabolism. The most valuable metabolic products in this biosynthesis are diacetyl and carbon dioxide, as they improve the texture and flavor of fermented dairy products.

Lactococcus species include a limited ability to synthesize amino acids de novo from inorganic carbon sources and are therefore auxotrophic. However, fermented dairy products can be enriched with amino acids through the proteolysis of milk caseins [10]. Extracellular proteinases of Lactococcus hydrolyze caseins into oligopeptides, which are then transported into the cell and cleaved by intracellular peptidases into amino acids [11]. The activity of proteolytic enzymes and transport systems depends on the availability of nitrogen, acidity and the content of free amino acids and is controlled by regulatory systems. Particularly, L. lactis subsp. cremoris cleaves β-casein with the similar protease specificity as L. lactis subsp. lactis, but has less active proteolysis [12]. Amino acids are used for cellular biosynthesis and as precursors of biologically active compounds. Several amino acids are degraded to volatile compounds (aldehydes, ketones and alcohols), which can affect the sensory characteristics of fermented dairy products [12, 13]. Partial hydrolysis of milk proteins by lactococci, with lactose reduction, enhances digestibility in the human gastrointestinal tract (GIT) and may decrease the risk of intolerance, compared to raw milk.

Amino acids play a critical role in the synthesis of the bacteriocin of nisin (a polycyclic peptide composed of 34 amino acid residues, including the non-proteinogenic amino acid lanthionine) which is active against a broad spectrum of Gram-positive bacteria [14]. Studies have shown that the presence of amino acids such as alanine, isoleucine, serine, glutamic acid, tyrosine and tryptophan in the culture medium enhances bactericidal activity against strains of Escherichia coli and Staphylococcus aureus [14, 15]. Regarding vitamin biosynthesis, Lactococcus species are largely auxotrophic due to their adaptation to nutrient-rich environments such as milk, where several vitamins are readily available. Nevertheless, certain strains have the ability to synthesize B-group vitamins, such as folates (B9) and riboflavin (B2) [16]. Folate production levels vary significantly between the strains. Efforts to enhance B9 biosynthesis in LAB have involved optimized cultivation conditions and genetic modification through strain selection and metabolic engineering [17]. The ability of Lactococcus strains to produce riboflavin is limited and highly strain-dependent. Some strains function as “riboflavin producers” and are used to enrich dairy products with improved nutritional value, as well as imparting a yellowish color to whey and cheeses [18, 19].

The metabolomic profile reflects the key characteristics of the microorganism that determine its technological potential in dairy production. Industrially associated characteristics of strains include technological significance in the production of dairy products (e.g., fermentation time and viscosity) Understanding the genetic and regulatory mechanisms underlying metabolite biosynthesis pathways in Lactococcus species enables the targeted selection of starter cultures and development of microbial consortia, which is critical for creating novel biotechnological products. The aim of this study was to analyze the metabolomic profile of Lactococcus strains, which allowed the researchers to study the characteristics of these microorganisms. As a result, principles for designing LAB consortia with pronounced production-significant charact-eristics for the creation of next-generation fermented milk products can be developed.

1. Materials and Methods

2.1. Experimental samples

The microorganisms of this study were Lactococcus strains from the collection of the All-Russian Research Institute of the Dairy Industry (“VNIMI”), selected based on an evaluation of the most important characteristics for dairy processing, including the viability after long-term storage, acidification rate, titratable acidity and pH, apparent viscosity and sensory characteristics. The selected strains included four strains of L. lactis subsp. lactis biovar diacetylactis (Ld1, Ld2, Ld3 and Ld4), eight strains of L. lactis subsp. cremoris (Lc1, Lc2, Lc3, Lc4, Lc5, Lc6, Lc7 and Lc8) and four strains of L. lactis subsp. lactis (Ll1, Ll2, Ll3 and Ll4). Before use, the strain was preserved at -80 °C under glycerol. The strain was then cultured in milk at 30 °C. A second-generation culture was used. The initial cell concentration was 10⁷ CFU·ml^-1. To obtain experimental samples, sterile skimmed milk (the "Standard" brand Complimilk, Belarus) was fermented with the test strain at a concentration of 5% and incubated at (30 °C ±2) until a clot formed (рН 4.5 ±0.2). The control was sterile skim milk. The milk was prepared by dissolving 90 g of dry skim milk in 1000 ml of tap water and sterilized by autoclaving using N-Bioteck sterilizer, Korea, at 121°C for 3 min. The initial pH was 6.8 and was set in a thermostat with the experimental samples at 30 °C. A limitation of this study was the number of strains (40). The study was limited to assessing the characteristics of representatives of a genus, lactococci, the most common type of LAB used in food production. The studied strains were isolated from homemade fermented products and natural sources in a temperate climate zone. Moreover, the study did not include genetically modified strains.

2.2. Assessment of organic and amino acids

The content of organic acids in the experimental samples was assessed using capillary electrophoresis and "Kapel 205" system, equipped with a spectrophotometric detector and a quartz capillary (75-µm inner diameter, 60-cm total length). Samples were pre-diluted with distilled water. The buffer electrolyte was prepared using benzoic acid, diethanolamine, cetyltrimethylammonium bromide and Trilon B. Separation was carried out at 20 kV with UV detection at 254 nm. Electropherograms were processed using "Elforan" software.

For the assessment of amino acid composition, the samples underwent acid hydrolysis and, alkaline hydrolysis for tryptophan, to convert protein-bound amino acids into free forms. All amino acids, except tryptophan, were derivatized into phenylisothiocarbamyl derivatives and quantified via capillary electrophoresis. Tryptophan was assessed directly without derivatization using borate buffer, +25 kV voltage and UV detection at 219 nm. Glutamic acid, aspartic acid and cystine were analyzed in phosphate buffer with β-cyclodextrin under +25 kV voltage, 50 mbar pressure and UV detection at 254 nm. The rest of amino acids (arginine, lysine, tyrosine, phenylalanine, histidine, leucine and isoleucine, methionine, valine, hydroxyproline, proline, threonine, serine, alanine and glycine) were assessed using similar method without pressure. All electropherograms were processed using "Elforan" software. For such amino acids as tryptophan, glutamic acid and glutamine, and aspartic acid and asparagine, the associated error was ±20%; for arginine, tyrosine, phenylalanine, histidine and methionine, the associated error was ±23%; for lysine, leucine and isoleucine, valine, proline, threonine, serine, alanine and glycine, the associated error was ±18%; for cystine and cysteine, the associated error was ±24%; and for organic acids, the associated error was ±20%.

2.3. Assessment of mono and disaccharide contents

The contents of mono and disaccharides in the experimental samples were assessed using high-perfor-mance liquid chromatography (HPLC) in accordance with GOST 54760-2011, “Component milk products and infant milk products. Assessment of mono and di-sugars mass concentration using HPLC method.” The analysis was carried out using MAESTRO liquid chromatograph (INTER-LAB, Russia), with a Zorbax carbohydrate analysis column (5 μm, 4.6 × 250 mm; Agilent Technologies, USA), CAU-X-320 electronic analytical balance (CAS, Korea), variable-volume single-channel pipette (100–1000 μl; BIOHIT, Finland; Sartorius Biohit Liquid Handling, Germany) and Sigma 1-14 microcentrifuge (Sigma Laborz-entrifugen, Germany). The associated measurement error was ±12%.

2.4 Assessment of vitamins

Vitamin C was assessed using Agilent 1260 chromato-graphy system equipped with a gradient 4-channel pump, a diode array detector, a column thermostat and an automatic sample delivery system. The associated measurement error was ±34%. The analyte was extracted by diluting the sample aliquot 1:2 with an extracting solution (3% metaphosphoric acid). It was vortexed and centrifuged at 10.850× g. The supernatant was filtered through a 0.22-μm filter. The resulting filtrate was used for analysis. Separation of components was carried out by reversed-phase chromate-graphy on an Agilent Extend-C18 4.6*250 mm column. Vitamin C was eluted under isocratic conditions using mobile phase of 30 mM phosphate buffer, pH 2.6, at a flow rate of 1 ml·min^-1. Detection was carried out using diode array detector at 254 nm.

B-group vitamins were assessed using HPLC with an Agilent 1260 Infinity II HPLC device combined with an Ultivo Triple Quad LC/MS mod. 6465, Singapore, Agilent Technologies and an Agilent 1260 Infinity II HPLC device with a diode array detector G7115A. Chromatographic separation was carried out using Agilent InfinityLab 120 Poroshell 120 Phenyl-Hexyl column (3.0 × 100 mm, 2.7 μm). For sample preparation, 1 g of the sample was mixed with 4 cm³ of deionized water, vortexed and then 5 cm³ of acetonitrile and 0.1 g of ascorbic acid were added to the mixture. The mixture was re-vortexed, ultrasonicated for 30 min and centrifuged at 3.650× g for 10 min. The sample was stored at -4 to -6 °C and filtered through a 0.22-μm filter.

2.5. Statistics

The reliability of data was verified through the execution of experiments in at least three independent replicates. Furthermore, MS Office Excel 2016 was used for data analysis and graph creation. Experimental data were presented in“mean value ± measurement error (relative)”. The data on vitamins were present as mean ± standard deviation. To calculate the significance of differences, ANOVA statistical method was used, with the accepted limitation of statistical significance of differences between options at p ≤ 0,05. The study was carried out using equipment of the Collaborative Center of the All-Russian Dairy Research Institute (CKP “VNIMI”).

1. Results and Discussion

3.1. Assessment of amino acids

Amino acids are promising metabolites for biotech-nological uses. These compounds are used as artificial sweeteners, flavoring agents and feed additives and for pharmaceutical purposes [19, 20]. During the production of fermented dairy products, specific amino acids generated via casein proteolysis are responsible for the formation of thiols, alcohols, esters and aldehydes, contributing to a broad range of flavors [20]. Sulfur-containing amino acids play a critical role in the development of sensory charac-teristics, as well as in the antioxidant activity of fermented dairy products [21]. In the experimental samples containing the industrially associated Lactococcus strains, the amino acids were assessed. Compared with the control (milk), a significant increase in the content of methionine and cystine was detected in the experimental samples. The results are presented in Figure 1.

A significant increase in amino acid content relative to the control was observed for methionine (Figure 1A) in Strains Lc2, Lc6, Lc8, Lc4, Ll2, Ll3, Ll4, with the highest content in Strain Ll3 at 311.2 mg·(100 g)^-1. The sulfur-containing amino acid methionine plays a role in the initiation of transcription and is critical in various methyltransferase reactions. Aromatic compounds such as 3-methylbutanal, methanethiol, dimethyl sulfide, 2-methylpropanol and dimethyl trisulfide are synthesized from methionine [22]. Among these, methanethiol is particularly important for its contribution to the desirable flavor of cheddar cheese, especially in combination with diacetyl and butyric acid [23]. Increase in cystine content (Figure 1B) was observed in Strains Ld1, Ld2 and Lc8, with the highest content in Strain Ld1 [45.6 mg·(100 g)^-1], compared to the control. Cystine is the oxidized form of cysteine, which serves as a key metabolite for synthesizing most sulfur-containing cellular compounds. Cysteine is involved in protein folding, assembly and stability through the formation of disulfide bonds. Additionally, cysteine-containing proteins such as thioredoxin and glutathione play essential roles in protecting cells against oxidative stress [24]. The strain with the conditional designation (Lc8) is capable of synthesizing methionine and cystine. The levels of other amino acids were similar to or less than those of the control, suggesting their active utilization by microorganisms during growth.

3.2. Assessment of organic acid content

One of the fundamental criteria for selecting Lactococcus strains for starter culture consortia is their ability to produce acid. Organic acids such as acetic, lactic, citric, succinic and propionic acids are produced by LAB during carbohydrate metabolism. As intermediates in metabolic pathways, these acids enhance the taste of products and extend shelf life through their antimicrobial and antioxidant characteristics; thereby, increasing consumer value [25,26]. The results of assessment of the content of organic acids are presented in Figure 2.

All the samples showed a significant increase in lactic acid content, compared to the control (Figure 2). The highest levels were observed in Strains Ld1 and Ll1, reaching 8116.5 and 8677.8 mg·kg^-1, respectively. Lactic acid synthesis indicates the β-galactosidase enzymatic activity of these microorganisms. Lactic acid in fermented milk products acts as an acidity regulator and preservative. When ingested, lactic acid accelerates metabolism, decreases the number of pathogenic bacteria and also has antioxidant characteristics.

Only trace quantities of citric acid were detected in Samples Lc1, Lc3 and Lc5, which indicated the presence of a citrate pathway for the breakdown of this substrate. The consumption of citric acid depends on the presence of genes encoding the enzymes citrate permease and citrate lyase, which ensure the transport of citrate into the cell and its catabolism into oxaloacetate [27]. As a result of breakdown by lactococci, gas formation occurs in dairy products. In the production of certain products such as cheeses, this is addressed as a positive phenomenon, leading to the formation of "eyes." However, if its presence is assessed, this can lead to product defects. The breakdown of citric acid by these microorganisms leads to the formation of acetate, lactic acid, diacetyl and acetoin. Diacetyl and acetoin create flavor and aroma in food products, which is promising for the production of starters for butter, sour cream, yogurt and certain types of cheese.

Significantly, formic acid was detected in Strains Lc1, Lc2, Lc4, Lc6, Lc7, Lc8, Ll2, Ll3 and Ll4. In Lactococcus, formic acid is produced primarily by the pyruvate-formate lyase enzyme, which catalyzes the breakdown of excess pyruvate into acetyl-CoA and formate, which is the initial step in the formation of mixed-acid products. In the L. lactis genome, the genes responsible for this pathway are pfl, encoding pyruvate-formate lyase and pflA, a formate activator. Formate is formed with acetyl-CoA, which is consumed by the cell for energy needs. Formate itself can be excreted from the cell to maintain pH and redox balance [28]. In fermented dairy products, formic acid has little effect on the taste of the product, but it can serve as a preservative to preserve nutrients and it also has a bactericidal effect against pathogens.

Succinic acid was detected in Strains Lc4, Lc8, Ll3 and Ll4, with contents of 83.5, 89.9, 74.9 and 63.3 mg·kg^-1, respectively. This acid is synthesized via heterofermentative lactic acid fermentation alongside lactic acid. Succinic acid is formed upon activation of the reductive branches of the tricarboxylic acid cycle. In Lactococcus strains, the complete tricarboxylic acid cycle is inactive, but succinate can be synthesized through carboxylation of C3–C4 units as pyruvate or phosphoenolpyruvate is carboxylated to oxalo-acetate by pyruvate carboxylase or phosphoenolpyruvate carboxy-lase and then oxaloacetate is sequentially reduced to malate and ultimately converted to succinate [29]. In the human body, succinic acid promotes the synthesis of propionic and butyric acids in microbiota.

Acetic acid production was observed in many strains, with the highest contents recorded in Strain Lc4 and Lc6 at 1148.7 and 1129.8 mg·kg^-1, respectively. Acetate is formed during mixed fermentation from acetyl-CoA by phospho-acetyltransferase and acetate kinase. This results in the generation of energy in the cell in the form of ATP [30]. Acetic acid enhances the sensory characteristics of the fermented dairy products and contributes to microbial safety and shelf-life extension due to its antimicrobial activity.

3.3. Assessment of mono and disaccharide contents

Lactose, the primary carbohydrate in milk, is a disaccharide composed of glucose and galactose residues, hydrolyzed by lactase and β-galactosidase. Lactose metabolism in Lactococcus differs from that of other LAB by allowing simultaneous catabolism of glucose and galactose [9]. Galactose participates in the formation of glycoproteins and glycolipids, which are essential for cell-cell communication. However, its presence in food poses risks for individuals with galactosemia.

Another important disaccharide is lactulose, formed from lactose during high-temperature treatment of milk through isomerization involving hydrogen transfer from the second to the first carbon atom in the glucose residue. Lactulose content in milk serves as an indicator of thermal processing [31]. It is composed of D-fructose and D-galactose linked via a β-1,4-glycosidic bond and shows prebiotic effects at low concentrations. Previous research [32] demonstrated that the presence of lactulose improves the viability of starter cultures during storage, freezing and freeze-drying.

The analysis of monosaccharides and disaccharides (Figure 3) revealed a slight decrease in lactose content, compared to milk in all experimental samples. Particularly, lactulose was detected in all samples except Strains Ld1 and Ld4, including control. A decrease in lactulose content was observed in Samples Ld2, Ld3 and Ll1, compared to the control. The presence of lactulose can enhance consistency, increase starter culture viability and extend the shelf life of fermented dairy products. Galactose was detected in Strains Ld1, Ld2 and Ld3, with contents of 0.295%, 0.605% and 0.354% by mass, respectively. This indicated that the other strains—Lc1 through Lc8, and Ll1 through Ll4 and Ld4—were capable of galactose utilization. After transport into the microorganism, lactose contained in milk is hydrolyzed by β-galactosidase to glucose and galactose. Galactose can either be released by cells into the environment or undergo further metabolism into glucose-1-phosphate by four enzymes of galactokinase (GalK), hexose-1-phosphate uridine transferase (GalT), UDP-glucose-4-epimerase (GalE) and aldose-1-epimerase (GalM). These enzymes make up the Leloir pathway, whose genes can be present either alone (gal genes) or in combination with lactose metabolism genes (lac genes), making the final fermented products safer for individuals with galactosemia.

3.4. Assessment of vitamin contents

Based on the results of amino acid analysis, nine strains that showed a significant increase in methionine and cystine contents in samples were selected, compared to the control (Lc2, Lc4, Lc6, Lc8, Ll2, Ll3, Ll4, Ld1 and Ld2). These strains are of interest for possible inclusion in consortia for the development of products with specific functional characteristics. The selected samples were additionally assessed for vitamin C and vitamin B group (B2, B6, B7 and B9) contents. The results of vitamin C assessment in the samples are presented in Table 1.

As the studies have shown, the content of vitamin C in the samples did not exceed the control levels. The results of the assessment of B vitamins in the samples are presented in Table 2.

A number of foods, including milk, contain vitamin B2 (riboflavin). It is a key enzyme in redox reactions and essential for hemoglobin synthesis, maintaining the normal condition of the GIT mucosa and supporting nervous system function [33]. The ability of lactococci to synthesize riboflavin is limited, but some strains produce the compound in significant quantities [18]. In this study, only Strain Ld2 showed an increased level of vitamin B2 [103.18 μg·(100 g)^-1], compared to the control [79.97 μg·(100 g)^-1]. In the other samples, the content of riboflavin was at the control level or decreased.

Vitamin B6 plays a critical role in the normal functioning of the human body systems, including the digestive, cardiovascular and immune systems. It participates in hundreds of biochemical reactions, affecting metabolism. It is involved in the synthesis of neurotransmitters such as serotonin and dopamine, improves cognitive functions and promotes a positive mood [34]. It affects collagen synthesis and regulates hormonal balance in women. In the samples of industrially associated Lactococcus strains Lc6, Ll4, Ll3, Ll8 and Ll2, a significant content of vitamin B6 was detected, exceeding the control by 3–5 times. The vitamin B7 content in all samples was lower than that in the control.

Analysis of vitamin B9 content showed a significant increase in eight out of nine samples, ranging from 10.02 to 69.59 μg·(100 g)^-1, compared to 4.37 μg·(100 g)^-1 in the control group. Significantly, the Lc2 strain contained less than 1 μg·(100 g)^-1 of vitamin B9. Folic acid (vitamin B9) is essential for appropriate cell growth and division, acting as a coenzyme in nucleic acid synthesis. Folic acid deficiency can lead to serious fetal development disorders in pregnant women [35]. Vitamin B9 plays a critical role in the conversion of homocysteine, a substance associated with an increased risk of cardiovascular disease, into methionine. This compound is necessary for the synthesis of several neurotransmitters, prevents neurological damage, improves cognitive function and decreases the risk of certain types of cancer (breast, cervical and intestinal cancers) [36]. Receiving vitamins through foods is more physiologically beneficial for the human body than introducing chemically synthesized preparations. A comparative assessment of lactococci from the VNIMI collection, stored since 1960, was carried out for the regenerative capacity of LAB strains, as well as for the production-significant characteristics of these microorganisms such as fermentation activity, active and titratable acidity, viscosity and organoleptic indicators. Of the 40 Lactococcus strains, 39 showed metabolic activity regardless of storage time [37].

1. Conclusion

The study included metabolomic profiling of industrially associated Lactococcus strains based on the analysis of amino acids, organic acids, monosaccharides, disaccharides and vitamins. All strains produced lactic acid, with the highest content observed in Strains Ld1 and Ll1, at 8116.5 and 8677.8 mg·kg^-1, respectively. The experimental samples showed a significant increase in the content of methionine, with the highest content in Strain Ll3 and cystine in Strain Ld1. The content of other amino acids was similar to the control level or decreased. Strain Lc8 showed the ability to synthesize methionine and cystine, which may be promising in cheese production, since these sulfur-containing amino acids contribute to the aroma and flavor of fermented milk products. This strain produced formic, succinic and acetic acids and could catabolize citric acid and galactose. The Strains Ld1, Ld2 and Ld3 showed galactose accumulation, indicating lack of enzymatic activity for the oxidation of this monosaccharide. Literature data on significant variability in vitamin synthesis levels between the strains have been verified [18,20]. Thus, only Strain Ld2 synthesized vitamin B2, the quantity of which significantly exceeded that in the control and reached 103.18 ±2.06 μg·(100 g)^-1. Five strains were capable of synthesizing vitamin B6 up to 247.69 ±5.45 μg·(100 g)^-1. The vitamin B9 content increased in eight of the nine samples, with maximum values of 69.59 ±1.46 and 51.16 ±1.28 μg·(100 g)^-1 for the strains Ll3 and Ll4, compared to the control. For further formation of consortia, strains Lc1, Lc8, Ll2, Ll3, Ld1 and Ld2 were selected. The results can serve as a basis for the design of consortia of LAB in biotechnological approaches to produce novel fermented milk products with specific characteristics, including functional ones.

1. Acknowledgements

The authors would like to thank Elena Yurova and Nikolai Zhizhin for their assistance.

1. Declaration of competing interest

The authors report no conflict of interest (If authors have any kind of interest mention please clearly).

1. Authors’ Contributions

Conceptualization, I.R., N.P., V.L., V.S., S.K., V.M.; methodology S.K.; validation, S.K., V.S.; formal analysis, I.R., V.M., S.K..; investigation, V.S., S.K.; data curation, S.K., V.S; writing—original draft preparation, V.S.; writing—review and editing, V.L., I.R., V.S., S.K.; visualization, V.S.; supervision, I.R. N.P.; project administration, I.R, N.P.

1. Using Artificial Intelligent Chatbots

No using Artificial Intelligent Chatbots

1. Ethical Consideration

The ethics committee's opinion is not required in this work

Background and Objective: Environmental concerns over petroleum-based packaging have increased interests in biodegradable edible films from renewable resources. Exopolysaccharides from lactic acid bacteria are promising due to their film-forming ability, safety and functionality. Weissella confusa NH02, isolated from Thai fermented pork (Nham), produces high-molecular-weight glucan exopolysaccharides appropriate for film formation. However, exopolysaccharides films alone are fragile and moisture-sensitive. Glycerol plasticization can improve flexibility and barrier characteristics by modifying intermolecular interactions. This study developed exopolysaccharides films from Weissella confusa NH02 and assessed the effect of glycerol content on film characteristics.

Material and Methods: Exopolysaccharides were produced by submerged fermentation and recovered by ethanol precipitation and drying. Film-forming solutions containing 2.5–10 g kg^-1 exopolysaccharides were prepared with glycerol (6.25–50 grams per 100 gram exopolysaccharides). Dispersions were cast, dried at 40 °C and conditioned under controlled humidity. Rheological behavior was analyzed using rotational rheometer and fitted to Ostwald-de Waele model. Mechanical characteristics were assessed using American Society for Testing and Materials methods. Moisture content, water activity, thickness, transparency, solubility and water-vapor permeability were assessed using standard techniques. Dynamic mechanical thermal analysis investigated viscoelastic behavior and glass transition temperature. All assessments were carried out in triplicate.

Results and Conclusion: Exopolysaccharides produced clear flexible films and dispersions showed shear-thinning behavior appropriate for casting. Glycerol strongly affected performance as water-vapor permeability decreased up to ~25% glycerol due to improved polymer packing then increased at higher levels due to increased hydrophilicity. Mechanical characteristics improved at moderate plasticizer levels, with optimal performance near an exopolysaccharide:glycerol ratio of 65:35 (TS ≈ 5.56 MPa; elongation ≈ 142%), while excess glycerol weakened the matrix. Dynamic mechanical thermal analysis revealed single transition temperature decreasing from ~85 to less than -15 °C as glycerol increased, indicating strong compatibility and efficient plasticization. Plasticized exopolysaccharides films from Weissella confusa NH02 show promise as biodegradable food-contact materials with tunable strength, flexibility and moisture barrier characteristics.

Keywords: Biopolymer materials; Microbial polysaccharides; Plasticizer effect; Barrier characteristics; Mechanical performance; Biodegradable packaging

1. Introduction

Protective edible coating and optimized packaging are increasingly adopted to extend the shelf life of foods. Conventional packaging depends heavily on non‑renewable inputs and can impose environmental costs. Therefore, bio‑based alternatives and edible and biodegradable films derived from renewable resources have been interested as a route to maintain quality while decreasing wastes [1]. In addition to waste decrease, next-generation edible films are increasingly designed to act as active interfaces capable of regulating moisture transfer, limiting oxidation and serving as carriers for natural antimicrobials; thereby, aligning food preservation strategies with circular bioeconomy principles and sustainable material transitions [2,3].

Weissella confusa NH02, isolated from Thai fermented minced pork (Nham), produces exopolysaccharides (EPS) characterized by high viscosity, molecular mass of approximately 1.13 × 10⁶ Da and glucose as the sole monomer [4]. Based on these characteristics and the demand for multipurpose food‑contact films, EPS from W. confusa NH02 is a promising film‑forming substrate. Films assembled from lactic acid bacterial EPS alone can be achieved [5], but limited resistance to water and vapor often limits practical uses. Mixing EPS with compatible, lower‑cost polymers and plasticizers is a common strategy to develop mechanical strength and barrier behavior [6]. A study highlighted that the functional performance of EPS-based films depended not only on composition but also on supramolecular organization and intermolecular interactions formed during drying, which governed transport characteristics and structural integrity [7].

A report has demonstrated that EPS produced by lactic acid bacteria (LAB), including W. confusa, possess excellent rheological characteristics and potential for food, biomedical and biodegradable material uses [8]. The EPS from W. confusa NH02 has been reported as a high-molecular-weight (HMW) glucan with strong viscosity and thickening ability, making it interesting for film formation [4,9]. In addition, research on polysaccharide-based edible films has shown that plasticizers such as glycerol can improve flexibility and decrease fragility [10].

However, several important issues are still unresolved. First, although EPS from W. confusa NH02 has been characterized chemically and rheologically, its intrinsic film-forming capability and appropriateness as a standalone film matrix have not been systematically investigated. Second, EPS-based films are typically limited by poor moisture resistance and mechanical fragility, restricting practical uses in food packaging. Third, while glycerol is widely used as a plasticizer, the optimal concentration needed to balance mechanical strength, flexibility and water-vapor barrier performance in EPS films from this specific strain has not clearly been established. Moreover, limited information are available on the thermomechanical behavior and molecular compatibility of EPS-glycerol systems, particularly regarding glass transition behavior and structure-characteristics relationships.

Therefore, the present study addressed these gaps by developing films from EPS produced by W. confusa NH02 and systematically assessing the effects of glycerol content on rheological behavior, mechanical characteristics, water-vapor permeability (WVP) and thermal transitions. This study aimed to clarify structure-function relationships and provide guidance for optimizing EPS-based biodegradable films for food-contact uses.

1. Materials and Methods

2.1. Microorganism and culture conditions

The W. confusa NH02 strain was recovered on modified MRS agar containing 2% sucrose at 4 °C and sub‑cultured every 3 w. Inoculum preparation followed the procedure of [4,9]. Production media were inoculated at 5% (v/v).

2.2. Exopolysaccharide production setup

The EPS production was carried out in shake flasks using synthetic fermentation media described by Wongsuphachat et al. [4].

2.3. Isolation of exopolysaccharides from fermentation broth

Culture supernatant collected on day 2 was clarified by centrifugation at 4,058× g for 15 min. The EPS was precipitated by adding two volumes of 96% chilled (v/v) ethanol and incubating the mixture at 4 °C for 1 h. The precipitate was recovered by centrifugation and dried under vacuum at 40 °C for 48 h. Materials from three fermentations were pooled and ground to a fine powder. Then, EPS content was quantified following Araujo-Rodrigues et al. [11] by assessing reducing sugars released by β‑glucanase (Sigma‑49101, USA) using DNS assay; β‑1,3‑glucan from Euglena gracilis (Sigma‑89862, USA) served as a positive control.

2.4. Preparation of film-forming solutions

Aqueous EPS solutions (2.5, 5.0, 7.5 and 10 g kg^-1) were prepared under continuous stirring. Glycerol (Baker, Mexico) was added as the plasticizer at 6.25, 12.5, 25.0, 37.5 or 50.0 g per 100 g EPS to assess plasticizer level effects.

2.5. Rheology of filmogenic dispersions

Rheological behavior of the selected EPS concentration was characterized using Haake ReoStress 600, PP35 plate and plate geometry (1 mm gap) at 25 °C. Shear rate increased from 0 to 500 s⁻¹ (4.167 s⁻² acceleration) and then decreased with a similar magnitude. Data were fitted to Ostwald-de Waele model, t = k·Dⁿ, where D was shear rate (s⁻¹), t was shear stress (Pa), k was the consistency index (Pa·sⁿ) and n was the flow index (dimensionless). Apparent viscosity was reported at 300 s⁻¹.

2.6. Film casting and conditioning

           Twenty‑five grams of each filmogenic solution were cast into 8.7‑cm Petri dishes and dried at 40 °C for ~6 h to constant mass using ventilated oven. Films were removed and equilibrated at 20 °C and 75% relative humidity (RH). Prior to mechanical and permeability assessments, specimens were conditioned based on ASTM D618‑61 at 25 °C ±2 and 51% RH over saturated magnesium nitrate for at least 48 h and then sealed using plastic bags and desiccators.

2.7. Dynamic mechanical thermal analysis

Glass‑transition temperature (Tg) of EPS/glycerol films was assessed using dynamic mechanical thermal analysis (DMTA) instrument (Triton Technology, UK) equipped with liquid nitrogen and film grips for uniaxial assessment. Samples (initial grip separation of 5.5 mm) were assessed using small sinusoidal strain (1 Hz, 0.02%), while heating from 69 to 159.6 °C at 5 °C min⁻¹. To minimize water loss, the exposed film surface was partially covered with aluminium foil, leaving the clamped ends uncovered [12]. Storage modulus (E′) and loss tangent (tan δ) were recorded versus temperature. The Tg was recorded at the midpoint between the tan δ peak and the onset of the sharp drop in E′. Assessments were carried out in triplicate.

2.8. Mechanical assessment

Tensile strength (TS) and elongation at break (E) were assessed at 25 °C using Testometric M350‑10CT (United Kingdom) based on ASTM D882 [13]. Strips (25 × 100 mm) were conditioned for 48 h at 51% RH (saturated magnesium nitrate), clamped with an initial grip separation of 50 mm and pulled at 50 mm min^-¹. Stress-strain curves were recorded; TS was calculated from the maximum load divided by the initial cross-sectional area (MPa) and E was the percentage increase in length at rupture associated to the initial gauge length. For puncture assessment, discs (3 cm in diameter) were probed perpendicularly with a 3-mm cylindrical tip at 50 mm min^-¹ to investigate puncture strength (PS) and puncture deformation (PD).

2.9. Physicochemical characteristics

Moisture content was achieved by drying films at 105 °C ±1 to constant mass and expressed as grams water per 100 g material. Water activity was assessed using quaLab meter (Decagon Devices, USA). Thickness was assessed using Check Line DCN‑900 (USA); fifteen random positions per specimen were averaged. Transparency was assessed from A600 assessed using Beckman DU650 (USA) and normalized by thickness (A600 mm^-¹) according to [14,15]. Solubility of films with or without 25 g glycerol per 100 g polysaccharide was assessed by incubating samples in deionized water at 20, 37 or 100 °C for 2 h under agitation; then, supernatant polysaccharides were quantified using anthrone method [16].

2.10. Water-vapor permeability

The WVP was assessed based on ASTM E96 with minor modifications [17]. Each film sealed a 0.00181 m² aperture on a permeation cell held at 20 °C with 75% RH gradient (0% inside with anhydrous CaCl₂; 75% outside with saturated NaCl). Mass gain under steady state after ~2 h was recorded at eight time points. The driving force included 1,753.55 Pa.

2.11. Statistical analysis

           All experiments were carried out in triplicate and reported as means. Statistical analyses were carried out using SPSS v.10.0 for Windows (IBM, USA).

1. Results and Discussion

3.1. Crude exopolysaccharides and films

Alcohol‑precipitated EPS achieved from W. confusa NH02 are sometimes labeled as dextran, when composition is not explicitly resolved. In practice, the precipitate can include multiple glucan-type and hetero-polysaccharide fractions depending on cell morphology during culture and the specific fermentation regime [18]. In the present study, pooled alcohol‑precipitated material from three independent fermentations comprised ~65% glucan and its average molecular mass exceeded 1.13 × 10⁶ Da using HPLC analysis [4]. Films cast from the melanin-deficient W. confusa NH02 EPS were optically clear and colorless (Fig. 1), water‑soluble and resisted cracking on folding. Introducing glycerol did not visibly alter color/clarity or qualitative solubility; however, high plasticizer loadings (> 50 g per 100 g EPS) yielded tacky films that adhered strongly to Petri dishes and were difficult to remove intact. The EPS level in the film‑forming dispersion was selected to balance transparency, flexibility and ease of plate release. As expected, film thickness increased with increasing polysaccharide concentration.

3.2. Rheology of filmogenic dispersions

The EPS film-forming dispersion showed non-Newtonian pseudoplastic behavior within the assessed shear range, as verified by fitting to Ostwald-de Waele model (Table 1). The flow behavior index (n = 0.42) indicated pronounced shear-thinning characteristics, while the consistency index (k = 3.18 Pa·sⁿ) reflected a structured polymer network with substantial intermolecular interactions. Apparent viscosity decreased with increasing shear rate, reaching 0.84 Pa·s at 300 s⁻¹ and suggesting appropriate flow characteristics for casting. This shear-thinning behavior was attributed to progressive alignment and disentanglement of dextran chains under shear, which decreased flow resistance while preserving structural integrity at rest. The relatively high k value indicated strong hydrogen bonding and chain entanglement within α (1→6) dextran backbone, promoting cohesive film formation. Such rheological characteristics facilitated uniform spreading and stabilization of the filmogenic matrix, while the presence of a structured network in dispersion form was consistent with the enhanced tensile strength and flexibility observed in the resulting films, as improved intermolecular interactions and chain packing enabled efficient stress distribution and mechanical resilience. Similar relationships between shear-thinning behavior, polymer network structure and improved mechanical performance have been reported for polysaccharide-based film systems [19–20].

3.3. Water‑vapor permeability

The WVP response showed a non‑monotonic trend with plasticizer content as WVP decreased as glycerol was increased from 0 to ~35% (w/w). However, this re-increased at higher fractions, with a maximum of approximately 3.36 observed at 50% glycerol (Table 2). At moderate loadings, extensive hydrogen bonding between EPS chains and glycerol likely restricted water‑transport pathways; thereby, decreasing WVP [21]. In addition to an EPS:glycerol ratio of 60:40, the permeability increased sharply, which could be rationalized by heterogeneous plasticizer distribution and the greater hydrophilicity of glycerol linked to EPS that together facilitated water diffusion through the matrix. In humid environments, minimizing WVP is a key performance target for practical packaging films [22]. In parallel, water contact angle decreased from 94.25 (0% glycerol) to 65.82° (50% glycerol), indicating increased surface hydrophilicity with plasticization (Table 2).

This behavior could be explained by plasticizer-polymer-water interactions and their effects on film microstructure. At moderate glycerol contents (up to ~35%), glycerol promoted intermolecular hydrogen bonding and closer chain packing, decreasing microvoids and limiting water-vapor diffusion. In contrast, excess glycerol disrupted polymer-polymer interactions, increased free volume and enhanced matrix hydrophilicity, promoting moisture sorption and transport. Similar non-monotonic WVP trends associated with plasticizer-induced structural rearrangements have been reported for polysaccharide films [23–24].

3.4. Mechanical characteristics

Literature reports show broad ranges for hydrocolloid‑film mechanics due to differences in biopolymer structure/composition and processing conditions [25–27]. For the EPS/glycerol plasticized films, tensile strength (TS) increased initially with glycerol addition, reached a maximum of 5.56 MPa and then decreased as plasticizer content increased (Table 2). The early strengthening could be attributed to favorable inter-polymer hydrogen bonding in EPS and glycerol that improves network integrity up to an optimal composition of EPS:glycerol of 65:35, whereas excessive plasticization over this point disrupted interchain interactions and introduced free volume that weakened the film [28].

As the available hydroxyl content in the filmogenic solution increased with higher polysaccharide/plasticizer proportions, hydrogen‑bond density and distribution changed as well [29]; similar to the observed turning point in TS. Decreased TS has been linked to phase‑separation‑like behavior in associated systems, when intra‑polymer interactions dominate over inter‑polymer ones [30–31]. Elongation at break (E) followed TS trends, with the an EPS:glycerol (65:35) formulation showing the largest extensibility (142.51%). At small to moderate plasticizer contents, glycerol enhanced ductility without a penalty in TS; similar behavior has been reported for kefiran‑based plasticized films [32].

           The dextran-based film developed in this study demonstrated tensile strength (5.56 MPa) and elongation at break (142%), exceeding typical values reported for bacterial α-glucan films (TS < 3 MPa, E < 60%). This enhanced performance could be attributed to the structural characteristics of dextran produced by W. confusa, whose α(1→6) backbone with α(1→3) branches promoted chain entanglement and extensive hydrogen bonding; thereby, improving film cohesion. Moreover, the optimized glycerol content likely provided effective plasticization by increasing polymer chain mobility while preserving matrix integrity. Controlled drying conditions might promote dense polymer packing and decreased microstructural defects. Similar improvements in strength-flexibility balance due to plasticizer optimization and intermolecular interactions have been reported for polysaccharide-based films in recent studies [23, 33]. Collectively, these factors explain the superior mechanical performance.

3.5. Dynamic mechanical thermal analysis

The DMA/DMTA is widely used to probe viscoelastic transitions in glycerol‑containing biopolymer films [34–36]. For EPS/glycerol blends, the E′ and tan δ curves showed single relaxation associated to the glass transition; Tg decreased significantly with increasing glycerol content, shifting from approximately 85 °C in the unplasticized EPS film to less than -15 °C at 50% (w/w) glycerol, as evidenced by the progressive left shift of the tan δ peak and the storage modulus drop (Fig. 2). This pronounced depression of Tg demonstrated the high plasticizing efficiency of glycerol and the increased mobility of EPS chains. Changes in Tg reflected molecular‑level compatibility and the evolving polymer network; EPS functional groups could form interpenetrating networks with the companion phase, modulating chain mobility; similar to classical structure-characteristics relationships in amorphous polymers [37]. The presence of one tan δ peak and a single‑step drop in E′, without additional α‑relaxations or separate melting events, indicated no detectable phase separation; single Tg in DMTA commonly signifiesd strong component compatibility [38].

1. Conclusion

This study provides the first evidence that EPS recovered from LAB W. confusa NH02 (from Nham) can be processed into continuous films and systematically maps how glycerol tuning governs physical, mechanical and barrier attributes. With increasing glycerol, WVP decreased from 0 to ~35% glycerol and then increased at higher fractions. Mechanical performance peaked nearly EPS:glycerol (65:35) (TS 5.56 MPa; E 142.51%), after which additional plasticizer decreased strength and extensibility. The DMTA revealed single glass transition in compositions, supporting good compatibility of the EPS and glycerol phases. Furthermore, Tg decreased dramatically with increasing glycerol content, shifting from approximately 85 °C in the unplasticized EPS film to les than -15 °C at 50% (w/w) glycerol, as indicated by the left shift of the tan δ peak and the corresponding drop in storage modulus. This strong depression of Tg verified the high plasticizing efficiency of glycerol and enhanced chain mobility within the EPS matrix. In conclusion, plasticized EPS films offer a promising platform for food-contact uses, where a balanced combination of strength, flexibility and moisture control is needed.

Background and Objective: Traditional cheese-making practices often rely on locally adapted fermentation strategies that contribute to product identity and sensory typicity. This study investigated the effects of commercial and natural starter cultures on the physicochemical, microbiological and sensory characteristics of a locally adapted Edam-type cheese produced under semi-traditional conditions in Algeria.

Material and Methods: Three cheese formulations were prepared, using (i) a combination of mesophilic and thermophilic lactic acid bacteria with red smear cultures, (ii) whey from a previous batch used as a natural starter and (iii) mesophilic bacteria with red smear cultures without thermophilic strains. Pasteurized cow milk used for cheese-making met standard physicochemical and microbiological quality requirements.

Results and Conclusion: The resulting cheeses showed significant differences (p < 0.05) in pH (4.8–5.7), moisture content (33.82–38.1%) and yield, depending on the starter culture used. From the three formulations, cheeses produced without thermophilic bacteria showed the highest yield (14.56%). Microbiological analyses verified the absence of Escherichia coli, Staphylococcus aureus and Salmonella in all samples, ensuring product safety. Sensory evaluation showed that cheeses produced with combined mesophilic, thermophilic and red smear cultures included the most balanced texture and flavor, whereas whey-inoculated cheese included characteristics associated with traditional fermented dairy products. These results highlight how starter culture selection and whey reuse shape the sensory identity and technological performance of ethnic Edam-type cheeses, supporting the sustainable valorization of whey within Algerian traditional dairy systems.

Keywords: Edam-type cheese, lactic acid bacteria, natural whey starter, sensory quality, traditional cheese.

1. Introduction

Cheese production is one of the most important pathways for milk valorization, providing a wide range of products with distinct technological and sensory characteristics. Cheese is a complex food matrix produced through the coagulation of milk proteins, by either enzymatic action (rennet) or acidification, followed by whey drainage and maturation. The diversity of cheese varieties is originated from variations in milk composition, coagulation processes, microbial communities and aging conditions [1]. The cheese industry aims to transform milk into a long‑lasting, flavorful product through microbial and enzymatic activities, which contribute to its distinctive texture and sensory characteristics.

Traditional cheese production is deeply rooted in agricultural heritage. In Algeria, traditional cheeses are an integral part of the country's dairy culture, reflecting local knowledge and unique processing techniques [2,3]. From these, Edam cheese (a semi-hard-pressed variety) is widely recognized for its characteristic texture and flavor, aligning with international cheese production standards [1].

The popularity of cheese within consumers stems from its desirable sensory attributes, high nutritional value and versatile uses in the food industry [4]. Regarding the critical role of microbial communities in defining cheese quality, it is essential to assess the physicochemical, microbiological and sensory characteristics of cheese to ensure optimal production and consumer satisfaction [5]. Sensory analysis is still an important tool for assessing cheese acceptability and quality [6]. Joudu et al. [6] demonstrated that ripening time leads to important changes in the texture and structure of Edam-type cheese. Mesophilic lactococci are mainly associated to early acidification and flavor development. Thermophilic starters such as Streptococcus thermophilus support rapid lactose fermentation and smear bacteria such as Brevibacterium linens contribute to surface-ripening aroma development [7-9]. In traditional Algerian dairy systems, whey reused as a natural starter introduces an indigenous mixed microbiota that may reinforce local sensory typicity while increasing process variability [2,7]. Despite the widespread production of Edam-type cheese, there is still limited research on how various starter cultures affect the quality of traditional pressed Edam-type cheese under locally adapted processing conditions. Therefore, this study aimed to assess how commercial and traditional starter cultures, including whey as a natural starter, affect the physicochemical, microbiological and sensory qualities of a locally adapted Edam-type cheese produced under semi-traditional conditions.

1. Materials and Methods

2.1. Cheese manufacture

The cheese‑making process was inspired by locally adapted artisanal practices commonly used in traditional dairy production, while maintaining controlled hygienic and technological conditions.

2.2. Starter cultures and formulations

Edam-type cheeses were prepared from whole cow milk standardized to 3.5% fat and pasteurized using LTLT, 63 °C for 30 min, then cooled to 32 °C before inoculation. Three Edam-type cheese formulations were assessed in triplicate.

2.2.1. Formulation 1: Commercial mixed starter and smear

Milk was inoculated with a mesophilic starter (Flora Danica; Lactococcus lactis ssp. cremoris, L. lactis ssp. lactis, L. lactis ssp. diacetylactis and Leuconostoc mesenteroides spp. cremoris; Chr. Hansen, Denmark), a thermophilic starter containing S. thermophilus (TA 40 Lyo 50 DCU, Yo-mix Danisco, France) and a red smear culture of B. linens (Chr. Hansen, Denmark), used as freeze-dried commercial cultures based on the manufacturers' recommendations.

2.2.2. Formulation 2: Natural whey starter (short-term storage)

Whey used as a natural starter was collected at the end of the previous batch immediately after whey drainage. The whey was filtered through sterile gauze to remove curd particles, then cooled and stored at 4 °C for 2–4 h (time needed to carry out quality control analyses on the milk). Prior to inoculation, whey physicochemical parameters were assessed as pH = 5.6 ±0.1 and titratable acidity = 18 °D ±2 (approximately 0.18% lactic acid). No preinoculation microbiological enumeration or species-level identification of the whey starter was carry out in the original experiment; accordingly, the natural starter was characterized by its short-term handling conditions and physicochemical screening only. The whey starter was added to pasteurized milk at a rate of 3% (v/v), corresponding to 30 ml per liter of milk. After whey addition, the milk entered the common 20 min activation step at 32 °C before rennet addition.

2.2.3. Formulation 3: Mesophilic starter and smear (without thermophilic strains)

Milk was inoculated with the mesophilic starter (Flora Danica, Chr. Hansen, Denmark) and the red smear culture (B. linens, Chr. Hansen, Denmark) without thermophilic strains, using freeze-dried commercial cultures based on the manufacturers’ recommendations.

2.2.4. Common processing steps

For all three formulations, after starter inoculation the milk was held at 32 °C for 20 min to allow starter activation. Calcium chloride (0.25 ml l^-1) and rennet (CHY-MAX, Chr. Hansen, Denmark) were then added to the formula and coagulation proceeded for 30 min at 32 °C. The curd was cut into approximately 1 cm cubes and primary whey was drained for 10 min. Scalding was carry out by gradually increasing the temperature to 38–40 °C over 30 min with gentle stirring. The curd was pressed at room temperature (RT) for 2 h (first pressing), followed by moulding and final pressing overnight (12–16 h) under progressively increasing weight to achieve appropriate consolidation and moisture removal. Cheese blocks were brined in 20% (w/w) brine at 13 °C for 24 h, coated with red wax, ripened for 6 w at 10–13 °C (80–85% RH), then stored at 4–6 °C. A 6-w ripening time was selected as a common early-ripening endpoint for comparing formulations under identical conditions and did not represent full Edam maturation. The cheese-making workflow followed standard semi-hard cheese practice while was consistent with locally adapted artisanal processing conditions [2,10].

2.3. Physicochemical analysis of milk

Milk physicochemical quality was assessed using routine dairy analytical methods [10]. Milk pH was assessed using pH meter. Titratable acidity was investigated using Dornic method and 0.111 N sodium hydroxide and expressed as degrees Dornic. Density was assessed using pycnometer, refractive index was assessed using refractometer, ash content after incineration at 500 °C and total solids were assessed using oven drying at 105 °C to constant mass.

2.4. Microbiological analysis of milk

Microbiological analyses were carried out on milk samples after preparing decimal dilutions in sterile physiological saline (0.85% NaCl). Aerobic mesophilic microorganisms at 30 °C were enumerated on plate count agar based on ISO 4833-1:2013. Thermotolerant coliforms were enumerated on violet red bile lactose agar based on ISO 4832:2006. Staphylococcus aureus was detected on Baird-Parker agar supplemented with egg yolk and potassium tellurite based on ISO 6888-1:2021. Salmonella detection followed ISO 6579-1:2017 after pre-enrichment, selective enrichment and isolation on Hektoen agar. When no colonies were observed on plates, milk results were reported as < 10 CFU ml^-1, based on the plated volume and dilution factor.

2.5. Physicochemical analysis of cheese

The pH, dry matter content, ash content, titratable acidity, density, refractive index and fat content of cheese were investigated in triplicate using standard dairy analytical methods [10]. Instrumental texture, objective color, porosity, protein content and shelf-life assessments were not included in the present study and were acknowledged as limitations.

2.6. Microbiological analysis of cheese

The methods for detecting Escherichia coli, S. aureus and Salmonella in cheese were similar to those used for milk after homogenization of 10 g of cheese in 90 ml of sterile physiological saline (0.85% NaCl). For cheese samples, results were reported as < 100 CFU g^-1 when no colonies were observed under the analytical conditions.

2.7. Organoleptic analysis

Sensory evaluation was carried out by a panel of 20 trained evaluators (ten males and ten females, aged 25–55 y) familiar with dairy product assessment and experienced in cheese sensory analysis. Cheese samples from each formulation were coded with random three-digit numbers and presented to panelists in a randomized order to minimize order effects. Samples (10–15 g cubes) were served at 10–12 °C under white light in individual booths and unsalted crackers and RT water were provided for palate cleansing between samples [11]. The sensory assessment was handled as a descriptive profile rather than numerical hedonic test. Panelists recorded predefined descriptors covering color, appearance, texture, odor and taste/overall appreciation for each formulation [11]. Because the available sensory dataset consisted of descriptive frequencies rather than numerical hedonic scores, sensory differences were interpreted comparatively and qualitatively.

2.8. Statistical analysis

Each formulation was produced in triplicate. One-way ANOVA was used to assess differences between the three cheese formulations, followed by Tukey's honestly significant difference (HSD) post-hoc test when ANOVA indicated a statistically significant result. P-values less than 0.05 were considered statistically significant. Statistical analyses were carry out using R software.

1. Results and Discussion

3.1. Physicochemical analysis of milk

The physicochemical parameters of the pasteurized milk used in this study are summarized in Table 1. The pH value (6.4) verified that the milk complied with the standard pH range of fresh milk (6.0–7.0) as recommended by [12]. The titratable acidity of the milk was recorded as 17 °D, which falls within the standard range (14–18 °D) and the milk density (1.030) was within the acceptable range (1.028–1.035), verifying a good balance between fat and solid contents [13]. The ash content (0.735%) was within the standard range (0.7–0.8%), reflecting the mineral composition of the milk. The total solid extract assessed as 12.9% was within the normal range (10.2–13%) [11], indicating good milk quality. The refractive index of the milk was investigated as 1.348, a value affected by the fat and protein compositions [10].

3.2. Microbiological analysis of milk

Microbiological analysis results (Table 2) indicated that aerobic mesophilic microorganisms, thermotolerant coliforms, S. aureus and Salmonella were all less than 10 CFU ml^-1 or absent in 25 ml in the pasteurized milk sample. These findings indicated good hygienic quality, in compliance with [2] and verified that pasteurization was effective in limiting potential contaminants [7].

3.4. Brining time and cheese yields

The brining time varied depending on the type of starter. The cheeses fermented with Formulation 1 including a brining time of 5 h and 37 min (±3 min), while the cheeses prepared with Formulation 2 needed 6 h (±5 min) and the cheeses made with Formulation 3 needed 7 h and 30 min (±6 min) (Table 3). Statistical analysis revealed a significant difference in yield between the various formulations (p < 0.05) (Fig. 1). The highest yield (14.56% ±0.06) was recorded for Formulation 3, followed by Formulation 2 (12.06% ±0.04) and Formulation 1 (11.24% ±0.06), verifying that bacterial composition affects cheese yield [14,15]. Mayo et al. [16] demonstrated that milk coagulation characteristics significantly affected cheese-making efficiency. In this study, since a similar milk was used for all formulations, the observed variations in yield could be attributed mainly to the effect of the bacterial starter cultures.

3.5. Physicochemical characteristics of cheeses

Table 4 presents the physicochemical parameters of the three cheese formulations. The pH values ranged from 4.8 ±0.04 (Formulation 2) to 5.7 ±0.06 (Formulation 1), with a significant decrease in Formulation 2, due to higher lactic acid production. Lower pH was associated to enhanced bacterial metabolism during ripening [17]. This was consistent with the ripening-linked physicochemical changes reported by Joudu et al. [6] for Edam-type cheeses. Titratable acidity ranged between 18 °D ±0.02 and 20.7 °D ±0.03, which was appropriate for cheese maturation [8]. The dry matter content varied from 61.90% ±0.5 (Formulation 2) to 66.18% ±0.4 (Formulation 3), affecting cheese texture and firmness. The moisture content was inversely correlated to the dry matter percentage, with Formulation 2 including the highest moisture (38.1% ±0.03) and Formulation 3 including the lowest moisture (33.82% ±0.05). The variations in rheological characteristics observed in Edam-type cheeses during ripening were consistent with those of Joudu et al. [6], who analyzed changes in these characteristics during the maturation of Edam-type cheeses. The present differences should therefore be interpreted as 6-w ripening outcomes rather than full Edam maturation characteristics. The fat content was highest in Formulation 1 (20% ±0.7) and lowest in Formulation 3 (15.9% ±0.3), indicating that the type of starter affects lipid retention during cheese formation [18,19]. The ash content was highest in Formulation 1 (2.40% ±0.04) and lowest in Formulation 3 (1.60% ±0.01), suggesting variations in mineral retention. The refractive index was consistent in all cheese samples (1.335 ±0.001 to 1.336 ±0.002), indicating similar compositional characteristics [20]. In addition to their technological effects, the observed differences in cheese formulations revealed the effects of fermentation strategy on the typicity of the final product. Cheeses produced using whey as a natural starter showed characteristics commonly associated with traditional fermented dairy products, including enhanced acidity and softer texture. Such attributes were often perceived as markers of authenticity in ethnic and artisanal foods. In contrast, the use of commercial starter cultures resulted in greater standardization and consistency, highlighting the balance between technological control and traditional identity in locally adapted cheese production.

3.6. Microbiological quality of cheese

Microbiological analysis results verified that E. coli, S. aureus and Salmonella were less than the limit of detection (< 100 CFU g^-1) in all cheese samples from the three formulations, meeting microbiological safety standards [10]. No colonies characteristic of these pathogens were detected on selective media under the conditions specified in the Methods section. The absence of pathogenic bacteria was attributed to good manufacturing practices, effective pasteurization and the antimicrobial activity of lactic acid bacteria (LAB), which produce organic acids and bacteriocins that inhibit microbial growth [11,21].

3.7. Organoleptic assessment

The sensory evaluation (Fig. 2) showed that cheese made with Formulation 1 included the highest organoleptic quality. The panelists associated this formulation most often with light-yellow color, further hydrating appearance, smooth texture, cheesy odor and the most favorable overall taste profile, compared with Formulations 2 and 3. Formulation 2, produced with the natural whey starter, included hydrating appearance and soft texture but was further frequently associated with higher perceived acidity and less balanced odor profile. Formulation 3, which lacked thermophilic bacteria, showed the driest appearance and lower overall sensory appreciation, with a less pronounced aroma. The results suggested that the combination of mesophilic, thermophilic and red smear bacteria contributed to superior cheese quality, a pattern that was broadly consistent with the sensory differentiation reported by Ercan et al. [5] and technological interpretation frameworks discussed by Possas et al. [18] as well as Thomareis and Dimitreli [19]. The use of a natural whey starter therefore was a feasible traditional alternative, but its sensory expression appeared less standardized than that achieved with the mixed commercial cultures.

1. Conclusion

This study investigated the effects of various bacterial starters on the production of semi-cooked pressed Edam-type cheeses and verified that starter culture composition significantly affected cheese yield, physicochemical characteristics and sensory attributes. Physicochemical and microbiological analyses verified the high quality of raw material and final product, as all assessed parameters were within acceptable standards. Sensory evaluation revealed significant differences between the three cheese types, with the best organoleptic quality observed in Edam-types 1, which contained mesophilic, thermophilic and red smear bacteria.

The findings suggested that bacterial starter composition played a critical role in assessing cheese texture, taste and overall quality. The use of whey as a natural starter (Formulation 2) provided an alternative with acceptable characteristics, while the absence of thermophilic bacteria (Formulation 3) negatively affected sensory characteristics. For local producers, these results illustrated the balance between process standardization and preservation of traditional sensory identity.

Future research should investigate the effects of aging duration, temperature variations, various bacterial cultures, proteolysis, lipolysis, whey microbial profiling and objective color, texture, porosity, protein-content and shelf-life assessments to further optimize cheese production techniques and improve product consistency. Because ripening in the present study was limited to 6 w, longer maturation times should be assessed.

From a broader perspective, this study demonstrates that the integration of traditional fermentation practices into the production of locally adapted Edam-type cheese can enhance sensory identity while maintaining microbiological safety and acceptable physicochemical quality. The valorization of whey as a natural starter contributes to the traditional knowledge and supports sustainable dairy production systems. These findings highlight the potential of combining traditional practices with controlled processing to promote culturally significant dairy products within modern food systems.

1. Declaration

5.1. Acknowledgements

This report was the result of an applied research project on traditional and industrial cheese productions carried out at University of Science and Technology of Oran - Mohamed Boudiaf. The authors thank the Laboratory of Applied Microbiology, Faculty of Natural and Life Sciences, University of Oran1 Ahmed Ben Bella, for providing laboratory and pilot-scale cheese-making facilities.

5.2. Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable requests.

5.3. Funding

This research received no specific grant from any funding agency in the public, commercial or not‑for‑profit sector.

5.4. Financial interests

The authors declare no financial interests.

5.5. Declaration of competing interest

The authors have no competing interests to declare that are relevant to the content of this article.

5.6. Authors’ Contributions

Nebia Zebboudj: conceptualization, methodology, cheese production, data curation, writing – original draft. Hanane Fatma Chentouf: supervision, methodology, validation, writing – review and editing. Wassim Yezli: formal analysis, visualization, statistical analysis, writing – review and editing; corresponding author. Abdelatif Boudra: supervision, resources, project administration, writing – review and editing.

5.7. Using Artificial Intelligent Chatbots

AI language models were used only for language editing and proofreading. The authors carefully checked all AI-assisted text and remain fully responsible for the scientific content and conclusions.

5.8. Ethical Consideration

This article did not contain any studies with human participants or animals carry out by any of the authors. Therefore, formal ethics committee approval was not needed.