Exploring the molecular interaction of bioactive peptides with antioxidative, antidiabetic and anti-inflammatory attributes from sheep milk fermented with lactobacillus and yeast

Abstract

Sheep milk is a rich source of proteins with potential to generate bioactive peptides through microbial fermentation. However, limited studies have focused on the Panchali sheep breed of India and the multifunctional health properties of its fermented products. This study investigated the generation of bioactive peptides with antidiabetic, antioxidative, and anti-inflammatory properties through the fermentation of sheep milk using a potent LAB, Limosilactobacillus fermentum (KGL4, MTCC 25515), in combination with the yeast strain Saccharomyces cerevisiae (WBS2A, MG101828). Optimal bioactivity was observed after 48 h of fermentation at 30 °C, with antioxidant activity reaching 40.08%, α-amylase inhibition at 72.48%, and α-glucosidase inhibition at 75.96%. Proteolytic activity was assessed at varying inoculum concentrations (1.5, 2 and 2.5%) and incubation time (0, 12, 24, 36 and 48 h), with the highest activity (6.74 mg/mL) observed at a 2.5% inoculum after 48 h. Additionally, fermented milk effectively lowered the production of pro-inflammatory cytotoxins and ROS in LPS-induced RAW 267.4 macrophage cells, indicating its anti-inflammatory properties. Protein profiling through SDS-PAGE revealed bands between 15 and 70 kDa, while 2D electrophoresis identified 39 distinct protein spots associated with peptide production. Reverse-phase HPLC of ultrafiltered water-soluble extracts (WSEs) was used to separate peptide fractions, and identified sequences were assessed through comparison with BIOPEP database to confirm antidiabetic and antioxidative activities. The most potent selected peptide showed strong binding affinity and stable interaction within the active pockets of human α-amylase and α-glucosidase, as well as high HADDOCK and HPEPDOCK scores in molecular docking analyses. These findings indicate that fermented sheep milk is a promising source of multifunctional bioactive peptides with potential applications in functional foods targeting diabetes and inflammation. Further in vivo studies are warranted to validate physiological relevance, bioavailability, and health benefits.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-30657-3.

Introduction

Milk and dairy products are vital components of the human diet due to their high-quality proteins, lactose, essential minerals (calcium, magnesium, potassium), and vitamins (A, B-complex, C, and D)^1,2. In addition, milk contains bioactive compounds such as caseins, whey proteins, and fatty acids that contribute to human health and metabolic functions³. Globally, sheep milk contributes a significant share to small ruminant milk production, with major producers including China, Turkey, and Greece⁴. Its composition varies depending on factors such as breed, lactation stage, diet, season, and udder health, which influence levels of proteins, fats, minerals, and vitamins⁵.

Sheep milk is notably high in protein, typically ranging from 45 to 66 g/kg of total protein, with casein making up approximately 42 to 52 g/kg. Among the casein fractions, β-casein predominated, accounting for approximately 61.60%, followed by αs₂-casein at 22.84%, κ-casein at 8.90%, and a smaller proportion of αs₁-casein at 6.66%. Several factors, such as breed, stage of lactation, nutritional intake, environment, number of previous births (parity), seasonal variation, and udder health, significantly influence protein content, which can differ widely across species⁶. Sheep milk is rich in proline, a key amino acid that participates in numerous physiological functions, including protein assembly, collagen synthesis and hemoglobin formation. It is crucial for promoting fetal development and maintaining optimal growth during the neonatal period. Additionally, proline contributes to immune system function by assisting in the maturation of T-lymphocytes. Research also highlights its neuroprotective potential, as it helps reduce the production of amyloid-beta (Aβ) a protein linked to Alzheimer’s disease thereby lessening its harmful impact on nerve cells⁷.

Lactic acid bacteria (LAB) are non-endospore-forming, gram-positive microorganisms that are typically present in human gut flora, where they play a crucial role in fostering and improving gastrointestinal health. LAB have long been utilized in food preservation through traditional fermentation processes, one of the earliest forms of biotechnology aimed at prolonging the shelf life of foods⁸. Beyond preservation, fermentation improves the sensory attributes of food and enhances its nutritional profile⁹. Advances in microbial and processing technologies have improved fermentation efficiency, product consistency, and industrial scalability, supporting cost-effective and sustainable production of clean-label fermented foods¹⁰.

During fermentation, microorganisms such as bacteria and yeast are grown in nutrient-rich environments fortified with protein-derived substrates. These microbes release proteolytic enzymes that degrade proteins, resulting in the release of bioactive peptides (BPs). Factors like, Substrate concentration, fermentation time, number of peptides formed, and the specific microbial strains used influence the extent of protein hydrolysis. Lactobacillus species are commonly employed in this process¹¹. The combination of lactic acid bacteria and yeast in co-fermentation offers synergistic proteolytic activity, resulting in enhanced release of bioactive peptides compared to single-strain fermentation. Yeast metabolites support LAB growth and activity, improving overall fermentation efficiency and functional properties of the final product. Bioactive peptides derived from sheep milk are recognized as a good source of antidiabetic and antioxidative properties, making them valuable for functional food applications.

During typical cellular and metabolic activities, the body naturally provides unstable reactive molecules, including reactive nitrogen species as well as reactive oxygen species. These are generally neutralized by the body’s inherent antioxidant defense mechanisms. However, under certain pathological conditions or upon exposure to harmful substances, either the production of free radicals increases abnormally, or the antioxidant defenses become compromised. This disruption leads to a condition known as oxidative stress, which is defined as an imbalance between the formation of free radicals and the capacity of the antioxidant system to counteract the formation of free radicals. Oxidative burden leads to structural and functional damage to cellular macromolecules, including lipids, proteins, and DNA, and disrupts intracellular signaling pathways. Oxidative lesions, strand breakage, and mutations are all kinds of DNA damage that can cause chronic illnesses¹².

The global incidence of diabetes has significantly increased in recent years. According to the 2022 statistics from the IDF¹³, approximately 537 million individuals worldwide are currently affected by the disease. Additionally, approximately 316 million individuals experience impaired glucose tolerance, placing them at an elevated risk of developing the disease. These numbers are expected to increase, with projections indicating that the diabetic population may reach 643 million by 2030. If effective and timely preventive strategies are not implemented, this number could increase to 783 million within the next 25 years. Bioactive peptides present in functional foods may regulate glucose absorption and insulin secretion¹⁴. Nearly 90% of all diabetes cases are caused by type 2 diabetes mellitus¹⁵. The incretin hormones Glucose-Dependent Insulinotropic Polypeptide (GIP) and Glucagon-Like Peptide-1 (GLP-1) stimulate pancreatic beta cells following the consumption of carbohydrate-rich diets, resulting in an increase in insulin secretion¹⁶. These hormones mediate biological functions through interactions with specific receptors. In individuals with type 2 diabetes mellitus, GLP-1 exhibits both insulin-promoting (insulinotropic) and glucagon-suppressing (glucagonostatic) actions, which contribute to the maintenance of normal blood glucose levels¹⁷.

The novelty of this study lies in the exploration of fermented Panchali sheep milk from India as a previously uncharacterized source of multifunctional bioactive peptides. Additionally, the application of a LAB with yeast fermentation strategy to enhance antidiabetic, antioxidative, and anti-inflammatory properties has not been reported earlier.

Although sheep milk from European and Middle Eastern regions has been investigated for functional peptide production, limited scientific data are available on the Panchali breed of Gujarat, India, despite its commercial and nutritional value in the local dairy sector. Moreover, the combined use of LAB and yeast for enhanced proteolysis and multifunctional peptide release has not yet been explored.

Given the scarcity of scientific studies on the Panchali sheep breed native to Gujarat, this study focused on exploring the biofunctional potential of its milk during fermentation. This study evaluated the enhancement of antioxidant and antidiabetic activities through fermentation with Lactobacillus and yeast strains. This process involves the extraction and identification of the peptides responsible for these bioactivities. In addition, the anti-inflammatory properties of milk fermentation were examined using RAW 267.4 macrophage cells. Advanced tools such as FTIR spectroscopy and CLSM were used to confirm the results.

Materials and methodology

Strain

The LAB strain Limosilactobacillus fermentum (KGL4, MTCC 25515) and the yeast strain Saccharomyces cerevisiae (WBS2A, MG101828) were procured from the Department of Dairy Microbiology, SMC College of Dairy Science, Anand, India.

Preparation of sample

Sheep milk (Panchali breed) was collected under hygienic conditions, from a small-scale, non-commercial source located in Anand, Gujarat. Raw sheep milk was first filtered through a clean muslin cloth to remove impurities and then pasteurized (85 °C/10 min). After heat treatment, the samples were stored in sterile glass vessels at a refrigeration temperature of 5 ± 1 °C until further use in experimental procedures.

To activate the pure cultures, a 2% inoculum was introduced into heat-treated sheep milk and incubated under specific conditions: Lactic Acid Bacteria (LAB) were cultured at 30 °C for 24 h, while the yeast strain was incubated (25 °C for 3–5 day). For combined fermentation, heat-treated sheep milk was co-inoculated with both Lactobacillus and yeast strains at a 2% level. The mixtures were incubated at 30 °C for 0, 12, 24, 36, and 48 h. The supernatants were filtered through 0.22 μm syringe filters after the fermented sheep milk (FSM) was centrifuged at 4193×g for 30 min at 4 °C. Subsequently, antioxidant and antidiabetic activities were conducted.

Antioxidative activity evaluation (ABTS method)

ABTS activity was measured according to the method described by Sah et al.¹⁸.

Evaluation antidiabetic properties

α-Amylase inhibition

Inhibition was assessed by incubating the culture supernatant with α-amylase and starch, followed by the addition of DNSA reagent, heating, and measuring absorbance at 540 nm, as described by^19,20.

α-Glucosidase inhibition

The inhibition of α-Glucosidase was measured by reacting the supernatant of milk with the enzyme and P-NPG substrate, followed by the addition of Na₂CO₃ and absorbance reading at 405 nm^21,22.

Proteolysis evaluation

Proteolytic activity was measured using the OPA assay, following the protocol outlined by Hati et al.²³. Heat-treated sheep milk was inoculated at varying concentrations such as 1.5%, 2.0%, and 2.5% and incubated across a range of time intervals (0, 12, 24, 36 & 48 h) for evaluate extent of proteolysis.

Purification and characterization of peptides exhibiting antioxidant and antidiabetic properties

Growth conditions were optimized by inoculating the strain at an inoculum level (2.5%) and incubated (48 h /30°C). Peptide concentration was determined using the OPA assay²⁴. Optimized fermentation with 2.5% inoculum at 30 °C for 48 h was followed by centrifugation (4193 × g, 30 min) to obtain WSEs as per standard protocol. This WSEs were then subjected to ultrafiltration (3 and 10 kDa) to separate different peptide fractions. This clarified WSE was then used for subsequent ultrafiltration and peptide purification analyses.

SDS-PAGE evaluation

The molecular weights of the protein fragments were determined following the protocols described by Carrasco-Castilla et al.²⁵ and Laemmli²⁶. A 12% resolving gel was used for electrophoretic separation, utilizing water-soluble extracts (WSE) derived from fermented sheep milk samples. Optimized fermentation with 2.5% inoculum at 30 °C for 48 h was followed by centrifugation (4193 × g, 30 min) to obtain WSEs as per standard protocol. This WSEs were then subjected to ultrafiltration (3 & 10 kDa) to separate different peptide fractions. This clarified WSE was then used for subsequent ultrafiltration and peptide purification analyses.

2D gel electrophoresis

Two-dimensional (2D) gel electrophoresis was performed to resolve peptide spots from sheep milk WSEs following the procedure described by Yang et al.²⁷; the electrical parameters were adjusted as specified by Panchal et al.²⁸.

Isoelectric focusing

For a 7 cm ready immobilized pH gradient (IPG) strip (pH 3–10) (Bio-Rad), 125 µg of protein from fermented sheep milk and camel milk sample was utilized. Panchal et al.²⁸ provide the isoelectric focusing electrical limits that are used for isoelectric focusing. A rehydration buffer was used to dilute the WSEs of fermented sheep and camel milk, resulting in a final volume of 130 µL. The strip was stored at 25 °C for the entire night in a rehydration buffer. After isoelectric focusing, the strip was equilibrated in Buffer I for 10 min followed by Buffer II for another 10 min. It was then briefly rinsed with 1× Tris–Glycine–SDS buffer and placed onto a 10% separating gel for the second-dimension SDS-PAGE.

Peptide identification and structural analysis using RPLC/MS

Liquid chromatography

The liquid chromatography (LC) was conducted using an Adhoc column (2.1 × 100 mm, 1.5 μm), with sample and column maintained at 20 °C and 40 °C, respectively. A 20 µL volume of 0.22 μm-filtered, trypsin-digested peptides was injected and separated at 0.3 mL/min using a solution of 0.1% formic acid in acetonitrile and water while the mobile phase B was only water, as described by Khakhariya et al.²⁹. The EksigentEkspert ultra-LC 100 system was used in conjunction with an AB SCIEX QTRAP 4500 with ESI to conduct mass spectrometric analysis. Peptides from fermented sheep milk were identified using RPLC-MS with EMS and EPI scanning (350–2000 Da for EMS; 100–2000 Da for EPI) under optimized DP (80 V), EP (10 V), and spray voltage (5500 V) using rolling collision energy. IDA was applied to target top 1–3 precursor ions (250 mDa window), and ER scans assessed isotopic distributions.

Analysis of data and peptide identification

Raw peptide mass data were analyzed using MASCOT software, and the obtained peptide sequences were cross-referenced with the BIOPEP resource to confirm their antioxidative and antidiabetic activities.

Fractionation of peptides using RP-HPLC

Peptides from FSM were analysed using a Thermo UMSil C18(3) column (5 μm, 250 × 4.6 mm) in RP-HPLC (Shimadzu LC-20, Japan). A 20 µL sample was injected via a HAMILTON micro-injector. Eluent A (0.01% TFA in water) and Eluent B (0.01% TFA in 80:20 acetonitrile: water) comprised the mobile phase. The Shimadzu SPD-20 A detector was employed to detect the sample at 214 nm, and the separation was performed at room temperature with a 0.25 mL/min flow rate using a gradient protocol. Peptide peaks were identified in accordance with Khakhariya et al.²⁹.

Production of peptides via RP-HPLC

Peptides were separated from ultra-filtered fractions of FSM obtained via the RP-HPLC technique. Optimized fermentation with 2.5% inoculum at 30 °C for 48 h was followed by centrifugation (4193 × g, 30 min) to obtain WSEs as per standard protocol. This WSEs were then subjected to ultrafiltration (3 and 10 kDa) to separate different peptide fractions. The antidiabetic potential of the samples was assessed through α-amylase inhibition (“α-amylase inhibition”) and α-glucosidase inhibition (“α-Glucosidase inhibition”) assays as well as for antioxidative capacity as outlined in “Antioxidative activity evaluation (ABTS method)”.

Fourier transform-infrared spectroscopy

ATR-FTIR spectroscopy (Alpha, Bruker, Germany) was utilized to examine the functional groups, capturing spectra within the 4000–500 cm⁻¹ range at a resolution of 4 cm⁻¹, averaging 24 scans, as per the methodology of Pipaliya et al.³⁰.

Molecular docking

The potential antidiabetic activity of the most potent identified peptides (MSFVSLLLVGILFHATQAEQLTK) was predicted through molecular docking with α-amylase and α-glucosidase, as described in^30,31. The 3D structure of the peptide was produced via the AlphaFold Colab tool of Google DeepMind, while the 3D structures of the receptors, human α-amylase (PDB ID: 1KB3) and human α-glucosidase (PDB ID: 8EMR), were retrieved from the RCSB PDB. AutoDock Vina v.1.2.3 was employed to execute the mooring procedure. Moreover, the HADDOCK v2.4 web server was also employed for docking, focusing on the top binding cavity, while the peptide’s docking score was estimated using HPEPDOCK v2.0.

Inflammation-inhibiting activity

Cell culture condition

RAW264.7 cells (NCCS, Pune, India) were cultured in high-glucose DMEM supplemented with 10% FBS and 1% penicillin–streptomycin, maintained at 37 °C in 5% CO₂, and passaged every 24–48 h in T25 flasks. All the in vitro experiments were performed between 19 and 22 passage number of the cell.

Cytotoxicity assay

The 96-well plates were inoculated with 1 × 10⁵ RAW264.7 cells per well, and the cells were incubated at 37 °C with 5% CO₂ for 16 h. The cells were subsequently incubated for 24 h and exposed to a variety of dilutions of fermented milk samples (0.25, 0.5, 0.75, 1, 2.5, and 5 mg/mL). One hundred millilitres of 0.5 mg/mL MTT was added for 4 h, and the supernatants were discarded. The absorbance of formazan was measured at 570 nm using a Tecan M200 PRO reader after it was solubilised with 100 µL of DMSO.

Reactive oxygen species production

In order to assess the total Reactive oxygen species (ROS) production from RAW264.7 cells following co-treatment with LPS (1 µg/mL) and fermented milks, approximately 1 × 10⁵ cells were inoculated per well of a 12-well plate. The plate was incubated at 37 °C with 5% CO₂ for 24 h. Following that, the cells were exposed to per se LPS (1 µg/mL), fermented milk (0.5 mg/mL), and a combination of LPS (1 µg/mL) and fermented milks for the next 24 h. After incubation, the cells were scraped and resuspended in standard media containing 20 µM DCFDA. Subsequently, they were incubated in the dark for 30 min at 37 °C. Following this, the cells were examined using a BD FACS Aria III flow cytometer in the FITC-A channel.

Assessment of pro-inflammatory cytokines (TNF-α, IL-6, IL-1β)

Briefly for the quantification of pro-inflammatory cytokene production from RAW264.7 cells following co-treatment with LPS (1 µg/mL) and fermented milks, approximately 1 × 10⁵ cells were inoculated per well of a 12-well plate. The plate was incubated at 37 °C with 5% CO₂ for 24 h. Following that, the cells were exposed to per se LPS (1 µg/mL), fermented milk (0.5 mg/ml), and a combination of LPS (1 µg/mL) and fermented milks for the next 24 h. After incubation, the supernatants were collected, and the concentrations of IL-6, IL-1β, and TNF-α were quantified using commercially available ELISA kits (Elabscience, USA) according to the manufacturer’s instructions³².

Visualization of protein biomolecules of fermented sheep milk using CLSM

The distribution of protein in the samples was determined using a Confocal Laser Scanning Microscope (Leica TCS-SP8, Solms, Germany). To stain the proteins in the samples, fluorescein isothiocyanate (FITC, 15% w/w in dimethylsulfoxide, 1% concentration using Milli-Q water) were used. For the analysis, 100 µL of the sample was mixed with 25 µL of the FITC solution. This mixture was then vortexed for 5 s before being subjected to microscopy. Fluorescence was activated by laser excitation at a wavelength of 540 nm. Visualization of the protein structures was performed using a 63× objective lens³⁰.

Statistical analysis

The methodology outlined by Steel and Torrie³³ was employed to analyze the collected data. A significance level of 5% was used for all statistical evaluations. Data were analyzed using one-way ANOVA, and Tukey’s post hoc test was applied to determine which specific group means differed significantly. Statistical analysis was performed using GraphPad Prism version 8.0 (GraphPad Software, La Jolla, CA, USA).

Results and discussion

Antioxidative activity evaluation (ABTS method) of FSM

Figure 1 presents the antioxidative potential of sheep milk fermented with the KGL4 + WBS2A over different incubation periods (0–48 h). The activity increased progressively from 1.7 ± 0.02 at 0 h to 19.82 ± 0.39% at 12 h, 26.34 ± 0.45% at 24 h, 35.69 ± 0.33% at 36 h, then highest 40.08 ± 0.53% at 48 h; results indicated a significant variation (p ≤ 0.05) in antioxidative potential across the different incubation periods, suggesting enhanced antioxidant potential with prolonged fermentation. These findings indicate that sheep milk fermented with lactic acid bacteria and yeast showed a substantial increase in antioxidative activity as the fermentation time progressed at 30 °C.

Fig. 1.

ABTS radical-scavenging activity and antidiabetic activities (% inhibition) of fermented sheep milk at different incubation periods. Note: Values with different superscripts differ significantly (p ≤ 0.05). Results are expressed as mean ± SD (n = 3).

Sheep milk fermented with Lactiplantibacillus plantarum KGL3A exhibited a progressive increase in antioxidative activity as measured by the ABTS assay. Sheep milk fermented with KGL3A showed an increase in antioxidative activity over time. The activity was 23.8 ± 5.42% at 12 h and reached the highest value of 34.5 ± 1.05% at 48 h. The increasing trend suggests that longer fermentation improves the antioxidant potential of sheep milk³⁴. This study correlates well with our data, demonstrating a similar trend of increasing antioxidative activity with extended incubation time.

The antioxidant potential of Brazilian kefir-fermented sheep milk, assessed through the ABTS assay, exhibited a marked enhancement corresponding with extended fermentation and incubation durations. Initially, in the unfermented pasteurized milk (T0), the radical scavenging capacity was minimal, beginning at 5.73% at 6 min and progressively increasing to 15.91% by 180 min. Following 1st day of refrigerated storage (4 °C, T1), antioxidant activity increased to 19.50% at the same incubation interval. This upward trend continued steadily, with T7 sample showing a slight improvement to 19.73%. A more notable elevation occurred at T14 and T21, where the scavenging activity reached 31.17% at 180 min by day 21. The most substantial antioxidant activity was observed after 28 days of cold storage (T28), with values rising sharply from 23.51% at 6 min to a peak of 41.49% at 180 min. These results emphasise the critical role of both incubation duration and fermentation progression in enhancing the antioxidant efficacy of water-soluble peptides found in kefir-fermented sheep milk³⁵. These findings collectively support our results, where antioxidant activity increased consistently with fermentation time, underscoring the significance of prolonged fermentation in enhancing the bioactive potential of sheep milk.

Assessment of antidiabetic activities of FSM

The Fig. 1 illustrates the α-glucosidase inhibition and α-amylase inhibition of sheep milk fermented with the KGL4 + WBS2A strains at 30 °C. A notable elevation (p ≤ 0.05) was detected following 48 h of fermentation, where α-glucosidase inhibition rose from 2.35 ± 0.88 at 0 h to 75.96 ± 0.8, and α-amylase activity increased from 3.74 ± 0.61% to 72.48 ± 0.69%. These findings indicate that sheep milk fermented with lactic acid bacteria and yeast showed a substantial increase in both antidiabetic activities (α-glucosidase inhibition and α-amylase inhibition) as the fermentation time progressed at 30 °C.

The limitation of glucose absorption in the intestinal tract has been demonstrated by the inhibition of α-amylase and α-glucosidase enzymes, which impede the digestion of carbohydrates³⁶. This regulatory effect on carbohydrate metabolism is largely attributed to the presence of bioactive peptides, particularly low-molecular-weight peptides, which possess the capacity to modulate enzymatic activity³⁷. Pipaliya et al.³⁰ investigated the α-glucosidase inhibitory activity of sheep milk fermented with Limosilactobacillus fermentum (KGL4), revealing a substantial enhancement in enzyme inhibition correlated with fermentation time. At the initial time point (0 h), the inhibition was relatively modest at 6.77%; however, following 48 h of incubation (37 °C), the inhibition markedly increased to 72.19%. These results underscore the critical influence of fermentation duration in augmenting the antidiabetic potential of sheep milk. In a subsequent study, Pipaliya et al.³⁸ examined the impact of Lactiplantibacillus plantarum KGL3A on α-amylase inhibition during sheep milk fermentation. The initial inhibition level at 0 h was 9.15%, which progressively escalated with continued incubation, reaching 67.14% after 48 h at 37 °C. Collectively, these findings highlight a strong positive correlation between extended fermentation periods and enhanced enzymatic inhibition. Bhuva et al.³⁹ also concluded that the fermentation of camel milk using a combination of Lactobacillus strains and yeast significantly enhanced its antidiabetic properties. This finding further substantiates our results, as a similar synergistic effect was observed when sheep milk was fermented with Lactobacillus in combination with yeast. The improvement in α-amylase as well as α-glucosidase inhibition can be ascribed to complementary biochemical actions of LAB with yeast, leading to the generation of bioactive compounds with potent antidiabetic potential.

Assessment of proteolysis of FSM

Figure 2 illustrates the proteolytic activity in sheep milk fermented with the KGL4 + WBS2A strain at 30 °C, assessed under varying inoculation percentages (1.5, 2 and 2.5%) and incubation durations (0, 12, 24, 36 and 48 h). The activity rose from 2.42 ± 0.006 mg/mL at the initial time point (0 h, 1.5% inoculum) to a peak value of 6.74 ± 0.14 mg/mL after 48 h using a 2.5% rate. Statistical analysis (p ≤ 0.05) confirmed that both inoculum concentration and fermentation time significantly influenced proteolytic activity. The highest enzymatic breakdown of proteins occurred at 48 h with 2.5% inoculation (6.74 ± 0.14 mg/mL), demonstrating that longer fermentation and higher inoculum levels promote greater proteolysis in sheep milk.

Fig. 2.

Effect of inoculation rate and incubation period on the proteolytic activity (mg/mL) of fermented sheep milk. Note: Values with different superscripts differ significantly (p < 0.05). Results are expressed as mean ± SD (n = 3).

The proteolytic activity of Lactiplantibacillus plantarum in fermented sheep milk was found to be substantially enhanced by both incubation time and inoculation rate, as reported by Ashokhbhai et al.³⁴. In their investigation, the concentration of free amino groups—serving as a marker for proteolysis was found to range from 6.10 mg/mL at 12 h (1.5% inoculation) to a peak value of 10.40 mg/mL (48 h at 37 °C with 2.5% inoculum rate). These findings support our results, where higher incubation periods similarly led to elevated proteolytic activity. Similarly³⁸, fermenting sheep milk with the Lactiplantibacillus plantarum strain led to different levels of proteolytic activity depending on strain concentration and fermentation time. Initially, at 0 h, milk fermented with 1.5% strain exhibited a proteolytic activity of 3.84 mg/mL. This activity increased notably to 9.38 mg/mL after 48 h at 2.5% rate. The maximum activity observed at 48 h with a 2.5% inoculation rate. These findings highlight that prolonged fermentation and higher inoculum concentrations enhance proteolytic activity in fermented sheep milk. Similar trends were observed in our data, where sheep milk fermented with Lactobacillus with yeast showed increased proteolytic activity.

SDS-PAGE evaluation

SDS-PAGE analysis (Fig. 3) was performed on WSEs obtained from sheep milk fermented with the KGL4 + WBS2A strain combination, using a molecular weight marker (10–200 kDa). A higher diversity of protein bands was observed in the unfermented milk compared to the fermented ones, indicating significant proteolytic activity by the Lactobacillus and yeast strains used in fermentation. The unfermented sheep milk displayed prominent protein bands primarily between 15 and 85 kDa, consistent with the presence of intact proteins. Conversely, fermented samples exhibited fewer bands within the same range, suggesting substantial protein hydrolysis. Additionally, no protein bands were detected in the permeate fractions, implying complete removal or degradation of proteins in these samples.

Fig. 3.

Protein and peptide profile of sheep milk fermented with KGL4 + WBS2A revealed by SDS-PAGE (1: Protein ladder, 2: Unfermented sheepmilk, 3: Fermented sheepmilk, 4: 3 kDa permeate, 5: 3 kDa retentate, 6: 10 kDa permeate, 7: 10 kDa retentate).

SDS-PAGE analysis was performed by Ashokbhai et al.³⁴ to assess the protein profile of sheep milk that had been fermented with Limosilactobacillus fermentum (KGL4). The results indicated that the unfermented sheep milk exhibited distinct protein bands in comparison to its fermented counterpart, underscoring the extensive proteolytic activity of the LAB. Within the molecular weight range of approximately 26 to 42 kDa, protein bands were predominantly observed in the unfermented samples. Conversely, the fermented milk samples presented a broader spectrum of protein bands, spanning from 10 to 70 kDa, indicating that the microbial fermentation process resulted in the partial degradation and diversification of protein structures. In addition, the authors also discovered that fermented milk samples exhibited a more extensive array of protein bands, which ranged from 10 to 70 kDa. In contrast, the fermented milk samples in our study exhibited a broader spectrum of protein bands, ranging from 15 to 85 kDa, suggesting that the microbial fermentation of milk proteins resulted in partial degradation and structural diversification.

2D gel electrophoresis

Figure 4 present the two-dimensional electrophoresis of fermented sheep milk with KGL4 + WBS2A. A total of 39 protein spots were observed, primarily within the 15–70 kDa range, using IPG strips (pH 3–10), indicating protein separation based on isoelectric point and molecular weight.

Fig. 4.

2D Gel Electrophoresis of sheep milk fermented with KGL4 + WBS2A.

Fermented sheep milk (Lactobacillus strains KGL4 and KGL3A) was subjected to two-dimensional gel electrophoresis (2D PAGE) as per Ashokbhai et al.^34,40. Subsequently, trypsin digestion was performed, and the sample was analysed using reverse-phase liquid chromatography coupled with mass spectrometry (RP-LC/MS). A total of 42 distinct protein spots were detected 18 from the KGL4 strain and 24 from KGL3A within molecular weight (10 to 124 kDa), indicating active proteolysis by lactic acid bacteria. All spectral data were processed using ProteinPilot software to determine peptide mass fingerprints. These identified masses were further queried against the BIOPEP database to investigate their potential as antioxidative and antimicrobial peptides. Multiple bioactive peptide sequences, including ITMPLW, FAWPQYLK, HKEMPFPK, LDQWLCEK, KADEKKFW, SPAQTLQWQVLPNAVPAK, GPFPILF, SCQDQPTAMAR, AMKPWTQPKTNAIPYVRY, IPAVFK, and KFWGKYLYEVAR, were characterized from both KGL4 and KGL3A strains. Their functional properties were corroborated through BIOPEP, confirming their antioxidant and antimicrobial potential.

Peptide identification and structural analysis using RPLC/MS

The analysis of peptides produced from trypsin-digested protein spots obtained through 2D-PAGE was described by Pipaliya et al.³⁸ using reverse-phase liquid chromatography coupled with mass spectrometry (RP-LC/MS). Figure S1 illustrates the peptide score distribution and total ion chromatograms for sheep milk fermented with the KGL4 + WBS2A strain. Mass spectrometry data were processed using PeakView software, and peptide identification was performed by comparing results with curated, species-specific protein databases for sheep milk. Only non-toxic peptides with a Peptide Ranker score above 0.45 were selected for further analysis (Table 1). Mascot software was used for spectrum evaluation, and identified peptides were matched against entries in the Swiss-Prot database. To assess potential bioactivities specifically antioxidant and antidiabetic properties the identified sequences were further compared with peptide entries listed in the BIOPEP database (Tables 2 and 3).

Table 1.

Amino acid sequences obtained from sheep milk fermented with KGL4 + WBS2A with peptide ranking score.

Sequences	Source	Peptide Ranking score	Mol. wt.	Prediction	Retention time	Net charge at pH 7	Hydro-phobicity	Hydro-pathicity	Hydro-philicity
ENINELSK	α -S1-casein	0.078	946.14	Non-Toxin	8.3791	− 1	− 0.33	− 1.3	0.76
SPAQTLQWQVLPNAVPAK	κ -casein	0.49	1948.52	Non-Toxin	15.973	1	− 0.07	− 0.21	− 0.43
ENINELSK	β -lactoglobulin	0.078	946.14	Non-Toxin	8.387	− 1	− 0.33	− 1.3	0.76
SPAQTLQWQVLPNAVPAK	α -S1-casein	0.49	1948.52	Non-Toxin	15.814	1	− 0.07	− 0.21	− 0.43
IIAEKTKIPAVFK	κ -casein	0.22	1458.01	Non-Toxin	16.676	2	− 0.03	0.51	0.09
YNVPQLEIVPK	α -S1-casein	0.17	1299.7	Non-Toxin	14.715	0	− 0.08	− 0.2	− 0.23
DIGSESIEDQAMEDAKQMK	α -S1-casein	0.15	2125.58	Non-Toxin	2.4242	− 4	0.156169	0.156169	0.156169
AGSSSSSEEIVPNSAEQK	α -S1-casein	0.12	2329	Non-Toxin	8.5639	0	− 0.05	− 0.02	− 0.63
QPMIAVNQELAYFYPQLFR	α -lactalbumin	0.49	2547.42	Non-Toxin	15.729	0.5	0.1	0.96	− 0.67
MSFVSLLLVGILFHATQAEQLTK	α -S1-casein	0.79	2125.58	Non-Toxin	15.584	− 4	− 0.29	− 1.13	0.94
YPLRYPEVFQNEPDSIEEVLNK	κ -casein	0.52	120	Non-Toxin	4.1246	0	4.1246	4.1246	4.1246

Table 2.

Amino acid sequences with antioxidant activity obtained from sheep milk fermented with KGL4 + WBS2A searched on BIOPEP database.

Sequences	ID	Matched Sequences	Molecular Mass	Source	References
ENINELSK	7888	EL	260.13	Casein	41
YNVPQLEIVPK	8107	PK	243.15	Soybean protein	42
QPMIAVNQELAYFYPQLFR	7866	AY	252.11	Okara Protein	43
	7888	EL	260.13	Casein	41
	7963	YFY	491.20	Synthetic	44
	8288	YFYPQL	829.39	Casein	41
	10,377	YPQ	406.18	Soybean protein	45
	10,471	YF	328.14	Sorghum proteins	46
	10,732	FY	328.14	Perilla	47
MSFVSLLLVGILFHATQAEQLTK	10,749	LT	232.14	Pilot-scale black‐bone silky fowl	48
YPLRYPEVFQNEPDSIEEVLNK	10,051	RY	337.17	Soybean protein	49

Table 3.

Amino acid sequences with antidiabetic activity obtained from sheep milk fermented with KGL4 + WBS2A searched on BIOPEP database.

Sequences	ID	Matched Sequences	Molecular Mass	Source	References
ENINELSK	8804	IN	245.13	Soy protein hydrolysates	50
	8841	NE	261.09	Soy protein hydrolysates	50
	8894	SK	233.13	Soy protein hydrolysates	50
SPAQTLQWQVLPNAVPAK	3179	PA	186.20	Rat intestinal brush border membrane	51
	3180	LP	228.28	Rice bran	52
	3181	VP	214.26	Rice bran	52
	8505	SP	202.20	Rice bran	52
	8678	WQ	332.14	Milk protein	53
	8764	AV	188.22	Soy protein hydrolysates	50
	8839	NA	203.09	Soy protein hydrolysates	50
	8860	PN	229.10	Soy protein hydrolysates	50
	8878	QT	247.11	Soy protein hydrolysates	50
	8879	QV	245.13	Soy protein hydrolysates	50
	8880	QW	332.14	Soy protein hydrolysates	50
	8905	TL	232.27	Soy protein hydrolysates	50
	8922	VL	230.30	Soy protein hydrolysates	50
IIAEKTKIPAVFK	3179	PA	186.20	Rat intestinal brush border membrane	51
	8304	IPA	299.18	Whey proteins	54
	8501	IP	228.28	Rice bran	52
	8525	IA	202.13	Casein	55
	8558	EK	275.30	Milk protein	56
	8620	IPAVF	545.32	Whey protein concentrate	16
	8621	IPAVFK	673.41	Whey protein concentrate	16
	8758	AE	218.20	Soy protein hydrolysates	50
	8764	AV	188.22	Soy protein hydrolysates	50
	8801	II	244.17	Soy protein hydrolysates	50
	8812	KI	259.34	Soy protein hydrolysates	50
	8816	KT	247.15	Soy protein hydrolysates	50
	8904	TK	247.15	Soy protein hydrolysates	50
	8917	VF	264.31	Soy protein hydrolysates	50
YNVPQLEIVPK	8772	EI	260.28	Soy protein hydrolysates	50
	8851	NV	231.12	Soy protein hydrolysates	50
	8858	PK	243.30	Soy protein hydrolysates	50
	8861	PQ	243.25	Soy protein hydrolysates	50
	8874	QL	259.30	Soy protein hydrolysates	50
	8942	YN	295.11	Soy protein hydrolysates	50
	3181	VP	214.26	Rice bran	52
SCQDQPTAMAR	3173	MA	220.08	Rice bran	52
	8531	TA	190.09	Casein	55
	532	QP	243.12	Casein	55
	8768	DQ	261.09	Soy protein hydrolysates	50
	8868	QD	261.23	Soy protein hydrolysates	50
DIGSESIEDQAMEDAKQMK	8768	DQ	261.09	Soy protein hydrolysates	50
	8773	ES	234.20	Soy protein hydrolysates	50
	8826	ME	278.09	Soy protein hydrolysates	50
	8831	MK	277.38	Soy protein hydrolysates	50
	8867	QA	217.22	Soy protein hydrolysates	50
	8893	SI	218.24	Soy protein hydrolysates	50
AGSSSSSEEIVPNSAEQK	3181	VP	214.26	Rice bran	52
	8758	AE	218.20	Soy protein hydrolysates	50
	8772	EI	260.28	Soy protein hydrolysates	50
	8860	PN	229.10	Soy protein hydrolysates	50
QPMIAVNQELAYFYPQLFR	3175	LA	202.25	Rat intestinal brush border membrane	51
	8521	YP	278.30	Rice bran	52
	8525	IA	202.13	Casein	55
	8532	QP	243.12	Casein	55
	8764	AV	188.22	Soy protein hydrolysates	50
	8765	AY	252.26	Soy protein hydrolysates	50
	8780	FR	321.37	Soy protein hydrolysates	50
	8830	MI	262.13	Soy protein hydrolysates	50
	8848	NQ	260.11	Soy protein hydrolysates	50
	8859	PM	246.32	Soy protein hydrolysates	50
	8861	PQ	243.25	Soy protein hydrolysates	50
	8869	QE	275.25	Soy protein hydrolysates	50
	8874	QL	259.30	Soy protein hydrolysates	50
	8924	VN	231.12	Soy protein hydrolysates	50
	8935	YF	328.14	Soy protein hydrolysates	50
	3175	LA	202.25	Rat intestinal brush border membrane	51
MSFVSLLLVGILFHATQAEQLTK	3182	LL	244.32	Rat intestinal brush border membrane	51
	3184	HA	226.23	Rat intestinal brush border membrane	51
	8560	SL	218.24	Milk protein	56
	8758	AE	218.20	Soy protein hydrolysates	50
	8763	AT	190.19	Soy protein hydrolysates	50
	8785	GI	188.22	Soy protein hydrolysates	50
	8802	IL	244.32	Soy protein hydrolysates	50
	8824	LT	232.27	Soy protein hydrolysates	50
	8825	LV	230.30	Soy protein hydrolysates	50
	8867	QA	217.22	Soy protein hydrolysates	50
	8874	QL	259.30	Soy protein hydrolysates	50
	8891	SF	252.11	Soy protein hydrolysates	50
	8904	TK	247.15	Soy protein hydrolysates	50
	8908	TQ	247.11	Soy protein hydrolysates	50
	8918	VG	174.19	Soy protein hydrolysates	50
	8926	VS	204.22	Soy protein hydrolysates	50

The peptide sequence MSFVSLLLVGILFHATQAEQLTK, identified from sheep milk fermented using KGL4 + WBS2A, showed similarity to previously reported peptides such as LT⁴⁸. Similarly, another peptide, YPLRYPEVFQNEPDSIEEVLNK, aligned with the peptide RY described by Amigo et al.⁴⁹ in the BIOPEP database and confirmed its antioxidant activity.

The peptide sequence ENINELSK, identified from sheep milk fermented using KGL4 + WBS2A, showed similarity to previously reported peptides such as IN, NE, and SK⁵⁰. Similarly, the peptide SPAQTLQWQVLPNAVPAK matched with known peptide fragments including PA⁵¹, LP, VP, and SP⁵², as well as WQ⁵³, and AV, NA, PN, QT, QV, QW, and TL⁵⁰ in the BIOPEP database, confirming their potential antidiabetic activity.

Ultra-filtered fractions of FSM through RP-HPLC

Peptide fractions from both 3 and 10 KDa (permeate & retentate) molecular weight ranges were analyzed using RP-HPLC. This analysis was carried out under optimized fermentation conditions. As shown in Table 4, RP-HPLC profiles were categorized into < 3 kDa, > 3 kDa, < 10 kDa, and > 10 kDa fractions. Comparative chromatographic data (Figures S2–S7) for both fermented and unfermented sheep milk samples revealed a marked increase in peptide content after fermentation. Unfermented samples mainly displayed peaks associated with intact proteins, while fermented samples demonstrated a broader range of hydrolysates, with distinct peaks appearing between 5 and 47 min of retention time.

Table 4.

Characterization of < 3 kDa, > 3 kDa, < 10 kDa, > 10 kDa peptides generated from fermented sheep milk by RP-HPLC analysis.

Milk	Sample	Number of peaks	Retention time (min)
Sheep milk (KGL4 + WBS2A)	< 3 kDa	45	5.75 to 46.66
	> 3 kDa	39	5.81 to 46.83
	< 10 kDa	41	6.12 to 46.66
	> 10 kDa	41	7.02 to 48.82

Figure 5 highlights the antioxidant and antidiabetic effects of ultrafiltered fractions from fermented sheep milk (FSM). Among these, the retentate fraction > 3 kDa showed the highest antioxidative activity at 50.73%. Additionally, the < 3 kDa permeate fraction exhibited the strongest α-amylase inhibition, reaching 76.25%. Notably, the same < 3 kDa fraction also recorded a significantly elevated α-glucosidase inhibition value of 77.69%, suggesting strong antidiabetic potential associated with the peptides produced under KGL4 + WBS2A fermentation conditions.

Fig. 5.

ABTS radical scavenging and antidiabetic activities of ultra-filtered fractions (3 kDa and 10 kDa permeate and retentate) from fermented sheep milk. *Values with different superscripts differ significantly (p < 0.05), ABTS radical scavenging and antidiabetic activities (%) Mean ± SD of three replicate experiments (n = 3).

The antioxidant activity of sheep milk that was fermented using KGL3A and KGL4 strains and subsequently subjected to ultrafiltration with 3 kDa and 10 kDa molecular weight cut-off membranes exhibited significant differences (p ≤ 0.05), as per Ashokbhai et al.^34,40. In the < 3 kDa permeate fraction, KGL3A displayed higher free radical scavenging activity at 32.98%, compared to 23.86% for KGL4. A similar trend was observed in the corresponding retentate fraction, with KGL3A achieving 33.59% antioxidant activity, outperforming KGL4’s 23.71%. In the < 10 kDa permeate, KGL3A again demonstrated superior antioxidant potential, recording the highest value among all tested samples at 34.65%, while KGL4 showed 26.9%. Interestingly, this pattern was reversed in the >10 kDa retentate fraction, where KGL4 exhibited greater antioxidant capacity (26.49%) than KGL3A (18.38%).

Shukla et al.⁵⁷ investigated the antidiabetic potential of fermented camel milk using Lactobacillus plantarum KGL3A, highlighting its effectiveness through enzymatic inhibition assays. In the < 3 kDa permeate fraction, the investigation recorded inhibition rates of 82.10% for α-amylase, 67.14% for lipase, and 64.98% for α-glucosidase. In the >3 kDa retentate fractions, inhibitory activity declined, with values of 53.89% for α-amylase, 74.62% for lipase, and 44.90% for α-glucosidase. For the < 10 kDa permeate fractions, inhibition levels reached 64.93% for lipase, 80.19% for α-amylase, and 60.35% for α-glucosidase, while the corresponding >10 kDa retentates exhibited reduced rates of 51.78%, 73.95%, and 39.32%, respectively. These findings reinforce the idea that lower molecular weight peptide fractions, especially those under 3 kDa, possess greater antidiabetic activity supporting similar trends observed in our fermented sheep milk samples. Fermentation of Camel milk using Lactobacillus strain M11 and yeast strain WBS2A exhibited notable α-glucosidase inhibition across its ultrafiltered fractions. The 3 kDa retentate showed an inhibition of 42.05%, while the 10 kDa retentate demonstrated a higher inhibition of 60.48%. Among all, the 10 kDa permeate exhibited the highest α glucosidase inhibition, reaching 63.33%, indicating that low molecular weight bioactive peptides present in this fraction may contribute significantly to the antidiabetic potential of the fermented camel milk²⁹.

Fourier transform-infrared spectroscopy

FTIR analysis was conducted to investigate the secondary structure and functional groups of proteins in sheep milk. Fermentation with the KGL4 + WBS2A strain led to a marked increase in absorbance at 1639 cm⁻¹, as shown in Fig. 6. This peak, located within the amide I region, corresponds to C=O stretching vibrations of peptide bonds. A significant rise in intensity within the 1620–1640 cm⁻¹ range characteristic of β-sheet structures was observed in the fermented milk and its ultrafiltered fractions. These β-sheet peaks were noticeably stronger in the fermented samples compared to the unfermented control, indicating considerable alterations in protein conformation. Additionally, a reduction in α-helix content was evident post-fermentation, suggesting extensive proteolysis, protein unfolding, and the formation of smaller, peptide-rich components.

Fig. 6.

Fourier Transform-Infrared Spectroscopy (FTIR) of sheep milk fermented with KGL4 + WBS2A (USM- Unfermented sheep milk, SL4 - sheep milk fermented with KGL4 + WBS2A, SM9-3P – 3 kDa permeate, SL4-3R- 3 kDa retentate, SL4-10P – 10 kDa permeate, SL4-10R- 10 kDa retentate).

FTIR spectral analyses of fermented camel and Gir cow milk, which demonstrated prominent and distinct peaks in the fingerprint region (1700–500 cm⁻¹), indicative of structural changes in milk proteins during fermentation. Specifically, fermented camel milk exhibited characteristic peaks at 1697 cm⁻¹ and 1568 cm⁻¹, corresponding to amide I and amide II regions, respectively, suggesting proteolytic activity and peptide formation. Similarly, unique peaks in fermented Gir cow milk at 1293 cm⁻¹ and 1492 cm⁻¹ were attributed to protein and peptide-related vibrations. These spectral features are consistent with the observed increase in amide I band intensity at 1639 cm⁻¹ in fermented sheep milk and 1637 cm⁻¹ in fermented camel milk in the present study, confirming the breakdown of native proteins and the generation of bioactive peptides due to microbial fermentation³⁹.

Molecular docking analysis of selected peptide

Molecular docking is an important tool used to analyze the conformation and orientation of various ligands within the active sites of a target receptor³¹. Therefore, the most potent peptide (MSFVSLLLVGILFHATQAEQLTK) identified in this study was selected for molecular docking with human α-amylase and human α-glucosidase enzymes to evaluate its potential antidiabetic activity. Three tools, AutoDock Vina v1.2.3, HADDOCK v2.4, and HPEPDOCK v2.0, were utilized for comprehensive analysis and validation of peptide-enzyme interactions. The findings from AutoDock Vina predicted binding affinity of the selected peptide with human α-amylase and α-glucosidase as -3.8 and 3.2 kcal/mol, respectively. The negative binding energy values for α-amylase suggest that the selected peptide has moderate potential to bind and affect the activity of these crucial carbohydrate-hydrolyzing enzymes. Previously documented docking studies show that acarbose (a standard antidiabetic inhibitor) generally exhibits binding energies around − 6.1 to − 9.2 kcal/mol for α-amylase^58,59 and approximately − 8.5 kcal/mol for α-glucosidase⁶⁰. Compared to the positive control, the chosen peptide showed weaker affinities, indicating limited inhibitory potential. The positive binding energy seen with α-glucosidase suggests a lack of favorable interaction⁶¹, possibly due to the peptide’s steric hindrance or improper orientation within the catalytic pocket. Typically, effective peptide–protein interactions have binding energies below − 5.0 kcal/mol, reflecting stronger and more stable complexes⁶². Therefore, the selected peptide exhibits moderate affinity for α-amylase and weak or non-specific interactions with α-glucosidase. This indicates that while the peptide may partially inhibit α-amylase, its effect on α-glucosidase may be minimal or allosteric in nature.

Furthermore, AutoDock Vina results illustrated with AutoDock MGL v1.5.7 indicated that the selected peptide predicted to form strong hydrogen bonds with HIS201 catalytic residue of human α-amylase, which plays a critical role in the enzyme active site (Fig. 7a). Furthermore, the selected peptide was predicted to form stable interactions with several other residues of human α-amylase, including GLN63, LEU162, THR163, TYR251, GLY308, LYS200, ILE235, LEU237, GLY238, and LYS257. A similar result was found with α-glucosidase, where the selected peptide was predicted to form five hydrogen bonds with residues such as VAL4, LEU12, THR16, LYS23, AND GLU19, along with other interactions (Fig. 7b). These residues are situated within or around the enzymes active pocket, and the interactions among these multiple residues indicate the potential of peptides to act as a as a competitive or non-competitive inhibitor.

Fig. 7.

AutoDock Vina results for the peptide (MSFVSLLLVGILFHATQAEQLTK), visualized using AutoDock MGL Tools v1.5.7, illustrating its interactions with various residues of human α-amylase (A) and human α-glucosidase (B) enzymes.

The HADDOCK web server was also employed to predict protein-peptide binding, with a more negative HADDOCK score indicating a stronger predicted binding affinity. The HADDOCK scores for the selected peptides binding to the active site of α-amylase and α-glucosidase were predicted to be − 48.3 ± 2.8 and − 41.0 ± 2.1, respectively. Concurrently, the peptide exhibited the highest HPEPDOCK scores, with a − 237.828 for α-amylase and a − 209.866 for α-glucosidase. The HPEPDOCK web server employs a hierarchical algorithm that eliminates the necessity for extensive modelling to improve the conformations of peptides. The results of the molecular docking analysis indicate that the peptides that were identified inhibit the activity of the target enzymes by interacting with various residues located within and near the active site. This inhibition is primarily achieved through hydrogen bonding and hydrophobic interactions.

Consequently, the presence of these hydrogen bonds and the intensity of the interactions were verified through molecular docking using both AutoDock Vina and HADDOCK software. The selected peptide’s potential as an effective diabetes management agent is underscored by the collective findings of the docking studies, which indicate that it has a high binding affinity for both α-amylase and α-glucosidase. The results of the observed interactions are more credible and serve as a foundation for future in vivo and human validation studies due to the consistency of the results across various docking platforms. It is noteworthy that the docking outcomes from HADDOCK and HPEPDOCK predicted more favorable interaction scores compared to AutoDock Vina, emphasizing the need for multi-software validation. Future structural optimization of the peptide sequence could enhance its α-glucosidase binding affinity.

Anti-inflammation activity

Figure 8A displays the viability of RAW 264.7 cells treated with sheep milk fermented using the KGL4 + WBS2A strain at various concentrations. At lower concentrations (0.25 and 0.5 mg/mL), no cytotoxic effects were observed, and cell viability remained close to 100%. However, as the concentration increased beyond 0.5 mg/mL, a dose-dependent reduction in cell viability was noted, indicating increasing cytotoxicity at higher levels of the fermented sample.

Fig. 8.

Anti-inflammatory activity and ROS of sheep milk fermented with KGL4 + WBS2A, (A) cell viability in RAW264.7 cells, (B) IL-1β, (C) TNF-α, and (D) IL-6 production; (E) Apoptotic cell population (F) ROS production. Data were expressed as mean ± SD and analyzed by one-way ANOVA followed by Tukey’s post hoc test.

The inflammation-suppressing activity of sheep milk fermented using KGL4 + WBS2A in LPS-stimulated RAW 264.7 macrophages. According to the results shown in Fig. 8B–D, treatment with 0.5 mg/mL of the fermented sample significantly lowered TNF-α and IL-1β levels. A higher concentration of 1 mg/mL resulted in a more substantial reduction in IL-6 expression. These observations indicate a dose-dependent suppression of IL-6, while TNF-α and IL-1β respond effectively even at lower concentrations. Notably, the cytokine levels in the treated groups closely resembled those of the unstimulated control, suggesting an almost complete mitigation of LPS-induced inflammatory responses. Furthermore, fermentation-derived bioactive peptides in dairy systems have been shown to exert anti-inflammatory effects primarily through suppression of NF-κB activation including inhibition of IκBα degradation and p65 nuclear translocation and by modulating MAPK and oxidative-stress pathways, providing a mechanistic basis for the reduced cytokine secretion observed in this model.

The Annexin-V/PI assay revealed that LPS exposure significantly increased the late apoptotic cell population in RAW 264.7 cells, reaching 40.5%, indicating severe inflammatory stress (Fig. 8E). Treatment with sheep milk fermented using the KGL4 + WBS2A strain reduced this late apoptotic population to 15.6%, demonstrating a notable protective effect against LPS-induced apoptosis and highlighting the cytoprotective potential of the fermented sample.

ROS levels were assessed to determine the antioxidant efficacy of FSM in LPS-activated RAW 264.7 cells. LPS exposure resulted in a marked increase in ROS production, reaching 44%, which reflects significant oxidative stress. In contrast, treatment with the fermented sheep milk notably decreased ROS levels to 11.7% (Fig. 8F), indicating strong antioxidant effects and effective protection against LPS-induced oxidative damage.

Overall, the results strongly suggest that fermentation of camel and sheep milk with selected strains enhances their anti-inflammatory, antioxidant, and cytoprotective effects. These fermented dairy products effectively suppress pro-inflammatory cytokine production, lower oxidative stress, and reduce apoptosis, highlighting their potential as therapeutic agents for protecting against inflammation-induced cellular damage.

The viability of RAW 264.7 cells was evaluated using the MTT assay after exposure to sheep milk fermented with the KGL4 strain. The assessment involved six concentrations (0.5, 1, 2, 4, 6, and 8 mg/mL) along with a control group that did not receive the bacterial strain. At lower concentrations of 0.5, 1, and 2 mg/mL, the fermented sheep milk exhibited no cytotoxic effects, maintaining nearly full cell viability. In contrast, cell viability significantly decreased at higher concentrations (4, 6, and 8 mg/mL), indicating increased cytotoxicity. The study further examined the impact of fermented sheep milk on nitric oxide (NO) production and pro-inflammatory cytokines in RAW 264.7 cells. A sharp rise in nitrite levels was observed in the LPS-stimulated group; however, these levels were significantly lowered when the cells were co-treated with fermented sheep milk. Cells treated with only the fermented milk displayed the lowest nitrite levels, nearly matching the control, suggesting that the peptides formed during fermentation did not negatively affect cell health. Moreover, treatment with 0.5 mg/mL of fermented sheep milk markedly reduced the overexpression of inflammatory cytokines TNF-α, IL-6, and IL-1β triggered by LPS stimulation. The cytokine profiles in these samples were closely aligned with the control, highlighting the fermented milk’s effective anti-inflammatory properties³⁰. Similar to the results reported by Pipaliya et al.³⁰, we also observed that cell viability significantly decreased at higher concentrations of fermented sheep milk, indicating dose-dependent cytotoxicity. Additionally, our study showed a notable reduction in the expression of pro-inflammatory cytokines such as TNF-α, IL-6, and IL-1β, aligning with their findings. Furthermore, we also recorded a substantial decrease in ROS production, further supporting the anti-inflammatory and antioxidant potential of fermented sheep milk.

Visualization of protein biomolecules of FSM using confocal laser scanning microscopy (CLSM)

Figure S8 presents the CLSM images of sheep milk fermented with the KGL4 + WBS2A strain, showing noticeable structural changes compared to unfermented milk. The fermented sample exhibited a more fragmented protein matrix and reduced fluorescence intensity, indicating enhanced proteolysis. Supporting this, both the 3 kDa and 10 kDa permeate fractions from the FSM contained smaller peptides compared to their respective retentates, confirming effective protein breakdown and peptide release.

In a recent study, Bhuva et al.³⁹ utilized Confocal Laser Scanning Microscopy (CLSM) to examine the microstructural changes in both unfermented and fermented camel and Gir cow milk treated with the KGL4 + WBS2A strains. Their findings revealed that in the unfermented milk, larger protein aggregates with simple protein arrangements were observed, characteristic of native proteins. However, in the fermented milk samples, a significant transformation occurred, with the formation of larger protein aggregates and more complex protein networks. This structural alteration is likely attributed to the ability of the KGL4 + WBS2A strains to break down and modify proteins into various polymeric forms during fermentation. Membrane filtration of fermented milk using 3 kDa and 10 kDa cut-offs revealed that the 3 kDa permeate comprised smaller protein fragments, indicating short peptides. In contrast, the 10 kDa permeate and retentate exhibited larger protein aggregates, indicating that higher molecular weight components were retained. This demonstrates the role of selective filtration in distinguishing protein fractions based on size and highlights the significant effects of fermentation on protein structures in camel and Gir cow milk.

Suetsuna et al.⁶³ employed cryogenic scanning electron microscopy (cryo-SEM) alongside confocal laser scanning microscopy (CLSM) to investigate the microstructural features of yogurt gels derived from cow, goat, and sheep milk. Their study revealed distinct differences in protein network organization depending on the milk source. Yogurt made from goat milk displayed the most porous and loosely connected protein structure, indicating a less dense gel matrix. In contrast, sheep milk yogurt exhibited a highly compact and uniform protein network, suggesting a denser and more cohesive structure. The microstructure of yogurt prepared from cow milk was found to be intermediate, with a protein arrangement that was denser than that of goat milk yogurt but less compact than that of sheep milk yogurt.

Conclusion

Fermentation of sheep milk with Lactobacillus and yeast strains led to the production of bioactive peptides exhibiting promising antioxidative, antidiabetic, anti-inflammatory, and proteolytic properties. Maximum activity was observed at 48 h of incubation with a 2% inoculum, where antioxidant activity reached 40.08%, α-amylase inhibition was 72.48%, and α-glucosidase inhibition reached 75.96%. Enhanced proteolytic activity, as measured by the OPA method, was associated with increased peptide release mainly within the 3–10 kDa range, known for better bioactivity and absorption. SDS-PAGE and 2D-PAGE analysis revealed extensive protein hydrolysis, showing peptide fragments ranging from 15 to 85 kDa, and 39 protein spots between 15 and 70 kDa, confirming proteolytic efficiency. FTIR spectroscopy further indicated structural modifications post-fermentation, especially at amide I and II bands, supporting peptide formation. CLSM imaging confirmed enhanced protein degradation and morphological changes in fermented samples compared to unfermented milk. The comprehensive molecular docking approach suggests strong antidiabetic effect of selected peptides by effectively binding to the active pocket of enzymes human α-amylase and α-glucosidase. Moreover, the fermented sheep milk significantly reduced LPS-stimulated cytokine synthesis in RAW 264.7 cell, indicating potential anti-inflammatory effects. The correlation between low-molecular-weight peptides and increased enzyme inhibition activity reinforces their therapeutic potential. Sheep milk fermented with LAB and yeast strains emerge as a potent source of multifunctional bioactive peptides. However, as the current study is limited to in vitro evaluations, these results primarily indicate the preliminary therapeutic potential of the fermented product. Future work will focus on in vivo validation through animal studies and, subsequently, human trials to confirm the bioavailability, safety, and clinical efficacy of the identified peptides.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ABTS

2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid)

ALA

α-linolenic acid

ATR

Attenuated total reflection

CLA

Conjugated linoleic acids

CLSM

Confocal laser scanning microscopy

DCFDA

2’,7’-dichlorofluorescin diacetate

DMSO

Dimethyl sulfoxide

DNSA

3,5-dinitrosalicylic acid

ELISA

Enzyme-linked immunosorbent assay

EMS

Enhanced MS

EPI

Enhanced product ion

ESI

Electrospray ionization

FAO

Food and agriculture organization

FBS

Fetal bovine serum

FITC

Fluorescein isothiocyanate

FSM

Fermented sheep milk

FTIR

Fourier-transform infrared spectroscopy

IDA

Information-dependent acquisition

IL-1β

Interleukin 1 beta

IL-6

Interleukin 6

IPG

Immobilized pH gradient

LAB

Lactic acid bacteria

LC

Liquid chromatography

LPS

Lipopolysaccharide

MPL

Milk polar lipids

MS

Mass spectrometry

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

OPA

o-phthaldialdehyde

P-NPG

p-nitrophenyl-α-D-glucopyranoside

RP-HPLC

Reverse phase high-performance liquid chromatography

SDS-PAGE

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis

TFA

Trifluoroacetic acid

TNF-α

Tumor necrosis factor alpha

WSE

Water-soluble extracts

Author contributions

Prashantkumar Natubhai Padhiyar designed and carried out the experiments, analyzed the data, visualization and prepared the manuscript; Subrota Hati helped in Supervision, Conceptual design, Resources, Investigation, selecting the experiments and edited & reviewed the manuscript; P Mankad helps in conducting the experiments, and analysis the data and visualization ; A. Sakure, A. Bhattacharya, K K Kondepudi helps in conducting the experiments, resources, and analysis the data and visualization ; B.P. Singh, S. Paul helps in Molecular docking analysis and edited the manuscript; P. Sarkar helps in conducting the experiments, resources and analysis the data; A. Patel helps in resources and editing the manuscript and Shaikh Adil reviewed, edited and formatted the manuscript.

Funding

This research was supported by institutional funding from SMC College of Dairy Science, Anand.

Data availability

The datasets used and/or analyzed during the current study available from the corresponding authors/Principal author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

The research was ethically approved by the PG Research committee, Kamdhenu University and Advisory Committee members from SMC College of Dairy Science, Anand, Gujarat, India. All the committee members were duly informed about their rights and responsibilities and they provided us with explicit written consent. The study was conducted in agreement with the guidelines governing research involving human participants, as outlined by the Ethics Committee of SMC College of Dairy Science, Anand, Gujarat, India.