Anti-Aging Potential of Bioactive Peptides Derived from Casein Hydrolyzed with Kiwi Actinidin: Integration of In Silico and In Vitro Study

Abstract

Background: Skin aging is mainly associated with oxidative stress and enzymatic degradation of collagen and elastin by protease activity. Peptides have antioxidant capacity and inhibitory effects on protease enzymes. Objective: The purpose of this study was to obtain peptides with in vitro anti-aging activity from the enzymatic hydrolysis of bovine casein with actinidin, a protease extracted from the green kiwi fruit (Actinidia deliciosa) Methodology: The enzyme actinidin was extracted from the pulp of the kiwi fruit, purified by ion exchange chromatography and characterized by polyacrylamide electrophoresis (SDS-PAGE). Subsequently, the extracted enzyme was used to hydrolyze commercial bovine casein at 37 °C for 30 min, precipitating the peptide fraction with trichloroacetic acid (TCA), and centrifuged. To determine the anti-aging potential of the peptides in vitro, antioxidant activity was evaluated using the ABTS (2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) radical. Additionally, the inhibitory capacity of the peptides against collagenase and elastase enzymes was also studied. To complement the in vitro results, the enzymatic hydrolysis of casein with actinidin was simulated. The binding energy (ΔG) of each of the hydrolysates with the collagenase and elastase enzymes was calculated using molecular docking to predict the peptide sequences with the highest probability of interaction. Results: Actinidin was extracted and purified exhibiting a molecular weight close to 27 kDa. The enzyme hydrolyzed the substrate by 91.6%, and the resulting hydrolysates showed moderate in vitro anti-aging activity: antioxidant (17.5%), anticollagenase (18.55%), and antielastase (28.6%). In silico results revealed 66 peptide sequences of which 30.3% consisted of 4–8 amino acids, a suitable size to facilitate interaction with structural targets. The sequences with the highest affinity were FALPQYLK and VIPYVRYL for collagenase and elastase, respectively. Conclusions: Despite the modest inhibition values, the use of a fruit-derived enzyme and a food-grade substrate is in line with current trends in sustainable and natural cosmetics. These findings highlight the great potential for laying the groundwork for future research into actinidin-derived peptides as multifunctional and eco-conscious ingredients for the development of next-generation anti-aging formulations.

1. Introduction

Human skin performs vital functions such as protection against physical and biological agents, thermoregulation, sensory perception, and immune surveillance. However, with aging and prolonged exposure to environmental stressors, its regenerative capacity diminishes, thereby increasing susceptibility to infections, dryness, autoimmune disorders, and structural deterioration [1]. Among the most visible changes are loss of elasticity, wrinkle formation, uneven pigmentation, and reduction in the abundance of dermal fibers like collagen and elastin, which are essential for skin firmness and resilience [2].

Collagen accounts for approximately 30% of total body protein and plays a crucial role in dermal architecture through its triple-helix structure [3]. Its degradation, along with that of elastin, is catalyzed by enzymes such as collagenase and elastase, classified as matrix metalloproteinases (MMPs) [4]. Although these proteases are physiologically regulated, their activity increases markedly with intrinsic aging and photoinduced mechanisms, particularly under chronic sun exposure [5]. Ultraviolet radiation induces the overproduction of reactive oxygen species (ROS), promoting epidermal oxidative stress, activating MMPs, and accelerating extracellular matrix (ECM) breakdown [6].

In response to this challenge, the search for natural compounds with anti-aging activity capable of inhibiting collagenase, elastase, and neutralizing ROS has intensified. Bioactive peptides have emerged as promising functional ingredients in cosmetic formulations due to their antioxidant properties, low allergenicity, and cost-effective production [7]. These peptides can be obtained through enzymatic hydrolysis of accessible proteins such as bovine casein, which has demonstrated multiple physiological properties, including antihypertensive, antioxidant, and antimicrobial effects [8]. The antioxidant capacity of peptides is mainly governed by structural properties such as peptide size or molecular weight and amino acid composition [9]. For example, negatively charged amino acids such as Glu and Asp or reducing amino acids, such as cysteine, could serve as potent scavengers of ROS due to the excess electrons that are readily donated from the side chain [10]. Additionally, some hydrophobic amino acids such as Ala, Pro, Leu, and aromatic amino acids have also been linked to high antioxidant capacity due to direct electron transfer [11]. Thus, peptide sequences rich in anionic and hydrophobic aromatic amino acids or cysteines have a high antioxidant potential.

Hydrolysis of casein enables the release of low-molecular-weight peptides, capable of interacting with enzymes implicated in aging processes. Specifically, peptides containing 4–8 amino acid residues have shown significant affinity with collagenase and elastase, enhancing inhibitory efficacy [7,12]. Previous studies have demonstrated that peptides <3 kDa penetrate the epidermis more effectively and exert beneficial effects [13,14,15]. It should be considered that not only the size of the peptide modulates the affinity with both enzymes. The type of amino acid plays an essential role in the interaction with these enzymes. For example, tyrosine participates directly in the elastase binding site; even the nitration of a single residue reduces the inhibitory capacity by up to 70% [16]. In addition, both valine and isoleucine allow covalent interaction with the serine oxygen present in the active site of elastase, imitating the transition state of substrate hydrolysis, and this generates inhibition [17]. On the other hand, lysine with its cationic groups establishes electrostatic interactions with carboxylates and aspartic and glutamic acid residues present in the collagen matrix, favoring the stability of its three-dimensional structure [18].

Among the enzymes used to generate functional peptides, actinidin—a cysteine protease from green kiwi (Actinidia deliciosa)—stands out for its broad catalytic activity range (pH 4–10) and preferential cleavage at carboxyl-lysine peptide bonds [19]. Its application in milk protein hydrolysis has been validated, showing reduced antigenicity of β-lactoglobulin and α-casein. Moreover, actinidin has exhibited capacity to produce peptides with documented antioxidant and enzyme-modulatory properties [20,21]. Despite these findings, its use as a biocatalyst for generating anti-aging peptides from bovine casein remains largely underexplored. The integration of bioinformatics tools such as in silico peptide hydrolysis and molecular docking enables prediction of peptide sequences and their affinity for target enzymes, optimizing selection and validation. Tools such as PeptideCutter simulate cleavage patterns of actinidin analogs like Lys-C, generating peptides of interest for computational evaluation [22,23]. Docking studies have shown that peptides with polar groups and aromatic residues exhibit higher binding affinity through hydrophobic interactions and hydrogen bonding, which are features considered critical for enzyme inhibition in cosmetic applications [24,25,26].

Previous studies on milk-derived peptides have reported variable antioxidant and enzymatic activities depending on the enzyme type, hydrolysis degree, and separation method [27,28,29]. For example, Zhao et al. [30] found that peptide profiles directly affect the antifatigue and antioxidant activities of whey protein hydrolysates. Moreover, parameters such as peptide size, aromatic residue content, and hydrophobicity influence enzyme interaction potential, particularly with targets implicated in skin aging [31].

The selection of actinidin as a biocatalyst offers several advantages over conventional proteases. Its natural origin, broad pH stability, specificity, and low immunogenicity make it a sustainable and efficient enzymatic alternative for cosmetic bioconversion [19,32]. Previous research has shown that actinidin achieves hydrolysis degrees above 90%, as reported in α-amylase degradation and milk protein optimization for allergenicity reduction [33,34,35].

Consequently, this study proposes an integrated experimental and computational approach to evaluate the anti-aging potential of casein-derived peptides obtained via actinidin-mediated hydrolysis. By combining in vitro biochemical assays with in silico peptide modeling and molecular docking, this strategy aims to identify bioactive sequences with predicted affinity for collagenase and elastase, agreeing with sustainable enzyme sourcing and supporting the development of natural cosmetic bioactives.

2. Materials and Methods

2.1. Materials

All reagents used in this study were of analytical grade. Key chemicals and buffers—including sodium hydroxide (NaOH), hydrochloric acid (HCl), bicinchoninic acid (BCA), bovine serum albumin (BSA), trichloroacetic acid (TCA), 2,2′-azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), potassium persulfate, ascorbic acid, mannitol, sucrose, Hepes-KOH buffer, cysteine, β-mercaptoethanol, phosphate-buffered saline (PBS), Tris buffer, and Tricine buffer (50 mM, pH 7.5, containing 400 mM NaCl and 10 mM CaCl₂)—were obtained from Sigma-Aldrich (St. Louis, MO, USA). Protein separation and enzymatic assays employed DEAE-Sepharose Fast Flow resin (Cytiva, Uppsala, Sweden), Clostridium histolyticum collagenase (EC 3.4.23.3, ≥125 CDU/mg), and the synthetic substrate FALGPA [N-(3-[2-furyl] acryloyl)-Leu-Gly-Pro-Ala]. Electrophoresis was conducted using precast SDS–polyacrylamide gels, mPAGE™ MES SDS Running Buffer Powder (Sigma-Aldrich, St. Louis, MO, USA), 2× Laemmli sample buffer (Bio-Rad, Hercules, CA, USA), Coomassie Brilliant Blue R250 stain, and Precision Plus Protein™ Dual Color markers (all from Bio-Rad, Hercules, CA, USA).

2.2. Extraction and Characterization of Actinidin

2.2.1. Extraction Protocol

Actinidin was extracted and partially purified based on the procedure described by Lo Piero et al. [36], with minor adaptations. Fresh green kiwifruits (Actinidia deliciosa var. Hayward) were purchased from local markets in Santiago de Cali, Colombia. The fruits were homogenized in a 1:1.25 (w/v) ratio using Buffer A composed of 200 mM mannitol, 70 mM sucrose, 20 mM Hepes-KOH (pH 7.5), and 10 mM cysteine. The homogenate was filtered and subjected to centrifugation at 24,400× g for 20 min at 4 °C using a 3–18R centrifuge (Hettich, Tuttlingen, Germany). The pH of the resulting supernatant was adjusted to 4.4 with 1 M HCl, followed by a second centrifugation at 12,000× g for 10 min under identical temperature conditions using the same centrifuge. The pellet was discarded, and the clarified supernatant (crude extract) was loaded onto a DEAE-Sepharose Fast Flow column (Cytiva, Uppsala, Sweden) previously equilibrated with Buffer B (25 mM sodium acetate, pH 4.4, containing 5 mM DTT and 1 mM EDTA). Ion-exchange chromatography was performed by applying a linear NaCl gradient (0–2 M), collecting fractions in 2 mL volumes. Dialysis of eluted proteins was carried out using the Pur-A-Lyzer™ Midi Dialysis Kit (Sigma-Aldrich, St. Louis, MO, USA) according to the manufacturer’s specifications. Protein content was quantified using the bicinchoninic acid (BCA) assay (Thermo Fisher Scientific, Waltham, MA, USA), with a standard curve constructed from serial dilutions of bovine serum albumin (BSA; Thermo Fisher Scientific). Each sample (50 μL) was mixed with 1 mL of BCA reagent and incubated at 37 °C for 30 min in an incubator (Memmert, Schwabach, Germany), and absorbance was recorded at 562 nm using a Spectroquant Prove 600 UV–Vis spectrophotometer (Merck KGaA, Darmstadt, Germany).

2.2.2. SDS-PAGE Electrophoresis

The running buffer was prepared by dissolving one sachet of mPAGE™ MES SDS Running Buffer Powder (Sigma-Aldrich, St. Louis, MO, USA) in 1 L of distilled water, according to the manufacturer’s instructions, while the loading buffer was prepared by mixing 25 μL of ≥98% pure 2-mercaptoethanol (14.2 M) (Bio-Rad, Hercules, CA, USA) with 475 μL of 2× Laemmli Sample Buffer (#1610737, Bio-Rad, Hercules, CA, USA). For electrophoretic analysis, 20 μL of each eluted protein fraction—corresponding to increasing NaCl concentrations (0, 0.5, 1.0, 1.5, and 2.0 M)—was combined with an equal volume of loading buffer, gently vortexed, and then denatured at 100 °C for 5 min using a digital dry bath incubator (Benchmark Scientific, Sayreville, NJ, USA). Subsequently, 10 μL of each denatured sample was loaded onto precast mPAGE™ 8–16% Bis-Tris polyacrylamide gels and subjected to electrophoresis at 120 V for 90 min at room temperature using a Mini-PROTEAN Tetra Cell System (Bio-Rad, Hercules, CA, USA). Gels were stained with EZBlue™ Gel Staining Reagent (G1041, Sigma-Aldrich, St. Louis, MO, USA) for 1 h and destained overnight in distilled water to visualize resolved protein bands according to the manufacturer’s instructions.

2.3. Enzymatic Hydrolysis

2.3.1. Preparation of Casein Solution

To prepare a 1% (w/v) casein solution, 0.1 g of commercial bovine casein (Sigma Aldrich, CAS 9000-71-9) was dissolved in 10 mL of distilled water. The pH of the solution was adjusted to 8.0 with NaOH 0.01 M.

2.3.2. Hydrolysis Procedure

Hydrolysis was initiated by mixing 200 μL of actinidin solution (3 mg/mL) with 1.8 mL of 1% casein solution. For the control (blank), 200 μL of distilled water was combined with 1.8 mL of the same casein solution. Both mixtures were incubated in a water bath at 37 °C for 30 min. Enzymatic activity was quenched by the addition of 3 mL of 6.25% TCA, followed by centrifugation at 12,000× g for 10 min at 4 °C. Supernatants were filtered and ultrafiltered using 10 kDa membrane tubes (WVR) at 12,000× g for 15 min, isolating peptides smaller than 10 kDa. The filtrates were stored at −20 °C.

2.3.3. Degree of Hydrolysis

The degree of hydrolysis (DH%) was determined based on soluble protein content using the BCA method, following Hoyle and Merritt [37]. A total of 50 μL of the filtered supernatant from the samples was mixed with 1.0 mL of the BCA solution (mixture of 16 mL of reagent A with 320 μL of reagent B) using a vortex mixer and incubated for 30 min at 37 °C. Subsequently, the absorbance readings were taken on a Spectroquant Prove 600 UV–VIS spectrophotometer (Merck KGaA, Darmstadt, Germany) at a wavelength of 562 nm. The concentration of soluble protein was calculated by interpolation in the calibration curve obtained at casein concentrations of 0, 200, 400, 600, 800, and 1000 μg/mL. The calculation was performed by quantifying the percentage of cleaved peptide bonds in relation to the initial protein load, based on the following equation:

$$DH\%=\left({\displaystyle \frac{PSti - {PSt}_0}{P_{total}}}\right) \times 100$$

(1)

where PSt_i is the protein concentration soluble in 6.25% TCA after enzymatic hydrolysis, PSt₀ is the protein concentration soluble in 6.25% TCA before hydrolysis, and P_total is the total initial amount of casein protein in the sample.

2.4. In Vitro Anti-Aging Activity

2.4.1. Antioxidant Activity (ABTS Assay)

The antioxidant activity of the casein-derived peptide hydrolysates was evaluated using the ABTS (2,2′-azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid)) radical scavenging assay, following the methodology described by Wang et al. [38], with minor modifications. The ABTS⁺• radical was generated by mixing equal volumes of 7.0 mM ABTS and 2.45 mM potassium persulfate in ultra-pure water and incubating the mixture in darkness at room temperature for 12–16 h. The radical solution was subsequently diluted using phosphate-buffered saline (PBS; 5 mM, pH 7.4) until a final absorbance of 0.70 ± 0.02 was reached at 734 nm. For each assay, 1 mL of the pre-diluted ABTS⁺• solution was mixed with 1 mL of the peptide sample, reaching a final volume of 2 mL. The mixture was incubated for 10 min at room temperature, and the absorbance was measured at 734 nm using a UV–Vis spectrophotometer (Spectroquant Prove 600, Merck KGaA, Darmstadt, Germany). PBS buffer was used as a negative control, while L-ascorbic acid solution (6.25 mg/mL; Merck KGaA, Darmstadt, Germany) was used as a positive control.

Antioxidant capacity was determined using the percentage of ABTS⁺• inhibition given by the test peptides, indicating their radical scavenging capacity. The following equation was used:

$$ABTS \left(\%\right)= {\displaystyle \frac{A_{control}-A_{sample}}{A_{control}}} \times 100$$

(2)

where A_control is the absorbance of the ABTS⁺• solution without peptide and A_sample is the absorbance of the ABTS⁺• solution in the presence of peptides.

2.4.2. Anti-Collagenase Activity

The collagenase inhibition assay was conducted based on the method of Thring et al. [39]. Clostridium histolyticum collagenase (ChC, EC 3.4.23.3) was prepared at 0.8 U/mL in 50 mM Tricina buffer (pH 7.5) containing 400 mM NaCl and 10 mM CaCl₂ and pre-incubated for 15 min to activate enzymatic function. The substrate FALGPA was dissolved at 2 mM in Tricina buffer. The enzymatic assay was initiated by mixing 1 mL of FALGPA substrate solution with 200 μL of peptide sample (0.844 mg/mL). The reaction was initiated by the addition of collagenase, after which absorbance at 335 nm was recorded immediately. Kinetic measurements were then conducted at 30 s intervals for 20 min. Collagenase inhibition was determined based on the reduction in substrate hydrolysis rate in the presence of peptides, calculated in accordance with the protocol provided by the manufacturer using the following equation:

$$\% \mathrm{C}\mathrm{o}\mathrm{l}\mathrm{l}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{ }\mathrm{i}\mathrm{n}\mathrm{h}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}={\%A substrate}_{sample}-{\%A substrate}_{blank}$$

(3)

where ${\%A substrate}_{sample}$ refers to the percent change in absorbance of the peptide-treated mixture over 20 min, while ${\%A substrate}_{blank}$ denotes the percent change in absorbance of the untreated control over the same time. The percentage change is derived from the difference in absorbance at minute 0 and minute 20 for each condition.

2.4.3. Anti-Elastase Activity

Elastase inhibition was evaluated following Nam et al. [40]. Peptide samples (0.844 mg/mL) were mixed in a 1:4 ratio with 0.4 mM of the elastase-specific substrate N-succinyl-trialanine-p-nitroanilide (N-Suc-(Ala)₃-pNA) in 1 M Tris buffer (pH 8.0). The mixture was incubated at 25 °C for 20 min. Afterward, elastase from porcine pancreas (PPE) was added, and absorbance was measured immediately at 410 nm, with readings taken every 30 s for 20 min. Thus, the ability of the peptide to block elastase activity by comparing substrate degradation rates in treated versus control samples was determined by means of the following equation:

$$\% \mathrm{E}\mathrm{l}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{ }\mathrm{i}\mathrm{n}\mathrm{h}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} = [1 - (\mathsf{\Delta }{A}_{410\mathit{sample}}/\mathsf{\Delta }{A}_{410\mathit{control}}\left)\right] \times 100$$

(4)

where $\Delta A_{410\mathit{sample}}$ is the change in absorbance at 410 nm over time in the presence of peptides and $A_{410\mathit{control}}$ is the change in absorbance under control conditions without peptides.

2.5. Computational Evaluation of Anti-Aging Peptides from Casein Hydrolysates

2.5.1. Protein and Ligand Preparation

Molecular docking analyses were performed using Discovery Studio Visualizer to evaluate the binding affinity between ten selected peptides—derived from in silico hydrolysis of casein proteins—and two target enzymes: collagenase (PDB ID: 4U7K) and elastase (PDB ID: 1QR3). The hydrolysates were obtained from the cleavage at the carboxyl side of a lysine residue [19] of casein proteins: CSN1S1 (Alpha-S1-casein, UniProt ID P02662), CSN1S2 (Alpha-S2-casein, UniProt ID P02663), CSN2 (Beta-casein, UniProt ID P02666), and CSN3 (Kappa-casein, UniProt ID P02668). Peptides enriched in polar and aromatic residues with lengths between 4 and 8 amino acids were docked according to the suggestions of Rama et al. [41], and interaction energy was quantified using Gibbs free energy (ΔG), expressed in kcal/mol, where lower values indicate stronger binding affinity. The selected peptides were energetically optimized using Avogadro software (version 1.2.0), employing the minimum energy conformers. The removal of water molecules and non-relevant heteroatoms from peptides and proteins was performed using PyMOL v3.0.4 (The PyMOL Molecular Graphics System, Schrödinger, LLC, New York, NY, USA). The enzymes (receptors) were prepared in AutoDock Tools (ADT 1.5.7,) preserving the polar hydrogens and assigning Kollman charges while the peptides (ligands) were assigned Gasteiger charges.

2.5.2. Docking Method and Validation

Each peptide–enzyme complex was analyzed using AutoDock Vina (version 1.2.7). A blind docking approach was employed, allowing the entire receptor surface to be explored for potential binding sites. Two well-characterized cosmetic inhibitors—epigallocatechin gallate (EGCG) for collagenase and FR901277 for elastase—were obtained from the Drugbank database (Accession Number DB12116) and Protein Data Bank (PDB ID: 1QR3), respectively. Both were included as positive controls to benchmark interaction profiles.

2.5.3. Docking Analysis

Subsequently, each complex was analyzed considering three types of molecular interactions: hydrogen bonds (HB), which involve donor–acceptor interactions primarily between electronegative atoms; hydrophobic interactions (HI), including alkyl and π-alkyl contacts stabilizing nonpolar regions; and electrostatic interactions (EI), referring to ionic or charged residue pairing. Binding affinity was determined by calculating ΔG using the expression ΔGbinding = Ecomplex − (Eprotein + Epeptide), where Ecomplex is the minimized energy of the peptide–enzyme complex and Eprotein and Epeptide represent the isolated energies of the enzyme and peptide, respectively. This docking strategy enabled the comparative assessment of theoretical peptide affinity; the dimensions of the docking grid box used for each enzyme were as follows: for Collagenase (PDB ID: 4U7K), center_x = 1.155, center_y = −18.899, center_z = 41.062; size_x = 116, size_y = 120, size_z = 106. For Elastase, the parameters were center_x = −6.544, center_y = 27.781, center_z = 43.294; size_x = 126, size_y = 126, and size_z = 126. This docking strategy considered molecular interactions and binding affinity as indicators to predict potential anti-aging activities.

2.5.4. Visualization

Visualization and analysis of docking complexes were performed using BIOVIA Discovery Studio Visualizer v17.2.0 (BIOVIA, Dassault Systèmes, 2016, San Diego, CA, USA) [42].

2.5.5. Web Tools and Software Utilized

Publicly accessible online databases were used to obtain reference sequences and structures: UniProt (https://www.uniprot.org, accessed 18 December 2024), Protein Data Bank (PDB) (https://www.rcsb.org, accessed 18 December 2024) and DrugBank (https://go.drugbank.com, accessed 18 December 2024). The analyses performed with Avogadro, PyMOL, AutoDock Tools, AutoDock Vina, and BIOVIA Discovery Studio Visualizer were carried out locally on a PC using their corresponding academic licenses.

3. Results and Discussion

3.1. Electrophoretic and Structural Characterization of Actinidin

The electrophoretic profile and structural model of the purified enzyme are shown in Figure 1. These results confirm the successful isolation of actinidin and validate its estimated molecular weight, which agrees with previously reported crystallographic data [43,44]. SDS-PAGE analysis revealed a distinct protein band between 25 and 37 kDa in the 0.5 M NaCl eluate (Figure 1, left panel), closely migrating near the 25 kDa marker and suggesting an approximate molecular weight of 27 kDa. This observation is consistent with the findings recorded for Dhiman et al. [45]. Additionally, the structural model derived from crystallographic coordinates (PDB ID: 1AEC) exhibited a compact globular conformation with an estimated mass of 23.9 kDa (Figure 1, right panel), thereby reinforcing the electrophoretic evidence. Collectively, the biochemical and structural analyses confirm that the protein eluted at 0.5 M NaCl corresponds to actinidin, rendering it suitable for downstream hydrolysis experiments.

3.2. Degree of Hydrolysis Assays

The degree of hydrolysis refers to the percentage of peptide bonds cleaved in a protein hydrolysate, reflecting the release of soluble peptides generated during enzymatic breakdown. This parameter plays a key role in determining the biochemical and bioactive properties of hydrolyzed proteins [46]. Based on the linear equation derived from the BSA standard curve (R² > 99.5%), the calculated hydrolysis percentage (DH%) was 91.6%, suggesting that only 8.4% of the casein remained uncleaved under the enzymatic model applied. A DH of 90% implies extensive proteolysis, with significant alterations in the functional properties of the native protein structure [47]. These results are consistent with observations reported by Dhiman et al. [45], where actinidin extracted from Actinidia deliciosa was shown to reduce α-amylase activity by 95%, leaving only 5% of the enzyme active. In another study, Alizadeh and collaborators (2005) reported enzymatic hydrolysis degrees up to 93% using actinidin in the conversion of glucans into fermentable sugars for subsequent biofuel production [48]. Similarly, previous research using corn forage reported hydrolysis degrees of 71% and 77% for plant biomass, concluding that pretreatment significantly enhanced hydrolysis yields [49]. It is important to highlight that a comprehensive review of the literature revealed no explicit reports on the degree of hydrolysis for casein treated with actinidin from green kiwi, limiting direct comparison with the present findings. However, in a study by Kaur et al. [50], actinidin was used to hydrolyze whey protein isolate (WPI) and milk protein concentrate (MPC) to reduce the immunoreactivity of β-lactoglobulin (β-LG) and αs1-casein (αs1-CN), achieving hydrolysis levels below 15% under the evaluated conditions. Another study described the hydrolysis of bovine casein using trypsin and pepsin at different time intervals, achieving hydrolysis percentages of 83.3% and 51.4%, respectively, after one hour of incubation [51]. Several factors—including enzyme concentration, protein purity and origin, substrate structure, and the methodology employed—may influence the hydrolysis degree obtained. Differences across assays are primarily attributable to the enzyme–substrate ratio, as well as the natural origin of both casein and the protease. Variability in the results may be related to incubation time, pH, and temperature, which all affect enzymatic catalysis. Altogether, these parameters collectively modulate the efficiency of enzymatic cleavage and define the peptide release profile within a given hydrolysis timeframe [52].

3.3. In Vitro Anti-Aging Activity Assays

To assess the biological relevance of the casein-derived peptides obtained through enzymatic hydrolysis with actinidin, a comparative analysis was conducted against benchmark compounds and previously reported hydrolysates documented in the indexed literature (Table 1). This comparison helped assess the strength and consistency of the antioxidant, anti-collagenase, and anti-elastase activities, while also offering insights into the possible reasons behind their variability.

In terms of antioxidant capacity, the ABTS⁺• radical-scavenging activity measured in this study was 17.5 ± 4.25% at a concentration of 0.85 mg/mL. This value is significantly lower than the inhibition observed for casein peptides hydrolyzed with trypsin (84.05% at ~3 mg/mL) [40] and for epigallocatechin gallate (EGCG), a green tea polyphenol that reached 78.5% inhibition at just 0.1 mg/mL [52]. κ-Casein hydrolysates derived from ovine milk using pepsin/trypsin also demonstrated ~60% activity [53]. The reduced antioxidant potential in our study may be explained from two points of view: (i) the composition of the peptide extract as a complex mixture of several peptides where those exhibiting antioxidant potential can be found in reduced concentrations and (ii) reduced exposure of amino acid side chains that participate in electron transfer reactions due to structural folding, especially in moderate-sized sequences above 10 amino acids [10]. Thus, the absence of molecular fractionation could have diluted highly active peptide sequences to sub-detectable levels. Peptides smaller than 3 kDa have been shown to exhibit greater radical-scavenging potential due to enhanced solubility and accessibility of reactive groups [54]. Nevertheless, despite its modest activity, the hydrolysate obtained under non-optimized conditions reflects the versatility of actinidin as a biocatalyst, capable of generating bioactive fragments suitable for further purification and potential application in skin-care formulations.

Table 1. Comparative in vitro anti-aging activities of casein-derived peptides and reference compounds.

Bioactivity	Source/Compound	% Inhibition	Applied Concentration	Reference
Antioxidant (ABTS+•)	Casein-derived peptides (actinidin hydrolysis)	17.5 ± 4.25%	0.85 mg/mL	This study
	Casein-derived peptides (trypsin hydrolysis)	84.05%	~ 3 mg/mL	Mokhtari et al. (2023) [46]
	Epigallocatechin gallate-EGCG (green tea polyphenol)	78.5%	0.1 mg/mL	González-Alfonso et al. (2019) [52]
	κ-Casein hydrolysate (ovine, pepsin/trypsin)	~60%	Not specified	Gómez-Ruiz et al. (2008) [53]
Anti-Collagenase	Casein peptides (actinidin hydrolysis)	18.55 ± 4.63%	0.844 mg/mL	This study
	Protein hydrolysates (PH) derived from Acheta domesticus	60.23%	Optimized via RSM	Yeerong et al. (2024) [55]
	Epigallocatechin gallate-EGCG (green tea polyphenol)	89.5%	0.1 mM	Madhan et al. (2007) [56]
	Casein peptides (trypsin hydrolysis)	Not quantified	~3 mg/mL	Mokhtari et al. (2023) [46]
Anti-Elastase	Casein peptides (actinidin hydrolysis)	28.66 ± 7.2%	0.844 mg/mL	This study
	Chia seed peptides (<3 kDa)	65.32%	0.43 mg/mL	Aguilar & Liceaga (2020) [57]
	Porcine skin peptides (1–3 kDa)	22.22%	Not specified	Hong et al. (2019) [58]
	White tea extract	89%	Not specified	Thring et al. (2009) [39]

Percentages indicate inhibition of ABTS⁺• radicals, collagenase, and elastase. Literature values were selected from studies using comparable assay conditions. Results underscore the moderate yet promising multifunctional potential of casein peptides for cosmetic use.

Table 1. Comparative in vitro anti-aging activities of casein-derived peptides and reference compounds.

Bioactivity	Source/Compound	% Inhibition	Applied Concentration	Reference
Antioxidant (ABTS+•)	Casein-derived peptides (actinidin hydrolysis)	17.5 ± 4.25%	0.85 mg/mL	This study
	Casein-derived peptides (trypsin hydrolysis)	84.05%	~ 3 mg/mL	Mokhtari et al. (2023) [46]
	Epigallocatechin gallate-EGCG (green tea polyphenol)	78.5%	0.1 mg/mL	González-Alfonso et al. (2019) [52]
	κ-Casein hydrolysate (ovine, pepsin/trypsin)	~60%	Not specified	Gómez-Ruiz et al. (2008) [53]
Anti-Collagenase	Casein peptides (actinidin hydrolysis)	18.55 ± 4.63%	0.844 mg/mL	This study
	Protein hydrolysates (PH) derived from Acheta domesticus	60.23%	Optimized via RSM	Yeerong et al. (2024) [55]
	Epigallocatechin gallate-EGCG (green tea polyphenol)	89.5%	0.1 mM	Madhan et al. (2007) [56]
	Casein peptides (trypsin hydrolysis)	Not quantified	~3 mg/mL	Mokhtari et al. (2023) [46]
Anti-Elastase	Casein peptides (actinidin hydrolysis)	28.66 ± 7.2%	0.844 mg/mL	This study
	Chia seed peptides (<3 kDa)	65.32%	0.43 mg/mL	Aguilar & Liceaga (2020) [57]
	Porcine skin peptides (1–3 kDa)	22.22%	Not specified	Hong et al. (2019) [58]
	White tea extract	89%	Not specified	Thring et al. (2009) [39]

Percentages indicate inhibition of ABTS⁺• radicals, collagenase, and elastase. Literature values were selected from studies using comparable assay conditions. Results underscore the moderate yet promising multifunctional potential of casein peptides for cosmetic use.

In terms of anti-collagenase activity, peptides from this study exhibited 18.55 ± 4.63% inhibition at 0.844 mg/mL, which is notably lower than the 60.23% reported for Acheta domesticus-derived hydrolysates [55] and markedly lower than the 89.5% inhibition observed with EGCG at 0.1 mM [56]. Trypsin-generated casein peptides also demonstrated collagenase inhibition in the study by [46]; however, no quantitative data were reported. These differences are likely due to variations in hydrolysis parameters, enzyme affinity, and peptide profile. Proteases such as alcalase or trypsin typically produce shorter peptides with enhanced access to the catalytic pocket of matrix metalloproteinases, while actinidin tends to generate longer chains with lower steric compatibility. The literature has consistently linked enhanced inhibition to peptides ranging from 7 to 16 amino acid residues, particularly those containing hydrophobic or aromatic side chains, which facilitate binding to metalloproteinase domains. Therefore, peptide size and sequence-specific binding interactions are key factors in achieving meaningful enzymatic suppression.

For elastase inhibition, the actinidin-hydrolyzed casein peptides produced 28.66 ± 7.2% activity at 0.844 mg/mL, a moderate value when compared to reference compounds. Chia seed peptides (<3 kDa), observed in Aguliar & Liceaga [57], reached 65.32% inhibition, whereas porcine skin collagen hydrolysates exhibited 22.22% inhibition in the <3 kDa fraction [58]. White tea extracts showed the highest inhibition at 89% [39], though the mechanism is primarily polyphenol-driven rather than peptide-based. The variability across these studies may be attributed to differences in peptide composition, molecular weight distribution, and experimental assay conditions. In general, inhibition of elastase is favored by small peptides with hydrophobic residues capable of interacting with the enzyme. Although the hydrolysates were filtered using a 10 kDa cutoff membrane, additional fractionation by molecular size may uncover smaller, more bioactive peptides with enhanced inhibitory potential.

Taken together, the antioxidant, anti-collagenase, and anti-elastase activities observed in this work demonstrate that casein-derived peptides generated by actinidin possess a multifunctional bioactivity profile suitable for dermo-cosmetic applications. While the values obtained are lower than those from optimized enzymatic systems or highly purified compounds, the results emphasize the feasibility of employing a natural enzyme system to produce physiologically relevant peptide mixtures. Moreover, this strategy offers a cost-effective and sustainable route for developing anti-aging ingredients from food-grade substrates.

To advance understanding of the molecular basis behind these activities and identify specific inhibitory peptide sequences, the next section presents a complementary in silico strategy. This computational approach integrates enzymatic cleavage simulation and molecular docking to predict sequence-specific binding affinities to collagenase and elastase, thus validating and expanding the biological potential of the hydrolysates characterized experimentally.

3.4. Computational Evaluation of Anti-Aging Peptides from Casein Hydrolysates Results

3.4.1. In Silico Hydrolysis of Casein Proteins

In silico hydrolysis of the four main bovine casein proteins—CSN1S1, CSN1S2, CSN2, and CSN3—was performed to complement the experimental enzymatic digestion. The endoprotease Lys-C (EC 3.4.21.50) was selected based on its specific cleavage at the carboxyl side of lysine residues, which agrees with the known substrate recognition preferences of actinidin. The simulation was conducted to generate peptide libraries and estimate theoretical hydrolysis efficiency. The peptide profiles obtained, along with the predicted degree of hydrolysis, are presented in

Table 2. Summary of peptides generated by in silico hydrolysis of bovine casein proteins using Lys-C. The number of peptides between 4 and 8 residues has been highlighted due to their biological relevance for enzymatic interaction and docking simulations. A theoretical hydrolysis degree of 100% was obtained for all protein sequences evaluated.

Protein (UniProt ID)	Total Peptides	Peptides of 4–8 Amino Acids	Theoretical Degree of Hydrolysis (%DHt)
CSN1S1 (P02662)	16	5	100%
CSN1S2 (P02663)	26	10	100%
CSN2 (P02666)	13	3	100%
CSN3 (P02668)	11	2	100%
Total	66	20 (30.3%)	—

.

The simulation results showed that complete theoretical hydrolysis (100% DHt) was achieved for all four protein chains, yielding a total of 66 peptide fragments. Among these, 20 peptides (30.3%) ranged from four to eight amino acids in length—an interval recognized in the literature as optimal for biological activity and enzyme interaction. CSN1S2 yielded the highest number of total peptides as well as short-length fragments, followed by CSN1S1. This trend was attributed to the greater lysine content and accessible sequence regions within these proteins.

The in silico data were found to be consistent with the experimental hydrolysis carried out using actinidin, which reached a degree of hydrolysis of 91.6%. The 7% deviation between simulated and experimental values may be explained by variables such as enzyme purity, steric hindrance, and kinetics under actual conditions. Prior studies involving actinidin-mediated hydrolysis of whey or milk proteins under non-optimized settings have reported degrees of hydrolysis below 15% [50], underscoring the improved cleavage efficiency achieved in this work.

Moreover, the predominance of short peptides in the theoretical dataset mirrored the <10 kDa filtration step applied during peptide recovery, reinforcing their relevance for subsequent molecular docking. Overall, the computational digestion not only validated the cleavage selectivity observed in vitro, but also produced a targeted peptide library enriched in polar and aromatic residues—features that are strongly associated with antioxidant and enzyme-inhibitory properties in the context of cosmetic peptide applications.

3.4.2. Molecular Docking

The molecular docking results of the ten casein-derived peptides with collagenase and elastase are presented in

Table 3. Results of molecular docking between peptides and the enzymes collagenase and elastase.

Enzyme	Peptide	Hydrogen Bond *	Electrostatic Bond *	Hydrophobic Bond *	ΔG kcal/mol
Collagenase	FALPQYLK	Ligand:PHE1 – A:GLU702 Ligand:PHE1 – A:THR699 Ligand:GLN5 – A:LYS697 Ligand:GLN5 – A:THR695 Ligand:GLN5 – A:TYR696 Ligand:TYR6 – A:LYS705 Ligand:TYR6 – A:GLU705 Ligand:LYS8 – A:THR695 Ligand:LYS8 – A:GLY694	Ligand:PHE1 – A:GLU702	Ligand:PHE1 – A:GLU702 Ligand:TYR6 – A:LYS705 Ligand:TYR6 – A:ILE704	−5.493
	HYQK	Ligand:HIS1 – A:GLU709 Ligand:TYR2 – A:GLU737 Ligand:GLN3 – A:LYS705 Ligand:GLN3 – A:GLU702 Ligand:LYS4 – A:THR695	Ligand:HIS1 – A:GLU709	Ligand:LYS4 – A:LYS705	−5.229
	TTMPLW	Ligand:THR1 – A:GLU737 Ligand:THR1 – A:ASN735 Ligand:THR2 – A:SER707 Ligand:TRP6 – A:GLU702 Ligand:TRP6 – A:LYS703 Ligand:TRP6 – A:LYS697 Ligand:TRP6 – A:LYS705	Ligand:THR1 – A:GLU737 Ligand:TRP6 – A:LYS705	Ligand:TRP6 – A:LYS705	−4.949
	HPIK	Ligand:ILE3 – A:LYS705 Ligand:LYS4 – A:ASN693 Ligand:LYS4 – A:THR695 Ligand:HIS1 – A:PHE706 Ligand:HIS1 – A:ASN735	Ligand:HIS1 – A:GLU709 Ligand:HIS1 – A:GLU737	Ligand:HIS1 – A:LYS705 Ligand:ILE3 – A:ILE704 Ligand:ILE3 – A:LYS705	−4.796
	HIQK	Ligand:HIS1 – A:ASN735 Ligand:HIS1 – A:SER707 Ligand:ILE2 – A:LYS705 Ligand:GLN3 – A:LYS705 Ligand:GLN3 – A:LYS697 Ligand:LYS4 – A:THR695	-	-	−4.704
	VIPYVRYL	Ligand:TYR4 – A:GLU734 Ligand:TYR6 – A:GLU733 Ligand:TYR6 – A:ASN735 Ligand:LEU7 – A:GLU734	Ligand:TYR4 – A:GLU734	Ligand:PRO3 – A:ILE718	−4.577
	EAMAPK	Ligand:GLU1 – A:GLU734 Ligand:GLU1 – A:SER711 Ligand:GLU1 – A:SER708 Ligand:ALA2 – A:TYR721 Ligand:MET3 – A:TYR721 Ligand:PRO5 – A:VAL719 Ligand:PRO5 – A:GLU734 Ligand:LYS6 – A:VAL719	Ligand:GLU1 – A:GLU734	Ligand:ALA2 – A:ILE718 Ligand:MET3 – A:LYS717 Ligand:PRO5 – A:TYR721 Ligand:LYS6 – A:VAL719	−4.452
	EMPFPK	Ligand:GLU1 – A:GLU709 Ligand:GLU1 – A:SER707 Ligand:MET2 – A:SER707 Ligand:LYS6 – A:LYS705	Ligand:GLU1 – A:GLU709	Ligand:MET2 – A:LYS705 Ligand:PHE4 – A:LYS69	−4.372
	ISQRYQK	Ligand:GLN6 – A:SER707 Ligand:TYR5 – A:ASN693 Ligand:TYR5 – A:LYS705 Ligand:TYR5 – A:TYR696 Ligand:TYR5 – A:SER707 Ligand:GLN6 – A:PHE706 Ligand:GLN6 – A:ASN735	Ligand:ARG4 – A:GLU709	Ligand:TYR5 – A:LYS705 Ligand:TYR5 – A:ILE704 Ligand:PRO5 – A:ILE704	−4.369
	EGIHAQQK	Ligand:GLU1 – A:VAL761 Ligand:ILE3 – A:SER762 Ligand:HIS4 – A:GLY690 Ligand:HIS4 – A:SER760 Ligand:ALA5 – A:VAL761 Ligand:GLN6 – A:VAL761 Ligand:GLN7 – A:LEU687 Ligand:GLN7 – A:PRO688 Ligand:LYS8 – A:TYR689	-	Ligand:HIS4 – A:SER762	−4.338
Elastase	VIPYVRYL	Ligand:TYR4 – A:THR96 Ligand:TYR7 – A:TYR35 L igand:LEU8 – A:SER36	-	Ligand:VAL1 – A:VAL217 Ligand:PRO3 – A:CYS42 Ligand:ARG6 – A:TYR35 Ligand:LEU8 – A:LEU151	−6.698
	ISQRYQK	Ligand:TYR5 – E:CYS191 Ligand:GLN6 – E:CYS42 Ligand:GLN6 – E:THR41	Ligand:ARG61 – A:LYS7	Ligand:ILE1 – E:PHE215 Ligand:TYR5 – E:VAL216	−6.286
	HYQK	Ligand:GLN3 – E:CYS191	Ligand:HIS1 – A:TYR2 Ligand:LYS4 – A:LYS4 Ligand:LYS4 – E:HIS57	Ligand:HIS1 – E:LEU143	−6.012
	HPIK	Ligand:HIS1 – E:CYS191 Ligand:LYS4 – E:THR41	-	Ligand:HIS1 – E:VAL216 Ligand:PRO2 – E:HIS57 Ligand:PRO2 – E:VAL99	−5.772
	FALPQYLK	Ligand:GLN5 – E:CYS191 Ligand:GLN5 – E:VAL216 Ligand:TYR6 – E:ARG217 Ligand:LYS8 – E:CYS58 Ligand:LYS8 – E:ARG61	-	Ligand:TYR6 – E:VAL99	−5.759
	HIQK	Ligand:HIS1 – E:HIS57 Ligand:HIS1 – E:SER195 Ligand:ILE2 – E:HIS57 Ligand:ILE2 – E:SER195 Ligand:GLN3 – E:GLY193	-	Ligand:HIS1 – E:VAL216 Ligand:ILE2 – E:HIS57 Ligand:LYS4 – E:LEU143 Ligand:LYS4 – E:LEU151	−5.669
	EGIHAQQK	Ligand:GLU1 – E:GLN23 Ligand:GLU1 – E:SER26 Ligand:GLU1 – E:GLN157 Ligand:GLY2 – E:SER26 Ligand:ALA5 – E:TYR207 Ligand:GLN6 – E:SER29 Ligand:GLN6 – E:VAL122 Ligand:GLN7 – E:GLN206	-	Ligand:ALA5 – E:TYR207 Ligand:LYS8 – E:VAL122	−5.661
	TTMPLW	Ligand:THR2 – E:TYR101 Ligand:THR2 – E:ALA99 Ligand:MET3 – E:TYR101 Ligand:TRP6 – E:SER232 Ligand:TRP6 – E:SER236		Ligand:LEU5 – E:HIS91 Ligand:LEU5 – E:TYR93 Ligand:LEU5 – E:TYR101 Ligand:PRO4 – E:ALA233 Ligand:TRP6 – E:ALA126	−5.531
	EMPFPK	Ligand:GLU1 – E:ASN25 Ligand:GLU1 – E:GLN119 Ligand:MET2 – E:SER26 Ligand:PHE4 – E:SER26 Ligand:LYS6 – E:ASN25	-	Ligand:PRO5 – E:TRP27 Ligand:PRO5 – E:TYR137 Ligand:PRO5 – E:TYR207	−5.143
	EAMAPK	Ligand:GLU1 – E:ARG217 Ligand:ALA4 – E:ARG217 Ligand:LYS6 – E:ARG217 Ligand:LYS6 – E:GLN192	Ligand:GLU1 – E:LYS177	Ligand:ALA2 – E:ARG217 Ligand:ALA4 – E:VAL99 Ligand:MET3 – E:VAL99 Ligand:LYS6 – E:ARG217	−4.950

* A is the chain of the enzyme (collagenase or elastase).

.

Table 3 shows consistent patterns in the molecular interactions of the amino acids present in the analyzed peptides. Electrostatic interactions play a key role in the formation of hydrogen bonds (HB); when they are strong enough, they favor the generation of multiple hydrogen bonds, which was predominant in most peptides. In particular, residues such as lysine and arginine confirm that their cationic groups can interact with negatively charged residues (Asp and Glu) of the enzyme [59]. In addition, aromatic residues such as phenylalanine, tyrosine, and tryptophan are essential for the three-dimensional structure of proteins, and their interaction modification can significantly alter protein function [60] which could generate an inhibition in collagenase and elastase enzymes.

In relation to Gibbs free energy, FALPQYLK was identified as the most affine peptide (ΔG = −5.493 kcal/mol), stabilized by 12 hydrogen bonds, 3 hydrophobic interactions, and 1 electrostatic contact, primarily involving Leu, Tyr, and Lys residues. HYQK followed with a ΔG of −5.229 kcal/mol, exhibiting seven hydrogen bonds and four electrostatic interactions, suggesting polarity-driven anchoring. **TTMPLW** displayed a moderate affinity (ΔG = −4.949), forming seven hydrogen bonds, five hydrophobic contacts, and two electrostatic interactions, facilitated by Trp and Leu. HPIK, with a ΔG of −4.796, showed eight hydrogen bonds and three hydrophobic and two electrostatic interactions, consistent with Pro- and His-mediated folding. In contrast, HIQK (ΔG = −4.704) presented a unique profile dominated by 10 hydrophobic interactions without hydrogen bonding, implying surface burial mechanisms. VIPYVRYL, though more affine to elastase, recorded a ΔG of −4.577, supported by eight hydrogen bonds and one hydrophobic and one electrostatic interaction. EAMAPK and EMPFPK yielded ΔG values of −4.452 and −4.372, respectively; both formed numerous hydrogen bonds (11 and 6) and moderate hydrophobic interactions, stabilized by Ala, Pro, and Lys residues. ISQRYQK exhibited a ΔG of −4.369, forming an extensive network of 15 hydrogen bonds and 3 hydrophobic and 3 electrostatic contacts. Finally, EGIHAQQK showed the lowest energy (ΔG = −4.338) yet formed 14 hydrogen bonds and 2 hydrophobic interactions. Overall, peptides with higher ΔG values (less negative) tended to form abundant hydrogen bonding but fewer hydrophobic and electrostatic contacts, whereas the most stable complexes—such as FALPQYLK and HYQK—combined bonding diversity with optimal residue positioning within the catalytic pocket, highlighting the importance of interaction type rather than quantity alone.

With respect to elastase, VIPYVRYL exhibited the strongest affinity among all evaluated peptides (ΔG = −6.698 kcal/mol), supported by three hydrogen bonds and four hydrophobic contacts, prominently mediated by Tyr, Val, and Arg residues. ISQRYQK followed closely (ΔG = −6.286 kcal/mol), forming 12 hydrogen bonds, 4 hydrophobic interactions, and 3 electrostatic contacts, suggesting robust polar and charged complementarity. HYQK (ΔG = −6.012 kcal/mol) displayed three hydrogen bonds and three hydrophobic contacts, consistent with compact polar anchoring. HPIK and FALPQYLK recorded ΔG values of −5.772 kcal/mol and −5.759 kcal/mol, respectively, forming between 12 and 3 hydrogen bonds, with moderate hydrophobic contributions. HIQK (ΔG = −5.669 kcal/mol) established five hydrogen bonds and five hydrophobic contacts, implicating Ile and Gln in hydrophobic modulation. EGIHAQQK (ΔG = −5.661 kcal/mol) formed 16 hydrogen bonds and 3 hydrophobic interactions, indicating rich polar engagement despite moderate energy stability. TTMPLW demonstrated balanced interactions (ΔG = −5.531 kcal/mol), with eight hydrogen bonds and seven hydrophobic contacts favored by Trp and Leu. EMPFPK and EAMAPK completed the series with ΔG values of −5.143 and −4.950 kcal/mol, respectively, both forming moderate networks of hydrogen bonds (10 and 7), hydrophobic contacts, and limited electrostatic interaction. Notably, the peptides with the most negative ΔG values—VIPYVRYL and ISQRYQK—combined hydrophobic anchoring and directional hydrogen bonding, reinforcing that interaction specificity and distribution are more determinant of binding energy than bond count alone. These docking patterns highlight distinct stabilization strategies for elastase inhibition, supporting the structural viability of these peptides as functional ingredients in anti-aging cosmetic formulations. Based on the docking data presented, the peptides FALPQYLK and VIPYVRYL were selected for structural analysis as they showed the highest binding affinities toward collagenase and elastase, respectively. In this way, non-covalent interactions, particularly hydrogen bonds and hydrophobic and electrostatic interactions, play a fundamental role not only in maintaining the three-dimensional structure of biomolecules, but also in the process of molecular recognition and affinity with the target protein [61].

On the other hand, Figure 2 displays the spatial arrangement of these peptides within the interaction sites of their corresponding enzymes. For comparison, the reference inhibitors EGCG (collagenase) and FR901277 (elastase) are also shown, allowing a visual assessment of binding orientation, residue-level interactions, and overall complementarity that may explain the observed energy values.

Analysis of interactions in 2D and 3D projections of peptide–enzyme and control–enzyme complexes (Figure 2A–D) allows visualizing binding sites and the spatial orientation of inhibitors at the interaction site, as well as comparing their likely inhibitory efficacy based on binding energy and docking-predicted interactions. This approach constitutes a fundamental procedure for the rational selection of bioactive candidates prior to experimental validation, as reported by Kang et al., [60]. In Figure 2A, the docking conformation of the active peptide FALPQYLK within the catalytic cleft of collagenase is shown alongside the control inhibitor EGCG. The peptide adopts an elongated structure that inserts deeply into the enzyme’s interaction site, which is zinc-dependent, as characterized in matrix metalloproteinases like MMP-1. Stabilization is achieved through 12 hydrogen bonds, 3 hydrophobic interactions, and 1 electrostatic contact, primarily involving Leu, Tyr, and Lys residues. These interactions are distributed around the zinc-coordinated catalytic center, suggesting that FALPQYLK may sterically hinder substrate access and interfere with enzymatic turnover. In contrast, EGCG—while exhibiting a more favorable ΔG (−7.087 kcal/mol)—binds more superficially, forming only three hydrogen bonds and two hydrophobic contacts, with limited penetration into the catalytic groove. This difference in spatial orientation and contact diversity highlights the peptide’s potential to mimic native substrate alignment and act as a competitive inhibitor.

In Figure 2B, the active peptide VIPYVRYL is shown docked with elastase, where it forms three hydrogen bonds and two hydrophobic interactions, primarily through Tyr, Val, and Arg residues. These contacts are positioned near the enzyme’s catalytic triad (Ser195–His57–Asp102), indicating strong complementarity and potential for direct inhibition. The control inhibitor FR901277, although more energetically stable (ΔG = −8.812 kcal/mol), exhibits a more rigid conformation and a broader interaction network—six hydrogen bonds, three hydrophobic contacts, and one electrostatic interaction—distributed across surface-exposed loops. In contrast, VIPYVRYL shows a more flexible and adaptive geometry, allowing targeted engagement with substrate-recognition regions. This spatial arrangement may enhance its functional integration into topical formulations, especially where dynamic binding and sustained inhibition are desirable.

Together, the structural models in Figure 2 emphasize that the active peptides not only form meaningful interactions but also adopt conformations that are consistent with known inhibitory mechanisms. Their ability to engage catalytic residues, occupy key binding pockets, and mimic substrate orientation supports their potential as bioactive agents in anti-aging cosmetic applications.

In this way, the molecular docking and structural visualization results offer complementary insights into the inhibitory potential of casein-derived peptides against aging-related enzymes. The active peptides FALPQYLK and VIPYVRYL demonstrated well-defined binding profiles and spatial conformations within the catalytic sites of collagenase and elastase, respectively, exhibiting contact patterns comparable to those of standard cosmetic inhibitors. While EGCG and FR901277 showed higher binding energies overall, the casein peptides displayed notable advantages in terms of residue-level engagement, depth of accommodation, and conformational adaptability. These structural features, combined with the biochemical and computational data presented earlier, provide a solid foundation for considering actinidin-hydrolyzed casein peptides as promising multifunctional agents in topical anti-aging formulations.

This study integrates in vitro and in silico analysis of the anti-aging potential of peptides, which constitutes a first exploratory approach with casein-derived hydrolysates. The design is adequate for the aim of the research since a high yield of enzymatic hydrolysis was obtained from an enzyme extracted and purified in laboratory scale. Subsequently, a moderate activity was determined using established and reliable enzymatic models that provide preliminary in vitro information on the anti-aging potential of the peptides. In addition, an exploratory predictive analysis is added to advance the recognition and identification of possible candidate sequences with high collagenase and elastase inhibitory activity. In this way, conclusive results confirm the presence of anti-aging peptides in this extract. This leaves a precedent to continue with future studies focused on the purification and exploration of this extract, including methods such as LC/MS. Furthermore, the sequences predicted by docking can be prepared by chemical synthesis and their performance evaluated in assays with in vitro cell lines and in vivo models, providing a more robust validation framework for the eventual development of cosmetic formulations.

4. Conclusions

This study demonstrated that enzymatic hydrolysis of bovine casein using actinidin from Actinidia deliciosa yields peptide hydrolysates with multifunctional cosmetic properties. A high hydrolysis degree (91.6%) confirmed actinidin’s efficiency under mild reaction conditions, while the resulting peptides exhibited measurable antioxidant, anti-collagenase, and anti-elastase activities. In silico hydrolysis via Lys-C predicted a diverse peptide library enriched in short fragments (4–8 amino acids), consistent with reported bioactive length ranges. Molecular docking identified sequences such as FALPQYLK and VIPYVRYL with strong affinities and interaction profiles at the catalytic sites of collagenase and elastase, respectively. Structural modeling revealed conformations comparable to commercial inhibitors, suggesting competitive and substrate-mimicking inhibition mechanisms. These findings highlight the potential of peptides derived from casein hydrolysis by actinidin as a new alternative of sustainable and effective active ingredients for use in topical formulations as adjuvants in anti-aging treatment. Future studies should focus on in vivo validation, skin permeation assays, and advanced peptide characterization to guide formulation scalability and clinical relevance.