EUVIMEDEUVIMED entdecken
Crossref

Innovative Computational Design of Sustainable Milk-Clotting Peptides for Enhanced Cheese Production

Shilan S. Saleem, Oluwasola Michael Akinola, Mohd Basyaruddin Abdul Rahman, Thean Chor Leow, Noor Dina Muhd Noor, Abu Bakar Salleh, Siti Nurbaya Oslan

Applied food biotechnology · 20.08.2025

DOI: 10.22037/afb.v12i1.48544 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Applied food biotechnology, erschienen am 20.08.2025. Verfasst von Shilan S. Saleem und 6 weiteren Autor:innen.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background and Objective: Identifying a milk-clotting enzyme (MCE) with high κ-casein specificity and heat sensitivity remains a challenge in cheese production. Current microbial, plant, and recombinant MCEs often exhibit low clotting activity, poor κ-casein specificity, and high thermostability, compromising cheese quality and increasing production costs. In this study, to address this, we developed a computational pipeline combining structural analysis, machine learning, and molecular dynamics simulation to design approximately 160,000 peptides from the Rhizomucor miehei protease–Pepstatin A complex (PDB ID: 2RMP). Material and Methods: Single-site mutagenesis, ML-driven affinity re-prediction, and physicochemical filtering yielded 84 peptides. Their specificity as aspartic proteases were validated via predicted Pepstatin A binding and further screened for cross-reactivity with αs1-, αs2-, and β-caseins. Results and Conclusion: Two candidates, Pep1 and Pep2, demonstrated superior κ-casein binding affinities (ΔG = -50.20 and -39.07 kcal/mol at 40 °C, respectively), lower melting indices (-4.05 and -3.09), and significantly enhanced specificity scores (-10.85) compared to Rhizomucor miehei protease (ΔG = -33.9 kcal/mol at 45 °C; melting index = 0.17; specificity score = 0.85). These peptides represent promising vegan- and halal-friendly alternatives to chymosins, pending experimental validation. Keywords: Milk-clotting enzymes, Computational peptide design, κ-casein specificity, Thermolabile peptides, Sustainable cheese production, Molecular dynamics simulations, Machine learning, Vegan cheese production, Halal cheese production, Aspartic protease. . Introduction Identifying chymosin substitutes with comparable enzymatic characteristics particularly high specificity for κ-casein (reflected in a high milk-clotting activity to proteolytic activity ratio, MCA/PA) and optimal heat sensitivity remains a significant challenge in cheesemaking [1-3]. Chymosin, traditionally derived from the stomachs of calves, has long been the gold standard for milk clotting due to its high specificity for κ-casein and its ability to produce high-quality cheese [4]. However, economic, ethical, dietary, and religious concerns, coupled with the rising demand for cheese production and consumption, have spurred the search for viable alternatives [1, 5]. Various sources of milk-clotting enzymes (MCEs) have been explored, including other animal-derived rennet (from lamb, buffalo, and camel), plant-based coagulants, and microbial and recombinant enzymes. Lamb and buffalo chymosin have lower MCA/PA ratios than calf chymosin, compromising cheese quality, though buffalo rennet’s stability suits mozzarella production [6, 7]. In contrast, camel rennet, while showing desirable clotting properties, is difficult to obtain in large quantities, limiting its widespread use [8]. Rennet from other animals, such as pigs and rabbits, often shows poor specificity and thermal stability, restricting their utility in cheesemaking [1]. Recently, marine organisms such as the shrimp Pleoticus muelleri have also been investigated as novel sources of coagulants [9-11]. Nevertheless, animal-based MCEs continue to face ethical and religious constraints. Plant-based coagulants, such as chymopapain from papaya, offer promising vegan- and halal-friendly alternatives [12]. However, their lower specificity toward κ-casein and tendency to generate bitterness limit their use, especially in aged cheeses [1, 13]. Microbial coagulants, particularly from bacterial and fungal sources, are increasingly favoured due to their cost-effectiveness and suitability for vegetarian and organic cheese production [14]. Although bacterial coagulants offer functional diversity [5, 15-18], they often exhibit high proteolytic activity, necessitating the development of improved strains through mutation [1]. Fungal coagulants are generally more compatible with cheesemaking conditions, particularly in terms of pH and temperature [1, 19]. Still, their high thermostability can cause undesired proteolysis during cheese ripening, adversely affecting texture and flavour [20]. Among fungal coagulants, mucorpepsin (EC 3.4.23.23) from Rhizomucor miehei is widely used due to its relative specificity. However, it still acts on α- and β-caseins and possesses high thermal stability, which can result in residual enzymatic activity after curd formation. This residual activity often leads to bitterness, altered texture, and reduced yield during cheese production [21-25]. Several studies have attempted to address these limitations using physical, chemical, enzymatic, and genetic modifications to enhance specificity, reduce thermostability, and minimise off-target proteolysis while maintaining milk-clotting efficiency. For example, some efforts have successfully reduced thermostability by removing N-linked carbohydrates, one of the primary factors contributing to R. miehei protease’s thermal resistance without significantly diminishing clotting activity [1, 25-29]. Other studies have enhanced milk coagulation without altering thermostability [30-35]. Genetic engineering, mutagenesis [36-39], substrate modification[40-43], and enzyme immobilisation [44, 45], have all been explored to develop efficient coagulants suitable for industrial cheese production. Nevertheless, regulatory restrictions on genetically modified organisms and technical and economic challenges have limited the commercial adoption of many of these approaches [46]. A novel strategy for identifying milk-clotting enzymes involves targeting conserved motifs—namely FDTSSD or FDTGSSE, found in all currently known commercial coagulants. Using this approach, several in silico BLAST searches have identified new candidate proteases for cheese manufacturing. Moreover, degenerate primers based on these motifs can identify relevant genes even when full gene sequences are unknown [26]. Computational enzyme design and molecular modelling have shown great potential in optimizing enzyme activities [47-49]. Advances in bioinformatics, molecular dynamics, and machine learning now enable the prediction and engineering of enzymes with enhanced specificity, catalytic efficiency, and thermostability. Structural insights into active sites support rational design through targeted mutations[50-52]. These in silico methods reduce experimental cost and time by predicting enzyme performance before validation. Notably, ancestral sequence reconstruction has been used to design a novel aspartic protease with improved κ-casein affinity and thermolability for cheesemaking [53]. Although milk-clotting enzymes (MCEs) have been widely studied, the application of advanced computational tools for their rational design remains underexplored [54]. Moreover, current MCEs often exhibit limitations such as low substrate specificity and excessive thermostability, which can negatively affect cheese quality and increase commercial costs. This study introduces a computational pipeline to design thermolabile, κ-casein-specific milk-clotting peptides as sustainable, vegan-, and halal-friendly alternatives for cheese production. An integrated computational pipeline was developed, incorporating machine learning, rational design, and structure-based screening. A combinatorial library of 160,000 peptide variants was generated based on conserved catalytic motifs from the Rhizomucor miehei protease–Pepstatin A complex, linked through a rationally designed sequence. The library was screened using predictive models trained on binding affinity data, followed by structure-based refinement, physicochemical profiling, and validation through molecular docking and molecular dynamics simulations. This in silico framework provides a foundation for the experimental development of next-generation coagulants for dairy applications. Materials and Methods 2.1. Aspartic peptide design strategy We present a computational pipeline for the de novo design of short peptide-based milk-clotting enzymes (MCEs) with enhanced affinity for κ-casein and improved chemosensitivity. The design strategy involves identifying interface residues containing catalytic motifs from Rhizomucor miehei protease (RMP), linking them via an optimised linker, and screening the resulting peptide candidates for specific binding and functional activity. To benchmark the design process, RMP was used as a positive control and also served as the structural template, while a peptide with low affinity for Pepstatin A was employed as a negative control. The results for both controls are provided in the supplementary file on GitHub. 2.2. Interface residue identification Interface residues from the crystal structure of 2RMP (RMP–Pepstatin A complex) were identified using the PDBePISA tool (https://www.ebi.ac.uk/pdbe/pisa/) [55] and served as the starting point for aspartic peptidase design. Two continuous interface segments containing the catalytic aspartic acid residues were selected: LFDTGSS (residues 36–42, harbouring the DTGS motif) and TIDTGTNFFI (residues 235–244, containing the DTGT motif). Although the ITYGT segment was identified among the interface residues, it was excluded from the design due to its lack of direct involvement in the conserved DTGS/DTGT catalytic motifs. Furthermore, incorporating non-catalytic segments would unnecessarily increase peptide length without functional contribution, potentially reducing the structural stability and target specificity of the designed peptides. 2.3. Continuous interface residues segment Catalytic motif of 2RMP (38-41, 237-240) LFDTGSS = 36-42 (7 residues) TIDTGTNFFI = 235-244 (10 residues) ITYGT = 80-84 2.4. Linker design and variant library construction The designed peptide sequence comprises two continuous interface segments, LFDTGSS (residues 36–42) and TIDTGTNFFI (residues 235–244), totalling 17 residues. The spatial distance between the terminal residues of these segments

Schlagwörter

milk-clotting enzymeaspartic proteasepepstatin Apeptide designcheese productionvegan cheesemolecular dynamics simulationbinding affinity

Zitieren

APA

Saleem, S. S., Akinola, O. M., Rahman, M. B. A., Leow, T. C., Noor, N. D. M., Salleh, A. B., & Oslan, S. N. (2025). Innovative Computational Design of Sustainable Milk-Clotting Peptides for Enhanced Cheese Production. Applied food biotechnology. https://doi.org/10.22037/afb.v12i1.48544

Vancouver

Saleem SS, Akinola OM, Rahman MBA, Leow TC, Noor NDM, Salleh AB, et al. Innovative Computational Design of Sustainable Milk-Clotting Peptides for Enhanced Cheese Production. Applied food biotechnology. 2025. doi:10.22037/afb.v12i1.48544

BibTeX
@article{Saleem2025,
  author = {Saleem, Shilan S. and Akinola, Oluwasola Michael and Rahman, Mohd Basyaruddin Abdul and Leow, Thean Chor and Noor, Noor Dina Muhd and Salleh, Abu Bakar and Oslan, Siti Nurbaya},
  title = {Innovative Computational Design of Sustainable Milk-Clotting Peptides for Enhanced Cheese Production},
  journal = {Applied food biotechnology},
  year = {2025},
  doi = {10.22037/afb.v12i1.48544}
}
Kostenlos starten →