Bioprocess Valorization of Brazilian Agro-Industrial Wastes for Enzyme Synthesis in Protease Production

Abstract

Proteases are key biocatalysts widely applied in the food, pharmaceutical, detergent, and environmental industries. One of the most costly steps in large-scale enzyme production is the preparation of the culture medium, making agro-industrial wastes attractive as low-cost nutrient sources and potential inducers. The non-conventional yeast Yarrowia lipolytica stands out in bioprocess engineering due to its high secretion capacity, GRAS status, and ability to metabolize diverse industrial residues. In this study, Brazilian agro-industrial by-products, namely Corn steep liquor (CSL), brewer’s yeast residue (BYR), and okara, were evaluated as alternative nitrogen sources for protease production by Y. lipolytica IMUFRJ 50678. Enzyme activity was quantified by the azocasein method at optimized conditions (40 °C, 40 min, pH 5 and 8). After an initial exploratory screening (n = 1), brewer’s yeast residue (BYR) and okara were identified as promising candidates for protease production. These preliminary findings guided subsequent experiments performed in biological triplicate (n = 3), which confirmed the reproducibility and comparative performance of these substrates, showing higher acid protease (AXP) activity in the BYR medium ((5.4 ± 0.3) U/mL), whereas alkaline protease (AEP) activities were comparable between the BYR ((8.4 ± 0.6) U/mL) and okara ((7.5 ± 0.9) U/mL) media. CSL was associated with higher lipase activity ((11.7 ± 0.9) × 10³ U/L), while esterase activity was higher in the BYR medium. These findings indicate that agro-industrial residues, particularly BYR and okara, can serve as effective nitrogen sources for protease production by Y. lipolytica IMUFRJ 50678, supporting their use in waste valorization and sustainable bioprocesses.

1. Introduction

Proteases (EC 3.4) are among the most industrially relevant hydrolases, catalyzing the cleavage of peptide bonds and converting proteins into peptides and amino acids. These enzymes are widely applied in the food, pharmaceutical, detergent, leather, and waste treatment industries [1,2,3,4,5,6,7,8]. The global protease market, valued at USD 2.03 billion in 2025, is expected to continue expanding, driven by growing demand for sustainable and eco-efficient biotechnological processes [9,10].

Microbial proteases dominate industrial production due to their high catalytic yields, robustness, and process scalability [11,12,13,14]. Among microbial hosts, the non-conventional yeast Yarrowia lipolytica has attracted considerable interest as a biotechnological chassis. This dimorphic, strictly aerobic yeast exhibits high secretion capacity, genetic stability, and a Generally Recognized as Safe (GRAS) status [15,16,17,18,19]. Its genome encodes several hydrolases, particularly two major extracellular proteases, an alkaline protease (AEP, XPR2) and an acid protease (AXP, AXP1), whose expression is strongly dependent on the medium composition and extracellular pH [19,20,21,22].

One of the main cost factors in enzyme production is the culture medium, in which commonly used nitrogen sources such as peptone and yeast extract are often applied. To overcome this limitation, agro-industrial by-products have emerged as renewable, low-cost nutrient alternatives that can serve as both carbon/nitrogen sources and as enzyme inducers [23,24,25,26,27]. This occurs because residues correspond to the fraction of biomass that is not converted during the production of high-value-added products [28]. Despite retaining a high content of valuable compounds, these materials are no longer relevant to the primary process and, due to their large generation volumes and the need for pretreatment before disposal, they impose additional costs on industry and may therefore be commercialized at prices significantly lower than those of the original biomass [29]. These residues also play a strategic role in circular bioeconomy frameworks by minimizing waste generation and enabling the recovery of value from industrial side streams.

Examples of studies that address the use of residues or coproducts for protease production by different microorganisms include the work of Novelli et al. [30], which explored different microorganisms of the genera Aspergillus and Penicillium for protease production using wheat bran or soybean bran as substrates. Singh and Bajaj [31] also investigated protease production by the bacterium Bacillus licheniformis K-3 using different residues as carbon sources, such as pine needle, wheat bran, potato peel, and cane bagasse, and as nitrogen sources, including gram husk, soybean meal, malt extract, mustard cake, and fish scales. Chimbekujwo et al. [32] used Aspergillus brasiliensis strain BCW2 to produce proteases from wheat bran. Liu et al. [33] obtained proteases from Bacillus licheniformis using soybean meal (SBM) and brewer’s spent grain as inducers. Gaonkar and Furtado [34] explored Haloferax lucentensis GUBF-2 MG076078 for the production of proteases and lipases using residues such as coconut oil cake, groundnut oil cake, rice bran, shrimp waste, fish processing waste, dairy waste, fish scales, crab shells, and feather waste.

In Brazil, the agro-industrial sector generates large volumes of residues, including corn steep liquor (CSL), brewer’s yeast residue (BYR), and okara (a soybean residue). Their utilization in microbial fermentation aligns with the United Nations Sustainable Development Goals (SDGs 9 and 12), which promote innovation and sustainable production systems.

To provide an overview of the magnitude of agro-industrial residues generated annually and, consequently, the need for their valorization in sustainable processes, corn steep liquor (CSL), a by-product of corn processing, represents approximately 40–50% of the processed corn mass [35]. Given that corn is a major commodity in countries such as Brazil, the United States, and China, global production is estimated at approximately 1.242 trillion tons per year in 2023 [36]. Brewer’s yeast residue (BYR) is obtained after 5–6 fermentation cycles in the brewing process, with an estimated generation of 15–18 tons of residual yeast per 10,000 hL of finished beer [37]. Okara is generated during soybean processing for tofu and soymilk production, with approximately 1.1–1.2 kg of okara produced per kilogram of processed soybeans [38,39]. Given the high consumption of these products in Asian countries, okara generation is estimated at approximately 800,000 tons per year in Japan, 310,000 tons in Korea, and 2,800,000 tons in China [39].

Therefore, the present study investigates the potential of the yeast Yarrowia lipolytica IMUFRJ 50678, still scarcely explored in the literature, as an industrially relevant microorganism for protease production under submerged fermentation. The study focuses on enzyme yield and substrate-induced effects using agro-industrial residues as alternative substrates, aiming to contribute to the development of sustainable bioprocesses. It is hypothesized that agro-industrial residues with higher protein content are effective inducers of protease production. To evaluate this hypothesis, different nitrogen sources were assessed in the culture medium, including peptone (used as a reference standard), egg albumin, and the agro-industrial residues corn steep liquor (CSL), brewer’s yeast residue (BYR), and okara.

2. Results and Discussion

2.1. Influence of Environmental Conditions on Protease Activity

The literature indicates that the yeast Yarrowia lipolytica produces two extracellular proteases: one active at acidic pH and another at alkaline pH. However, it was necessary to investigate whether the Y. lipolytica IMUFRJ 50678 enzymes behave under variations in proteolytic assay parameters.

In the present work, preliminary assays were performed to determine optimal pH and temperature for the proteolytic activity of Y. lipolytica IMUFRJ 50678. Azocasein, the chosen substrate, is poorly soluble at pH below 4 [40]; thus, assays were carried out at pH 5 and pH 8, representing acid and alkaline conditions, respectively, as previous studies on proteolytic activity at these pH values exist in the literature, such as Lario et al. (2015) [41] and Laishram and Pennathur (2016) [42]. Buffer molarity was adjusted to maintain pH stability during reaction: 0.1 M citrate buffer (pH 5) and 0.4 M Tris-HCl buffer (pH 8).

AXP is an aspartic protease whose active site is formed by the residues Asp77 and Asp264 [43]. In this system, one aspartate is protonated, while the other acts as a general base, activating a water molecule that then nucleophilically attacks the carbonyl carbon of the peptide bond. A tetrahedral intermediate is formed, whose decomposition leads to peptide bond cleavage and regeneration of the enzyme [44].

In turn, AEP is a serine protease with an active site composed of the residues Asp200, His231, and Ser397 [45], forming the catalytic triad Ser–His–Asp. The mechanism occurs in two main steps: during acylation, the serine is activated by histidine, attacks the substrate, forming a tetrahedral intermediate, and subsequently the acyl–enzyme complex; during deacylation, a water molecule activated by histidine hydrolyzes this intermediate, releasing the product and regenerating the enzyme [46].

Assays using the YPD medium supernatant were performed to evaluate the influence of incubation time and temperature (Figure 1 and Figure 2). Proteolytic activity reached a plateau after 40 min of incubation (Figure 1), with minimal standard deviation (2.5%), indicating steady-state enzyme kinetics. Temperature assays revealed maximum activity at 40 °C for both pH values (Figure 2), consistent with previous findings by Tobe et al. [47] for Y. lipolytica proteases. Consequently, all subsequent assays were conducted under these optimized conditions (40 °C, 40 min). For instance, the alkaline protease from Bacillus cereus BG1, described by Ghorbel, Sellami-Kamoun, and Nasri (2003) [48], exhibits an optimal temperature of 50 °C; however, in the presence of calcium ions, its activity doubles and the optimal temperature shifts to 60 °C, making it particularly attractive for industrial applications.

Comparable temperature profiles have been reported for alkaline proteases from Bacillus licheniformis [49] and Aspergillus parasiticus [50], and for acid proteases from Aspergillus clavatus [51], highlighting 40 °C as a favorable compromise between catalytic efficiency and enzyme stability.

Martinez et al. (2011) [52] performed the purification and stability characterization of alkaline protease (AEP) over a temperature range of 40–60 °C for 60 min, reporting optimal stability at the lowest temperature tested, with a half-life of 143 min, which increased to 247 min in the presence of 10 mM Ca²⁺. These results indicate that lower temperatures favor the stability of this enzyme; however, to maintain higher enzymatic activity, as shown in Figure 2, the temperature should be close to 40 °C.

2.2. Enzyme Production in Standard YPD Medium

The YPD medium was used as a control to evaluate the natural enzyme production profile of Y. lipolytica IMUFRJ 50678. During fermentation, glucose concentration gradually decreased until depletion at 72 h, while pH increased from 5.5 to 8.3 after 24 h (Figure 3).

Cellular biomass growth correlated with glucose consumption, reaching the stationary phase upon substrate exhaustion (Figure 4).

Cell viability, evaluated by direct counting, decreased slightly after 72 h, from 97.3% to 94.4%. However, due to the imprecision of the cell-counting technique, this decrease was not significant. Spectrophotometric and microscopic methods produced comparable trends, although direct counting was preferred for agro-industrial media containing particulate residues (e.g., okara, BYR) that could interfere with optical readings.

Hydrolase production during YPD fermentation is shown in Figure 5. AXP activity peaked at 96 h ((5.4 ± 0.1) U/mL) under acidic conditions (pH 5), whereas AEP activity reached a maximum at 72 h ((11.0 ± 0.1) U/mL) under alkaline conditions (pH 8). Lipase and esterase activities remained low, showing peaks of (5.4 ± 0.1) × ${10}^3$ U/L and (50 ± 3) × $10$³ U/L, respectively.

These results are consistent with the pH-dependent regulation of Y. lipolytica proteases, as described by González-López et al. [20] and McEwen & Young [53], who reported that AEP is expressed only above neutral pH. At the same time, AXP is suppressed under the same conditions. However, in this study, AXP remained detectable even under mildly alkaline pH, suggesting persistent secretion possibly due to prolonged fermentation or co-regulation mechanisms involving AEP.

Several studies in the literature have employed different strains of Y. lipolytica, in addition to the work of Buarque, Carniel, Ribeiro, and Coelho (2023) [54]. The latter used the strain Y. lipolytica IMUFRJ 50682 and observed cellular growth and pH variations similar to those shown in Figure 3 and Figure 4. However, they produced lipase and protease in a potassium phosphate-buffered medium (pH 7), achieving a peak lipase activity of 455.96 U/L at 32 h and a secondary proteolytic peak (23.7 U/L) at 40 h. Similarly, Carvalho, Finotelli, Bonomo, Franco, and Amaral (2017) [55] used the same strain in a 3 L aerated bioreactor at 650 rpm for 24 h of fermentation, yielding lipolytic activity of 58.3 U/mL and proteolytic activity of 0.152 U/mL. In another study, Brígida, Amaral, Gonçalves, Rocha-Leão, and Coelho (2014) [56] employed a biphasic reactor for lipase production with Y. lipolytica IMUFRJ 50682. They reported cellular growth and glucose consumption rates comparable to those in Figure 3 and Figure 4 for assays at 250 rpm. However, under biphasic conditions at 650 rpm, lipase production showed an anticipation of the two activity peaks (also observed in the control condition at 24 h and 48 h), with a maximum activity of 32,000 U/L at 20–24 h and a later peak of 100,000 U/L; this represented the best condition among the cited studies and was 18.4-fold higher than the peak lipolytic activity shown in Figure 5c. In contrast, protease activity was below 18 U/L in all assays.

Sales (2022) [57] evaluated the depolymerization of food packaging based on polyethylene terephthalate (PET) by Yarrowia lipolytica IMUFRJ 50682, using four culture media varying in nitrogen source (peptone or tryptone) and presence or absence of dextrose, formulating the YP, YPD, YT, and YTD media, with the addition of 500 mg/L of PET. Despite the presence of PET, cell growth in the YPD and YTD media showed a profile similar to that observed in Figure 4. Lipolytic activity peaked before 24 h in all media, with the best results observed with tryptone. YPD medium had the worst performance, with activity around 150 U/L and a smaller secondary peak at 24 h. Esterase activity also peaked in the YT medium. At the same time, YPD showed a profile similar to that observed in Figure 5d, with a reduction in activity at 48 h, followed by an increase to approximately 500 U/L at the end of fermentation. These values are 36 times lower for lipase and approximately 100 times lower for esterase than the results obtained in the YPD media assays under the evaluated conditions. However, it should be noted that the initial esterase activity values were already high, which contributed to the elevated levels observed. As for proteolytic activity (at pH 5), two trends were observed: media without dextrose showed a peak before 20 h, with values around 2750 U/L, while media with dextrose showed a gradual increase in activity. The YPD medium exceeded 3000 U/L for protease, corresponding to slightly more than half the value observed in Figure 5a, measured after 96 h of fermentation.

Botelho (2023) [58] also evaluated PET degradation by Y. lipolytica IMUFRJ 50682, using different conditions: YPD, YPD + DMSO, YPD + PET, and YPD + DMSO + PET. The YPD medium consisted of 1% yeast extract, 0.64% peptone, and 2% dextrose. The concentrations of DMSO and PET were 5% and 0.5%, respectively. The cell growth profile was similar across conditions, with the YPD medium reaching a biomass concentration of 13 g/L within 72 h. The pH variation was identical to that observed in Figure 3, starting from pH 6.5 and exceeding 7 throughout fermentation. Lipolytic activity in the YPD medium was extremely low, while the medium with DMSO and PET reached 380 U/L. Esterase activity in YPD reached 800 U/L at 24 h, dropping by half by the end of fermentation. The lipolytic activity values are approximately 0.8 times lower (at 24 h) and approximately 2% of esterase activity (at 96 h) than those observed in the YPD media for the IMUFRJ 50678 strain under the evaluated conditions.

Costa et al. (2020) [59] also investigated the consumption of PET-related compounds, cell growth, and enzyme production by Y. lipolytica using YP and YPD media with various PET derivatives, with stirring at 160 rpm for 96 h. Lipolytic activity peaked at 24 h in all conditions. In the YPD control medium, activity reached 40 U/L, then dropped to almost 0 U/L at 96 h. In the YP control medium, activity was 60 U/L. Proteolytic activity (pH 5, 37 °C) also peaked at 24 h, following a growth profile similar to that observed in Figure 5a. However, the maximum value at 96 h was approximately 700 U/L in YPD medium and 900 U/L in YP medium.

The studies discussed above use YPD culture medium and the yeast Y. lipolytica as the microorganism in the fermentation, with only the employed strain varying. Although they exhibit the pattern of at least one activity peak in lipolytic activity, the levels observed for the Y. lipolytica IMUFRJ 50678 strain were higher under the evaluated conditions compared to those of the other strains, and a similar trend was observed for proteolytic activity. This suggests that Y. lipolytica IMUFRJ 50678 can efficiently produce lipases, esterases, and proteases.

2.3. Effect of Different Inducers on Hydrolase Production

After establishing baseline activity in YPD, four alternative nitrogen sources were evaluated: corn steep liquor (YCD), brewer’s yeast residue (YBD), okara (YOD), and ovalbumin (YAD).

Fermentation profiles revealed distinct pH (Figure 6a) and growth patterns (Figure 6b). The YCD medium remained acidic (pH < 6) throughout fermentation, while YPD and YBD reached alkaline values above pH 8. Biomass accumulation was highest in YPD and YOD, and moderate in YBD.

Proteolytic activity results (n = 1) are summarized in Figure 7a,b. Under acidic assay conditions (pH 5), 96 h of fermentation, AXP activity followed the order of YBD ((5.61 ± 0.03) U/mL) > YPD ((5.4 ± 0.1) U/mL) > YOD ((4.5 ± 0.3) U/mL) > YAD ((4.3 ± 0.1) U/mL). Under alkaline conditions (pH 8), AEP activity ranked as follows: YPD (11.0 ± 0.1) U/mL at 72 h, decreasing later) > YOD ((8.7 ± 0.2) U/mL) > YBD ((8.2 ± 0.1) U/mL) > YAD ((6.9 ± 0.1) U/mL).

The substitution of peptone with brewer’s yeast residue or okara effectively sustained protease yields comparable to or exceeding those obtained with YPD medium. Notably, YBD induced the highest AXP production, whereas YOD favored AEP synthesis, but in the case of AEP production, the activity levels observed in the residue-based media were lower than those in the YPD medium.

Lipase and esterase activities (Figure 7c,d) showed higher values in the YAD medium, reaching (11.5 ± 0.3) $\times 10$³ U/L at 48 h, and in the YCD medium, reaching (11.7 ± 0.9)$\times 10$³ U/L at 96 h of fermentation. At the same time, YBD substantially enhanced esterase activity, reaching (74 ± 4) × 10⁴ U/L at 96 h. However, as these results were obtained from an initial screening performed with a single biological replicate (n = 1), they should be interpreted as preliminary observations rather than definitive comparative conclusions.

The observed differences among inducers may be attributed to their biochemical composition. BYR is rich in peptides and free amino acids, stimulating extracellular protease synthesis [60], while okara provides complex nitrogen and fiber, supporting balanced enzyme induction [61]. Conversely, CSL performed poorly, likely due to inhibitory compounds or the limited bioavailability of its nitrogen fraction. Another important factor that may influence the induction of the proteins of interest is variation in carbohydrate concentration and type, as well as in other micronutrients present in each residue, since the methodology used in this study standardized protein concentration and maintained constant levels of yeast extract and dextrose. Therefore, the influence of these components on protease production was not explored and could be addressed in future studies to better explain the observed differences in protease production among the evaluated residues.

These preliminary results indicated that BYR and okara exhibited favorable trends for protease production under the tested conditions. Based on this exploratory screening (n = 1), these substrates were selected for further evaluation using biological triplicates to assess reproducibility. The results for AXP production are presented in Figure 8a, and those for AEP production are shown in Figure 8b.

At the time point of highest protease production (96 h), the mean AXP activity remained higher in the YBD medium, with a value of (5.4 ± 0.3) U/mL, whereas the YOD medium showed an activity of (4.3 ± 0.2) U/mL. In contrast, for AEP production, mean activities of (8.4 ± 0.6) U/mL for YBD and (7.5 ± 0.9) U/mL for YOD were obtained, indicating that AEP production in these culture media is statistically comparable, with overlapping standard deviations. The mean lipolytic activity remained lower than that observed for the YMD medium shown in Figure 7c (n = 1). Esterase activity in the YBD medium exhibited greater variability among the biological triplicates at 72 h, and at 96 h, the mean activity was (71 ± 5) × 10⁴ U/L.

Proteolytic activities obtained with okara and brewer’s yeast residue media compare favorably with other reports of Y. lipolytica and related species. López-Flores et al. [62] observed protease activity of 67 U/mL in solid-state fermentation using soybean meal, while the present submerged fermentation achieved 8–11 U/mL, indicating lower volumetric activity but higher process scalability. Azeredo et al. (2006a, 2006b) [63,64] evaluated the effect of feather meal in the first study and supplementation with CSL in the second, using assays performed at 50 °C for 10 min at pH 6 with Streptomyces sp. 594 in SmF. These studies demonstrated that protease activity increased from (7.2 ± 0.2) U/mL to (13.4 U/mL), with two distinct activity peaks observed on the second and fourth days of cultivation. Based on Figure 1, after 10 min, the activity was 2.2-fold higher compared to 40 min. Therefore, when converting proteolytic activity measured in YOD and YBD media, the resulting AXP and AEP values are above the range reported in the literature.

Jamrath, Lindner, Popovic, & Bajpai (2012) [65] aimed at evaluating amylase and protease production by the bacterium Bacillus caldolyticus using food industry residues. Seven medium conditions were assessed: a control medium (hydrolysed starch and peptone, fermented for 8.8 h); a medium with pig blood (24 h); one with whey (6 h); one with CSL (6 h); and variations in the pig blood medium (10%, 5%, or 3%) + starch (10 h). Proteolytic assays were performed at pH 6.8 at 55 °C for 3 h, resulting in an average activity (0.5 ± 0.1) U/mL for the medium with CSL. The best medium was pig blood (3%) + starch, (0.9 ± 0.3) U/mL, although the control medium had the highest activity (1.2 ± 0.5) U/mL. They all exhibited proteolytic activity levels lower than those observed in the YOD and YBD media.

A comparative analysis of proteases produced by different species of the genus Yarrowia and some other yeasts regarding AEP was conducted by Ciurko, Neuvéglise, Szwechłowicz, Lazar, and Janek (2023) [66]. They then evaluated proteolytic activity against casein substrate in a cultivation medium containing brewer’s grain residue (BSG). The best result was obtained from Y. lipolytica W29 (2.47 U/mL/min), followed by Y. galli (1.96 U/mL/min), Y. alimentaria (1.38 U/mL/min), Y. keelungensis (1.07 U/mL/min), and Y. parophonii (0.80 U/mL/min). Despite Y. lipolytica’s performance under the evaluated conditions, proteolytic activity measured with casein was low, although higher values would be expected compared to the azocasein method. Moreover, the assay condition at 55 °C for only 10 min may have contributed to partial enzyme denaturation, potentially affecting the measured activity.

Mathias, Aguiar, Silva, Mello, & Sérvulo (2017) [67] conducted a study with a culture of Lactobacillus delbrueckii ssp., a microorganism also classified as GRAS, to evaluate three brewery solid residues: malt bagasse, hot trub, and brewer’s yeast residue as alternative media for lactic acid bacteria cultivation, focusing on the production of proteolytic enzymes. A proteolytic activity yield of 4.9 and 4.6 U/mL was achieved at pH 8.3 for 3 and 6 h, respectively. Although these activity levels have increased rapidly, they remain lower than the AEP activity levels observed in both YOD and YBD media under the evaluated conditions.

Kotlar, Belagardi, & Roura (2011) [68] also developed a study using brewery industry residues to produce hydrolases and other compounds by Bacillus cereus, utilizing another industry residue, malt bagasse. The study evaluated the proteolytic activity of the microorganisms, namely B. cereus, Pseudomonas sp., P. putida, E. hirae, and L. lactis subsp. lactis, cultivated at 60 rpm in an orbital shaker at 32 °C for 24 h, as well as the synergy of B. cereus with these other microorganisms. The proteolytic assay was performed at pH 7 for 30 min at 32 °C. The following results were obtained for microorganisms cultivated individually with 5% (v/v) inoculum: B. cereus (109.17 ± 7.50) U/mL, Pseudomonas sp. (85.00 ± 4.17) U/mL, P. putida (79.17 ± 1.67) U/mL, E. hirae (171.67 ± 10.00) U/mL, L. lactis subsp. lactis (100.83 ± 6.67) U/mL. The combination of B. cereus (2.5% inoculum) and other microorganisms (2.5% inoculum) showed the best results with B. cereus + Pseudomonas sp. at (213.67 ± 22.50) U/mL and B. cereus + P. putida (220.00 ± 9.17) U/mL. Therefore, the activity levels obtained tended to be higher than those observed for AXP and AEP under the evaluated conditions.

In a study by Slivinskia et al. (2012) [69], which focused on surfactin production by Bacillus pumilus UFPEDA 448, protease production was evaluated in both submerged and solid-state fermentations using soy peptone, non-hydrolysed okara, and hydrolysed okara as nitrogen sources. Protease production was significantly higher in solid-state fermentation across all media tested. Specifically, in the soy peptone medium, solid-state fermentation yielded 79 U/mL compared to 87 U/mL in submerged fermentation at 72 h; in the medium containing non-hydrolysed okara, protease activity was 38 U/mL in submerged fermentation and 442 U/mL (11.6-fold higher) in solid-state fermentation; and in the hydrolysed okara medium, activity increased from 82 U/mL in submerged fermentation to 356 U/mL in solid-state fermentation at 72 h. These results identify Bacillus pumilus UFPEDA 448 as an excellent protease producer, particularly under solid-state fermentation conditions. Comparing the submerged fermentation medium with non-hydrolysed okara to the YOD assays performed with Yarrowia lipolytica, the proteolytic activity levels were approximately 4.52 times higher than the mean activity obtained for the assays in the YBD medium under the evaluated conditions. This difference becomes even more pronounced when considering results obtained under other tested conditions or with solid-state fermentation using okara as a protease inducer. However, such comparisons should be interpreted with caution due to differences in experimental setups.

However, bacteria from the genus Bacillus continue to be better protease producers than the yeast Y. lipolytica. One example that supports this statement is the work by Bernardo, Kopplin, and Daroit (2023) [70], who used the strain Bacillus sp. CL18 to evaluate protease production using 12 different inducers. The proteolytic assay conducted in 100 mM Tris-HCl buffer at pH 8 and 55 °C for 15 min obtained the best results for ground fish scales (360 U/mL in 4 days), chicken feathers (332 U/mL in 5 days), and ground feathers (316 U/mL in 4 days); the medium with soybean meal, a residue from the soybean industry similar to okara, showed 200 U/mL on the third day. These results can be improved by a slight increase in the inducer concentration and by adding a second residue, such as milled feather at 5 g/L; ground fish scales increased the activity to 780 U/mL, a value substantially higher than the proteolytic activity levels in YOD and YBD media fermentations under the evaluated conditions.

To facilitate the comparison among the studies,

Table 1. Previous works about protease production.

Author	Fermentation Type	Microorganism	pH	Time (h)	Activity (U/mL)
This work	SmF	Y. lipolytica IMUFRJ 50678	5	96	5.4 ± 0.3
This work	SmF	Y. lipolytica IMUFRJ 50678	8	96	8.4 ± 0.6
López-Flores et al. [62]	SSF	Y. lipolytica	7	24	67
Azeredo et al. (2006a) [63]	SmF	Streptomyces sp. 594	6	72	7.2 ± 0.2
Azeredo et al. (2006b) [64]	SmF	Streptomyces sp. 594	6	72	13.4
Jamrath, Lindner, Popovic, & Bajpai (2012) [65]	SmF	Bacillus caldolyticus	6.8	10	1.2 ± 0.5
Ciurko, Neuvéglise, Szwechłowicz, Lazar, and Janek (2023) [66]	SmF	Y. lipolytica W29	7.5	72	2.47
Mathias, Aguiar, Silva, Mello, & Sérvulo (2017) [67]	SmF	Lactobacillus delbrueckii ssp.	8.3	3	4.9
Kotlar, Belagardi, & Roura (2011) [68]	SmF	B. cereus + Pseudomonas sp.	7	24	213.67 ± 22.50
Kotlar, Belagardi, & Roura (2011) [68]	SmF	B. cereus + P. putida	7	24	220.00 ± 9.17
Slivinskia et al. (2012) [69]	SSF	Bacillus pumilus UFPEDA 448	8.5	72	442
Bernardo, Kopplin, and Daroit (2023) [70]	SmF	Bacillus sp. CL18	8	72	780

was prepared.

The results confirm that Y. lipolytica IMUFRJ 50678 can efficiently utilize agro-industrial by-products as nitrogen sources, achieving relevant protease yields in submerged culture. The strain’s ability to metabolize diverse substrates and sustain proteolytic secretion highlights its potential for sustainable enzyme bioprocesses integrated into circular bioeconomy models, particularly in the valorization of residues from agro-industrial chains (e.g., brewing and soybean processing) and their integration into industrial fermentation platforms for enzyme production after further scale-up and process optimization.

3. Materials and Methods

3.1. Materials

Yeast extract, citric acid, disodium hydrogen phosphate heptahydrate, monosodium phosphate monohydrate, sodium citrate dihydrate, azocasein, trichloroacetic acid, 4-nitrophenyl dodecanoate (p-nitrophenyl laurate), 4-nitrophenyl butyrate, and 4-nitrophenol were purchased from Sigma-Aldrich^® (St. Louis, MO, USA). Bacteriological peptone was obtained from Oxoid^® (Basingstoke, UK). D(+)-Glucose (dextrose), acetic acid, and tris(hydroxymethyl)aminomethane were obtained from Vetec^® (Rio de Janeiro, Brazil). Sodium acetate, sodium hydroxide, and potassium hydroxide were purchased from Isofar^® (Rio de Janeiro, Brazil).

Hydrochloric acid (37%) supplied Loba Chemie Pvt Ltd.^® (Mumbai, India), and dimethyl sulfoxide (DMSO) by Tedia^® (Fairfield, OH, USA). Commercial enzymes—FAN Boost, Flavourzyme 1000L, Alcalase 2.4L, and Neutrase—were kindly provided by Novozymes^® (Bagsværd, Denmark).

Natural ovalbumin was obtained from Naturovos^® (Rio de Janeiro, Brazil).

All reagents were of analytical grade.

Agro-industrial residues were sourced as follows: CSL from Ingredion^® (Mogi Guaçu, Brazil), okara from Ecobras^® (Rio de Janeiro, Brazil), and BYR from Ambev^® (São Paulo, Brazil). Only okara was subjected to pretreatment through lyophilization.

3.2. Microorganism and Preservation

The wild-type yeast Yarrowia lipolytica IMUFRJ 50678 was isolated and identified by the Microbiology Institute of the Federal University of Rio de Janeiro (UFRJ), as previously reported by Mendonça-Hagler et al. [71]. Cells were preserved at −50 °C in YPD medium (1% yeast extract, 2% peptone, 2% glucose) supplemented with glycerol (50:50, v/v).

3.3. Inoculum Preparation

The pre-inoculum was prepared by cultivating Y. lipolytica IMUFRJ 50678 in 200 mL of YPD medium (1% yeast extract, 2% peptone, 2% glucose) in a 500 mL Erlenmeyer flask. The culture was incubated at 28 °C for 72 h at 160 rpm.

3.4. Fermentation Conditions

Fermentations were conducted in 500 mL Erlenmeyer flasks containing 200 mL of medium, without pH adjustment after sterilization. The cultures were inoculated with Y. lipolytica at an initial cell concentration of 10⁷ cells/mL and incubated at 28 °C and 250 rpm for up to 96 h. Five media were evaluated for their protein composition, with each formulated medium containing 2% (w/v) protein, as summarized in

Table 2. Culture medium composition.

Medium	Protein Proportion	Additional Components
YPD	Peptone (2%)	Yeast extract (1%), dextrose (2%)
YAD	Albumin (2%)	Yeast extract (1%), dextrose (2%)
YCD	Corn steep liquor (2%)	Yeast extract (1%), dextrose (2%)
YBD	Brewer’s yeast residue (2%)	Yeast extract (1%), dextrose (2%)
YOD	Okara (2%)	Yeast extract (1%), dextrose (2%)

.

After fermentation, cell-free supernatants were obtained by centrifugation at 8000 rpm for 6 min and used for extracellular enzyme activity assays.

3.5. Biomass Determination

Cell biomass was quantified using a Neubauer chamber (EMS Catalog Nos. 68052-14 and 68052-15) according to the manufacturer’s protocol [72]. The cell concentration (A, cells/mL) was calculated according to Equation (1):

$$A(\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}/\mathrm{m}\mathrm{L})=N/4\ast 1/V\ast f$$

(1)

where

A: Cell concentration (cells/mL);

N: Total number of cells counted;

f: dilution factor;

V: Neubauer chamber volume (mL).

Methylene blue was used to determine cell viability. The information on cell biomass growth presented in Section 3.3 was treated using a natural logarithmic (ln) transformation.

Cellular biomass growth was also monitored spectrophotometrically (SpectraMax M2E, Molecular Devices, San Jose, CA, USA) at 570 nm, and optical density values were converted to dry cell weight (g/L) using a previously established calibration curve for Y. lipolytica.

3.6. Glucose Quantification

Residual glucose was quantified by High-Performance Liquid Chromatography (HPLC, Shimadzu, Japan) using an Aminex^® HPX-87H column (300 mm × 7.8 mm, Bio-Rad Laboratories Ltd., Hercules, CA, USA) coupled with a cation-exchange guard column and a RID-10A refractive index detector. The mobile phase consisted of 5 mM H₂SO₄ flowing at 0.6 mL/min. The column temperature was maintained at 55 °C, and 20 μL of sample was injected after filtration through a 0.45 μm CHROMAFIL^® membrane. Glucose concentration was determined using a standard calibration curve.

3.7. Enzyme Assays

3.7.1. Protease Activity

Proteolytic activity was determined using the azocasein method [73], with modifications, according to the equation described by Carvalho, A.S.S. et al. [69]. Assays were performed at pH 5 (0.1 M citrate buffer) and pH 8 (0.4 M Tris-HCl buffer). The reaction mixture (280 μL) contained appropriately diluted enzyme extract and 0.5% azocasein. After incubation at 40 °C for 40 min, reactions were stopped with 140 μL of 10% trichloroacetic acid (TCA).

Samples were centrifuged (10,000 rpm, 10 min), and 100 μL of the supernatant was mixed with 100 μL of 1 N KOH. Absorbance was measured at 428 nm. Protease activity (U/mL) was calculated according to Equation (2):

$$A(\mathrm{U}/\mathrm{m}\mathrm{L})={\displaystyle \frac{Ab{s}_{sample}-Ab{s}_{blank}}{t\ast V_{sample}}}\ast D$$

(2)

where

A (U/mL): Enzyme activity, where 1 U is defined as the amount of enzyme that catalyzes the conversion of 1 micromole (µmol) of substrate per minute;

$Ab{s}_{sample}$: Absorbance of the sample;

$Ab{s}_{blank}$: Absorbance of the blank;

V: Sample volume in mL;

t: Reaction time in minutes;

D: Dilution factor.

3.7.2. Lipase Activity

Lipase activity was determined according to Carvalho A.S.S. et al. [74] and adapted from Buarque et al. (2023) [54], using p-nitrophenyl laurate (p-NPL) as the substrate. Reactions were monitored spectrophotometrically at 410 nm for 300 s at 37 °C. One unit (U) was defined as the amount of enzyme releasing 1 μmol of p-nitrophenol per minute under assay conditions. The lipolytic activity was calculated by Equation (3):

$$A\left(\mathrm{U}/\mathrm{L}\right)={\displaystyle \frac{∆abs}{∆t}}\ast {\displaystyle \frac{f\ast Vr\ast D}{Vs}}$$

(3)

where

A (U/L): the calculated enzyme activity, where 1 U is defined as the amount of enzyme that catalyzes the conversion of 1 micromole (µmol) of substrate per minute;

∆abs/∆t: Change in absorbance over the time interval;

Δt: Time (in minutes) during the linear phase of absorbance increase;

D: Dilution of the enzyme extract;

Vr: Reaction volume of the assay (in µL);

Vs: Volume of enzyme solution used in the assay (in µL);

$f$: Conversion factor obtained from the standard curve of p-nitrophenol ($f$ = 64.5 µmol/L).

3.7.3. Esterase Activity

Esterase activity was measured following Carniel et al. (2017) [75] using p-nitrophenyl butyrate (p-NPB) as the substrate and the equation described by Carvalho, A.S.S. et al. [38]. Absorbance was monitored at 410 nm for 300 s at 28 °C. One unit (U) corresponded to the release of 1 μmol of p-nitrophenol per minute under the defined assay conditions. Esterase activity was calculated by Equation (4):

$$A\left(\mathrm{U}/\mathrm{L}\right)={\displaystyle \frac{∆abs}{∆t}}\ast {\displaystyle \frac{f\ast Vr\ast D}{Vs}}.$$

(4)

where

A (U/L): the calculated enzyme activity, where 1 U is defined as the amount of enzyme that catalyzes the conversion of 1 micromole (µmol) of substrate per minute;

∆abs/∆t: Change in absorbance over the time interval;

Δt: (in minutes) during the linear phase of absorbance increase;

D: Dilution of the enzyme extract;

Vr: Reaction volume of the assay (in µL);

Vs: Volume of enzyme solution used in the assay (in µL);

$f$: Conversion factor obtained from the standard curve of p-nitrophenol ($f$ = 62.5 µmol/L).

3.8. Statistical Analysis

In this study, the initial screening for comparison among different culture media was conducted using a single biological replicate (n = 1). After identifying the most promising media for protease production, reproducibility was assessed using biological triplicates (n = 3) run simultaneously in independent flasks within the same fermentation batch. Protease assays were conducted in duplicate, while lipase and esterase assays were performed in triplicate.

Results were expressed as mean ± standard deviation and processed using Microsoft Excel for Microsoft 365 MSO (Version 2507, Microsoft Corporation, Redmond, WA, USA).

4. Conclusions

This study evaluated the valorization of Brazilian agro-industrial residues for protease production by Yarrowia lipolytica IMUFRJ 50678 under submerged fermentation. The results of the initial screening (n = 1) of the tested nitrogen sources indicated that brewer’s yeast residue (BYR) showed the highest acid protease (AXP) activity (5.61 ± 0.03 U/mL), followed by YPD medium (5.4 ± 0.1 U/mL). In contrast, alkaline protease (AEP) production was higher in the YPD medium, with an activity of 10.9 ± 0.4 U/mL, followed by the YOD medium (8.7 ± 0.2 U/mL), while the YBD medium ranked third (8.2 ± 0.1 U/mL). Corn steep liquor (CSL) was the residue most strongly associated with lipase production ((11.8 ± 0.9) × 10³ U/L), whereas BYR was also the most effective inducer of esterase production ((74 ± 4) × 10⁴ U/L). On the other hand, the YAD medium, for which a potentially higher induction of protease production was expected due to its commercial use as a human dietary supplement, did not confirm this expectation and was outperformed by the YBD and YOD media. Subsequently, reproducibility was assessed using biological triplicates (n = 3) for the most promising media. The YBD medium showed a mean AXP activity of (5.4 ± 0.3 U/mL), whereas the YOD medium exhibited a lower activity (4.3 ± 0.2 U/mL). For AEP production, mean activities of 8.4 ± 0.6 U/mL and 7.5 ± 0.9 U/mL were obtained for the YBD and YOD media, respectively. The mean esterase activity among the biological triplicates in the YBD medium was (71 ± 5) × 10⁴ U/L.

When comparing these results with those reported in the literature, bacterial systems such as Bacillus spp. typically achieve higher proteolytic activity levels. However, the yeast Y. lipolytica IMUFRJ 50678 demonstrated competitive performance considering its GRAS status, secretion capacity, and compatibility with renewable substrates, highlighting its potential as a sustainable microbial platform for enzyme bioprocesses and offering perspectives for the development of circular bioeconomy approaches in industrial biotechnology. In addition, it is important to note that the application of agro-industrial residues in enzyme production may face practical limitations, including variability in residue composition, the need for pretreatment steps, and uncertainties regarding supply stability, which should be carefully considered in future studies aimed at process optimization and scale-up.