All the fungal strains used in the present study were obtained from the laboratory of the Department Microbial Chemistry at the National Research Centre in Cairo, Egypt. The strains were maintained by sub-culturing them on slants of solid Czapek Dox medium containing: glucose, 30 g/l; NaNO, 2 g/l; KH₂ PO₄, 1 g/l; MgSO₄·7H₂O, 0.5 g/l; KCl, 0.5 g/l and agar, 20 g/l (BDH Chemicals Limited) at 28˚C and then stored at 4˚C. The inoculum size was adjusted by manually counting the spores.

All organisms were cultivated in liquid Czapek Dox medium (BDH Chemical, Ltd.), containing (g/l): glucose, 30; NaNO, 2; KH₂PO₄, 1; MgSO₄.7 H₂O, 0.5; KCl, 0.5 and following incubation at 28˚C for 5 days, the cultures were centrifuged at 5,000 x g, for 10 min at room temperature to obtain cell-free supernatants as crude enzyme extracts. These extracts were assayed for milk-clotting activity (MCA) by mixing 0.5 ml enzyme with 4.5 ml pre-warmed (35˚C) reconstituted skim milk containing CaCl₂, and the clotting time was recorded visually. The activity was expressed in arbitrary units relative to the time required for visible coagulation. To assess non-specific proteolysis, proteolytic activity (PA) was determined by incubating enzyme with azocasein, (Merck KGaA) (or casein), followed by trichloroacetic acid (TCA; El-Gomhouria Company), precipitation and measurement of soluble peptides spectrophotometrically. The ratio of MCA to PA was then calculated for each organism to evaluate enzyme suitability for cheesemaking. Organisms with high MCA and high MCA/PA ratio were considered the best candidates and selected for further purification and characterization, as shown Table I.

All chemicals and analytical-grade reagents used in the present study were procured from Merck KGaA. Skimmed milk powder (for MCA assays) was obtained from a local commercial supplier (El-Nasr Co. for Dairy Products). Wheat bran substrate was provided by ARMA Food Industries.

The effects of various culture media on MCE production were investigated using both submerged fermentation (SmF) and SSF under static and shaking conditions. A total of 10 liquid media were evaluated, including natural potato broth, Sabouraud dextrose broth, Czapek Dox broth, modified Czapek Dox with 5% sucrose and peptone, modified Czapek Dox with 5% sucrose and NaNO₃, starch nitrate broth, malt extract broth, yeast extract peptone dextrose (YPD) broth, synthetic potato dextrose agar (PDA) broth and a specific medium (Merck KGaA) (11). For each medium, 50 ml were inoculated with 1 ml spore suspension (7x10⁶ spores/ml) and incubated at 28˚C for 7 days under static and shaking (180 rpm) conditions. Additionally, eight agro-industrial residues, namely wheat bran, sugarcane bagasse, banana peel, guava seeds, corn stover, pomegranate peel, tangerine peel and sawdust (purchased from the local Arma Food Industry, Egyptian Company for solid waste recycling) were screened as solid substrates for MCE production in SSF. For SSF experiments, 5 g of each substrate were moistened with 5 ml distilled water, sterilized (121˚C for 15 min) and then inoculated with 1 ml of the same spore suspension. The mixtures were homogenized and incubated at 28˚C for 7 days. For enzyme extraction, the content of each flask was mixed with 100 ml of 0.05 M extraction buffer (Na-citrate, citric acid, pH 6.0) (1:10, w/v). The culture was mechanically agitated at 180 rpm in a shaking incubator (Benchmark Scientific Inc.) at 28˚C for 30 min. The mixture was then filtered to separate the mycelia from the medium. The supernatant (crude-enzyme extract) developed was utilized for MCE activity and determination of protein content, the method described by Nema et al (12). Through this optimization process, the ideal cultivation conditions for Penicillium purpurescens were determined to be as follows: A moisture content maintained between 60-70%, an aeration level of 75%, shaking at 180 rpm, a temperature range of 28-30˚C, an initial pH of 6.0-6.5 and an incubation period of 5 days.

Initial pH. The effect of incubation pH on MCE production was examined in the pH range of 3.0 to 10. The media pH was adjusted using either dilute HCl or dilute NaOH (El-Gomhouria Company) prior to autoclaving the bran (13).

Incubation temperature. Temperature was optimized for the growth and production of MCE by incubating the flasks containing the inoculated bran with pH 6, at 20, 25, 30, 35, 40 and 45˚C for 7 days (13).

Fermentation time. The MCE activity was monitored at 24-h intervals for up to 240 h. MCE production at optimum pH 6 and temperature 28˚C was followed. At the end of each time point, the extent of MCE production was determined as described by Vishwanatha et al (13).

Inoculum size. A total of 5 g of wheat bran was moistened with 5 ml distilled water (pH 6) and autoclaved. The substrate was inoculated with different volumes of the same spore suspension (0.25, 0.5, 1, 2, 3, 4 and 5 ml). The flasks were incubated at 25˚C for 5 days.

Different volumes of the substrate (milk + CaCl₂) were tested (0.125, 0.25, 0.5, 0.75, 1, 1.5, 2, 2.5 and 3 ml) against the same volume of the enzyme.

The MCA was determined using the method previously described by Kumar et al (14), based on the appearance of the first discontinuous particles and expressed in Soxhlet units (SU). The enzyme solution was incubated at 35˚C for 10 min before being added to skim milk (10% wt/vol, containing 10 mM CaCl₂ (Merck KGaA), which had been pre-incubated at the same temperature for 5 min. The reaction was terminated once discontinuous particles formed. The MCA was calculated using the following formula: SU=(2,400 x VS x N)/(T x VE), where VS represents the skim milk volume (ml), N is the MCE dilution factor, T is the milk-clotting time (seconds) and VE is the volume of the MCE (ml) used in the assay.

PA was determined using N,N-dimethyl casein (DMC) (Merck KGaA) as the substrate, following the method described in the study by Fan et al (15). In brief, 45 µl enzyme solution were combined with 45 µl substrate solution (10 mg/ml DMC in 20 mM potassium phosphate buffer, pH 5.8) and incubated at 35˚C for 30 min. The reaction was terminated by the addition of 350 µl of 100 mg/ml trichloroacetic acid (TCA) (El-Gomhouria Company), followed by incubation on ice for 20 min to precipitate undigested proteins. The samples were centrifuged at 2,000 x g for 15 min at 22˚C, and the absorbance of the supernatant was measured at 280 nm (UNICO advanced UV-VISIBLE spectrophotometer). A blank was prepared for each sample by adding TCA before the enzyme to correct for non-enzymatic hydrolysis. A total of one unit of PA [measured in proteolytic units (PU)] was defined as the amount of enzyme that releases 1 µg tyrosine per minute at 35˚C, using a tyrosine extinction coefficient (ε) of ml/µg for quantification:

The crude enzyme solution was fractioned by acetone and the active fraction with high MCA was further purified by passing through a column (1.5x40 cm) of DEAE-cellulose (Pharmacia Fine Chemicals AB) pre-equilibrated with 0.02 M sodium phosphate buffer at pH 6. The elution of protein was then carried out by the batch-wise addition of 40 ml portions of increasing molarities (0.0-0.4 M) of NaCl in 0.02 M phosphate buffer at pH 6. Fractions of 5 ml each were collected at room temperature (25˚C) at a flow rate of ~20 ml/h. The eluted fractions from the DEAE-cellulose column were dialyzed against cold distilled water to remove excess NaCl. The resulting purified enzyme solution was aliquoted into multiple test tubes, each containing 2 ml, and stored at -20˚C for later use. Under these storage conditions, the enzyme remained stable for >1 month. However, subjecting the enzyme to more than three freeze-thaw cycles resulted in approximately a 21% loss of activity. The enzyme remains active in the refrigerator for up to 1 week (16).

The molecular weight of the enzyme was determined using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) following the method described by Laemmli (17). A 12% acrylamide gel was prepared, and protein bands were visualized by staining with Coomassie Blue R-250 (Merck KGaA) at room temperature for 1 h, followed by de-staining until clear background was obtained. Zymography was performed in two steps: First, native-PAGE was conducted using a 12% polyacrylamide gel without SDS, as described by Snoek-van Beurden and Von den Hoff (18) with minor modifications. Samples were loaded without prior boiling.

To evaluate the effect of pH on the MCA of Penicillium purpurescens, the enzyme activity was assayed at different pH values ranging from 3.0 to 8.0. Substrate buffer solutions were prepared using: 0.05 M citrate buffer for pH 3.0-6.0, and 0.05 M phosphate buffer for pH 6.5-8.0. Each reaction mixture adjusted to the desired pH. The optimal pH was defined as the pH exhibiting the highest MCA. To assess the influence of temperature on MCA, enzyme assays were performed at temperatures ranging from 20 to 80˚C.

The Penicillium purpurescens MCE was dissolved in solutions containing 10 or 50 µM Na⁺, Ca²⁺, Mg²⁺, Hg²⁺ and Cu²⁺, respectively, and incubated for 30 min at 40˚C. The changes in MCA were measured, with the MCA of the untreated control (no metal ions) set as 100%.

The kinetic parameters of fungal MCE were determined using a modified version of the method described by He et al (19). Casein solutions at varying concentrations (0.1 and 3.0 g/l) were used as substrates to measure the PA following the protocol described in the study by Kumar et al (14). The Michaelis-Menten constant (Km) and maximal velocity (Vmax) were derived from a Lineweaver-Burk double-reciprocal plot, as established by Lineweaver and Burk (20). This analysis provided insight into the catalytic efficiency and substrate affinity of the enzyme.

All experiments were performed in triplicate. All reported values are expressed as the mean ± standard deviation (SD). Statistical analysis was conducted using SPSS 17.0 (IBM Corp.). Comparisons between two groups were performed using a t-test. P<0.05 was considered to indicate a statistically significant difference.

The MCA of various fungal and bacterial strains, measured in MCA/ml is presented in Table I. Among the tested strains, only a few exhibited detectable MCA, while the majority did not exhibit any activity. Penicillium purpurescens demonstrated the highest MCA (190±3.4 MCA/ml), suggesting its potential as a robust MCA. Aspergillus oryzae under shaking conditions in PDA medium (Materials and methods) also exhibited notable activity (102±3.4 MCA/ml), indicating that cultivation conditions significantly influence enzyme production (9,21). Curvularia lunata DHE5 and Noh2 displayed moderate activity (60.8±2.1 and 45±1.2 MCA/ml, respectively), which may warrant further optimization for industrial applications. The absence of MCA in the majority of strains, including Scopulariopsis brevicaulis, Penicillium javanicum and various Aspergillus and Penicillium species, suggests that MCEs are strain-specific and not universally produced (22). The variability in MCA between different growth conditions (static vs. shaking, dox vs. PDA) highlights the importance of optimizing fermentation parameters for maximal enzyme yield (23).

The results obtained (Table II) revealed variations in MCE production by Penicillium purpurescens across different media, aligning with findings from other fungal studies in Table I. For example, Aspergillus oryzae exhibited high MCE activity in wheat bran-based SSF (24), comparable to the optimal performance observed in Penicillium purpurescens under similar conditions. Likewise, Thermomucor indicae-seudaticae N31 was investigated using submerged fermentation with 4% wheat bran in a 0.3% saline solution, incubated at 45˚C and 150 rpm for 72 h (25), Furthermore, in solid-state fermentation of Mucor racemosus, molasses and casein were identified as significant carbon and nitrogen sources (26). Further emphasized is the role of substrate composition in fungal enzyme production, demonstrating that lignocellulosic substrates significantly influence MCE yields in SSF. The lack of clotting activity in Sabouraud and natural potato media for Penicillium purpurescens is consistent with reports on Neurospora intermedia, which also displayed low enzyme production in similar nutrient-limited substrates (27). These comparisons underscore the critical influence of media composition on fungal MCE production, with SSF often proving superior due to better fungal growth and enzyme induction, as demonstrated by Penicillium purpurescens and other fungi such as Aspergillus oryzae.

As illustrated in Fig. 1, the pH of the culture medium significantly influenced both MCE production and PA in Penicillium purpurescens, with optimal activity observed at near-neutral pH (6.0-6.5). No activity was detected at highly acidic (pH 2-3) or alkaline (pH 7) conditions, while a progressive increase was noted from pH 4 to 6, peaking at pH 6 (188 MCE units, 120 PU). This trend is in accordance with findings in other fungi, such as Aspergillus oryzae, which exhibited maximal MCE activity at pH 5.5-6.0(24), and Mucor racemosus, which exhibited optimal production at pH 4.8 (3,26). Similarly, Neurospora intermedia demonstrated reduced enzyme yields at extreme pH levels, emphasizing the importance of near-neutral conditions for fungal protease stability and secretion (27). The sharp decline in Penicillium purpurescens activity at pH 7 suggests enzyme denaturation or impaired fungal metabolism under alkaline conditions, a phenomenon also observed in Bacillus spp (28), which are known to produce enzymes that remain stable across a broad pH range (6.0-10.0).

The initial temperature significantly influences the production of MCE and PA in Penicillium purpurescens. As demonstrated in Fig. 2, the MCE and PA increased with temperature, peaking at 50˚C (145 U/ml MCE, 120 U/ml PU), followed by a sharp decline beyond 60˚C due to enzyme denaturation. Similar trends have been observed in other fungi, such as Aspergillus oryzae and Rhizopus miehei, which exhibit optimal MCE production at 30-40˚C and 45-55˚C, respectively (24,29). Aspergillus niger has also been shown to exhibit a reduced PA >50˚C, suggesting that thermal stability is common among fungal enzymes. These findings highlight the importance of optimizing temperature for maximal enzyme yield and stability in microbial MCE production.

The effect of fermentation time on Penicillium purpurescens MCE production and proteolytic activity follows a trend where enzyme activity initially increases with time before declining due to proteolytic degradation or nutrient depletion. As demonstrated in Fig. 3, optimal MCE production often occurs between 4-5 days, coinciding with peak fungal biomass and secondary metabolite secretion (30). Beyond this period, prolonged fermentation leads to reduced enzyme stability and increased proteolytic degradation, as observed in Penicillium purpurescens, where MCE activity decreased after day 4. Similar patterns have been observed in other fungi, such as Aspergillus oryzae and Rhizomucor miehei, where maximal MCE yield was achieved within 3-5 days before the activity diminishes (24,31). In the present study, the MCE demonstrated a ratio of 320, which is higher than the reported ratios for commercial counterparts and moderately close to that of calf rennet (~433), suggesting promising industrial potential, As shown in Table III, Aspergillus niger exhibits peak protease secretion at 72-96 h, after which activity declines due to autolysis. In contrast, the enzyme from Thermomucor indicae-seudaticae N31 was successfully used in the production of high-quality Prato cheese, maintaining stability without changes over 60 days (32). These findings highlight the importance of optimizing fermentation duration to balance enzyme yield and stability, which may vary across species.

As illustrated in Fig. 4, the inoculum size significantly influenced the production of MCE and PA in Penicillium purpurescens. Research has shown that increasing the inoculum size from 0.25 to 3 ml enhances both enzyme activity (from 370 to 631 SU) and proteolytic activity (from 120 to 285 U/ml), likely due to improved fungal biomass and nutrient utilization (33). However, beyond 3 ml, a decline in MCE activity (533 SU at 4 ml) and PA (290 U/ml), occurs, possibly due to nutrient depletion or metabolic stress. Similar trends have been observed in other fungi, such as Aspergillus oryzae, and Aspergillus clavatus, where optimal inoculum size (1-2x10⁶ spores/ml) maximizes protease production, while higher densities reduce yields due to oxygen limitation (24,33). Likewise, Rhizopus microsporus exhibits peak MCE activity at 5% (v/v) inoculum, with declines at higher concentrations (34,35). These findings suggest that inoculum optimization is critical for balancing microbial growth and enzyme synthesis across fungal species.

As demonstrated in Fig. 5, MCA peaked at 1% CaCl₂ (1,140 U/ml), while PA reached its maximum level (685 U/ml) at the same concentration. Higher CaCl₂ levels (>1%) lead to a decline in both MCA and PA, suggesting enzyme inhibition or destabilization at an elevated ionic strength. Similar trends have been observed in other fungi, such as Bacillus velezensis DB219, where optimal MCE production occurred at 0.5-1% CaCl₂ beyond which activity decreased due to possible metal ion interference with enzyme stability (36). Rhizopus miehei also exhibited reduced proteolytic activity at CaCl₂ concentrations above 1.5%, attributed to conformational changes in the enzyme (37). These findings highlight the critical role of Ca²⁺ in modulating fungal MCE activity, with species-specific optimal thresholds. The decline in Penicillium purpurescens MCE beyond 1% CaCl₂ aligns with previous research on Neurospora crassa, where excessive Ca²⁺ disrupted secretion pathways (38) underscoring the delicate balance between cation stimulation and inhibition in fungal enzyme systems.

As illustrated in Fig. 6, the purification of the MCE from Penicillium purpurescens involves fractionation steps that enhance specific activity, as evidenced by the data exhibiting increased activity in fractions 7-9, with peak specific activities of 286. 296 and 320 U/mg U/mg, respectively. These fractions likely contain the purified enzyme, as indicated by their high specific activities compared to earlier fractions with negligible activity. Similar purification trends have been observed in other fungal species, such as Aspergillus oryzae and Rhizomucor miehei, also produce proteases with milk-clotting properties (24,37). For instance, Rhizomucor miehei yields a highly active aspartic protease (rennin-like enzyme) after purification, with specific activities comparable to those of P. purpurescens (37). However, some fractions (e.g., 10-12) exhibit variability, possibly due to contamination or partial denaturation. Further purification steps, such as ion-exchange or gel-filtration chromatography, could improve homogeneity, as demonstrated in Aspergillus niger protease purification (23). These data suggest that Penicillium purpurescens is a promising source of milk-clotting enzymes, with purification efficiency comparable to other industrially relevant fungi.

The MCE from Penicillium purpurescens has a molecular weight of ~29 kDa (Fig. 7), which is comparable to other fungal aspartic proteases used in cheese production. For instance, Rhizomucor miehei produces a 38.6-kDa aspartic protease (rennet) (37), while Aspergillus oryzae secretes a 43-kDa MCE (39). Similarly, Endothia parasitica yields a 33.8 kDa protease (40). The lower molecular weight of Penicillium purpurescens enzyme suggests structural differences that may influence substrate specificity and thermal stability. These fungal enzymes are preferred over animal rennet due to their cost-effectiveness and suitability for vegetarian cheese production. Further comparative studies on their biochemical properties could optimize their industrial applications.

The MCA of the enzyme produced by Penicillium purpurescens exhibited a distinct dependence on pH, with optimal activity observed at pH 6, as shown in Fig. 1, where the enzyme reached a maximum activity of 336 MCE/ml. This result suggests that the enzyme functions best under mildly acidic to neutral conditions, aligning with observations from other microbial MCEs. For instance, similar pH optima have been reported for enzymes derived from Mucor pusillus and Rhizomucor miehei, which also exhibit peak MCA around pH 5.5-6.5 (37,41). The activity was significantly lower at both extreme acidic (pH 3, 120 MCE/ml) and alkaline (pH 10, 200 MCE/ml) conditions, indicating a narrow optimal pH range for maximum proteolytic efficiency. Fungal milk-clotting enzymes, as all enzymes, are sensitive to pH; deviations from their optimal pH range can lead to decreased stability and a reduced ability to bind to their substrate, milk proteins. This phenomenon underscores the critical need to control pH during dairy processing to ensure optimal enzyme performance, which directly impacts the efficiency and quality of products such as cheese (42).

The MCE derived from Penicillium purpurescens exhibited a pronounced temperature-dependent activity profile. Enzyme activity increased steadily from 30˚C, reaching its maximum at 50˚C with an activity of 360 U/ml, which indicates this temperature as optimal for catalytic function, as shown in Fig. 2. A sharp decline in activity was observed at 55˚C, suggesting the onset of thermal denaturation; however, the enzyme retained partial activity at 60˚C. Beyond this temperature, activity decreased markedly, with complete inactivation occurring at 90˚C. This thermal behavior is consistent with MCEs from other fungal species, such as Rhizomucor miehei and Mucor pusillus, which generally show optimal activity within the 45-55˚C range and lose enzymatic function at elevated temperatures due to structural instability (37,41). These observations underscore the critical need for precise temperature control when employing fungal MCEs such as MCE from Penicillium purpurescens in industrial processes such as cheese manufacturing.

As demonstrated in Table IV, a marked influence of various metal ions was observed on the MCE activity of Penicillium purpurescens. The data revealed that Ca²⁺ (50 µM) enhanced enzyme activity by 150%, suggesting a potential stabilizing or activating role, while Mg²⁺ (50 µM) also promoted activity (120%), albeit to a lesser extent. By contrast, Cu²⁺ (10 µM) and Hg²⁺ (10 µM) markedly inhibited MCE function, reducing activity to 40 and 10%, respectively, likely due to metal-induced protein denaturation or active-site interference (43). The negligible effect of Na⁺ (10 µM; 95% activity) implies that monovalent ions may have minimal impact on this enzyme system. These findings align with recent studies highlighting the dual role of metal ions as either cofactors or inhibitors in fungal MCE kinetics (44).

The Michaelis constant (K_m) of the MCE from Penicillium purpurescens was determined to be 5x10^-2 M (Fig. 8), indicating its moderate affinity for the substrate compared to other fungal proteases. For instance, Aspergillus oryzae produces a milk-clotting enzyme with a lower Km (2.5x10^-2 M), suggesting higher substrate affinity, while Rhizomucor miehei exhibits a markedly lower K_m (3x10^-3 M), reflecting superior catalytic efficiency in cheese-making (37,44). By contrast, Endothia parasitica has a higher Km (12x10^-1 M), indicating weaker substrate binding (40). The K_m value of Penicillium purpurescens positions it as an intermediate among fungal coagulants, potentially offering balanced proteolytic activity for dairy applications.

In the authors' laboratory, the solid-state fermentation process described herein demonstrated good reproducibility, as repeated batches produced under identical conditions consistently yielded comparable enzyme activity and quantity. This reproducibility at the laboratory scale supports the robustness of the optimized process. However, the scalability of the method has not yet been assessed due to some limitations we faced, including the lack of a pilot at the current time, a lack of funds, and the absence of an experienced team to deal with a pilot and its maintenance. Future research is required to focus on pilot-scale trials to evaluate the feasibility of large-scale enzyme production. Scaling up SSF may present challenges, including maintaining uniform moisture distribution, temperature control and adequate aeration in larger bioreactors, which need to be addressed to ensure consistent enzyme quality and yield in industrial applications (45).

Based on the results presented herein, the MCE from Penicillium purpurescens produces high-quality cheese curds with optimal characteristics, including curd yield, firmness and sensory properties (Table V). Further comparative studies are required to optimize its use in industrial processes. Collectively, Penicillium purpurescens demonstrates strong potential as a microbial rennet source, with performance comparable to established fungal producers. Further optimization and industrial-scale validation could enhance its applicability in cheese manufacturing.

The present study had certain limitations which should be mentioned. The activity of the enzyme against α- and β-casein fractions could not be assessed due to logistical challenges beyond the authors' control, including delays in chemical imports.

In conclusion, the present study highlights Penicillium purpurescens as a highly promising fungal strain for MCE production, demonstrating superior clotting efficiency compared to other screened strains. Optimal MCE production was achieved under SSF with wheat bran, near-neutral pH (6.0-6.5), and moderate temperatures (50˚C), with fermentation kinetics peaking at 4-5 days. The enzyme exhibited enhanced activity at 1% CaCs₂, but declined at higher concentrations, suggesting ionic strength sensitivity. Purification yielded fractions with high specific activity (up to 320 U/mg), and the molecular weight of the enzyme (29 kDa) and kinetic properties (K_m=5x10^-2 M) position it as a competitive alternative to commercial fungal rennets such as those and Aspergillus oryzae. These findings underscore Penicillium purpurescens as a viable candidate for industrial cheese production, though further scale-up and stability studies are needed to confirm its commercial applicability. Future research is warranted.