Taste characterization of okara and okara-derived substrates fermented by Pleurotus ostreatus mycelium

Abstract

Given its degradative capacity, mycelium may valorize okara, an underutilized agro-industrial by-product. This study aimed to conduct solid-state fermentation of okara and mixtures (okara with sawdust or wheat bran) using Pleurotus ostreatus mycelium and explore the potential of fermented okara as an alternative food, focusing on its taste characteristics. The fastest mycelial growth with dense cultures was achieved on non-dried okara, followed by wheat bran (11–12 days). Fermentation enhanced the nitrogen content in okara, indicating an improved protein level. Following biotransformation, substrates obtained higher concentrations of glucose (2.78- to 6.39-fold rise), free amino acids (2.16- to 6.58-fold rise), and 5′-nucleotides (3.24- to 6.31-fold rise), with succinic acid being the predominant organic acid. Electronic tongue measurement revealed reduced sourness and enhanced umami taste and richness of okara. This work demonstrated the potential of P. ostreatus for the bioconversion of okara and mixed media to acquire products with a palatable taste.

Highlights

• •Valorisation of okara and mixtures by fermentation via P. ostreatus mycelium.
• •Medium differences and fermentation significantly affected taste properties.
• •Fermentation increased free amino acids, 5′-nucleotides, and succinic acid levels.
• •The electronic tongue test showed umami taste improvements in fermented okara.
• •Fermented okara and okara-mix may serve as palatable protein alternatives.

1. Introduction

Annually, 1.4 billion tons of soybean residue (okara), a byproduct from the tofu and soymilk industry, are generated in the world. The utilization of okara is constrained by its perishable nature, high insolubility caused by high fiber content, rough taste, and poor flavor (Kamble & Rani, 2020; Liang et al., 2023). However, fermentation can be a promising path for okara biovalorization, as okara, which has a rich nutritional profile, is a potential growing medium for microorganisms. For instance, when biotransformation of okara was performed using L. plantarum and R. oligosporus, microbial hydrolases were released to break down the complex macromolecules into simpler forms of amino acids, sugars, and short-chain fatty acids, leading to improved digestibility (Gupta & Chen, 2021).

Fungal fermentation can lower the levels of crude fiber while enhancing the levels of soluble fiber, resulting in modified processing parameters and greater nutritional and sensory value to food residues. The contents of off-flavor volatile substances in okara were reduced significantly by the fermentation through edible fungi named W. cocos and T. fuciformis (Z. Wang et al., 2022). In addition, fermentation by two types of yeast (Hansenula sp. and S. cerevisiae) ameliorated the antioxidant capacity, physicochemical parameters, proximate composition, and sensory characteristics of okara (Shi, Zhang, Wang, & Devahastin, 2020). HongJun Tofu prepared using multiple strains (N. crassa, L. plantarum, and M. guilliermondii) gained decreased bitterness and sourness and a richer taste and freshness (Qiu et al., 2023). Even though there are considerable number of studies assessing the nutritional and antioxidant attributes of fermented okara, its taste characterization through non-volatile taste component analysis and electronic tongue test is an underexplored area.

To date, only a few research studies have been performed to facilitate the solid-state fermentation (SSF) of okara using mushroom mycelium, particularly with species such as W. cocos, T. fuciformis, G. lucidum, and L. edodes (Wang et al., 2022; Yang, Fu, & Yang, 2020). Pleurotus ostreatus (oyster mushroom) is one of the widely cultivated fungal species, and its mycelium is recognized for its substantial ability to break down and decay various organic compounds, including food waste (Bonatti, Karnopp, Soares, & Furlan, 2004). To the best of the author's knowledge, as of now, there is only one available study attributable to SSF of okara using P. ostreatus to obtain fermented okara with increased nutritional or sensory attributes. Samarasiri and Chen (2025a) reported changes in composition, taste characteristics, and antioxidant capacity of fresh okara following SSF with Pleurotus spp., highlighting the importance of strain and species selection. However, based on the taste characterization of mushrooms, substrate composition and the drying technique are two other parameters that can influence taste characteristics (Samarasiri & Chen, 2022).

Therefore, this study aims to examine the capability of okara (both fresh okara and dried okara) as a substrate for P. ostreatus mycelium in comparison to sawdust and wheat bran, which are two widely used substrate ingredients in commercial mushroom farming. It is hypothesized that SSF with P. ostreatus enhances the value of okara-based substrates by altering their taste component profile through fungal metabolism and degradation of complex macromolecules, resulting in the release and transformation of sugars, organic acids, amino acids, and nucleotides. Since okara can be effectively preserved through drying, two drying techniques (freeze-drying and thermal drying) were considered for okara pre-treatment in the present work. More importantly, this study evaluates mycelial growth performance, elemental composition, and the taste characteristics of fermented okara and fermented mixed substrates, highlighting their potential as alternative foods.

2. Materials and methods

2.1. Materials

Sawdust and fresh okara were kindly provided by Xcel Industrial Supplies Pte Ltd., Singapore, and Vitasoy International Singapore Pte Ltd., Singapore, respectively. Fresh okara was maintained at −20 °C. Food-grade Bob's Red Mill wheat bran was obtained from the local market. Calcium carbonate, methanol, disodium ethylenediaminetetraacetate dihydrate (EDTA), hexane, acetonitrile, standards (sugar, organic acid, amino acid (AAS18), and 5′-nucleotide), and potassium phosphate were supplied by Sigma-Aldrich (St. Louis, MO, USA). Standards, solutions, and sensors for the electronic tongue system were obtained from Intelligent Sensor Technology, Inc. (Kanagawa, Japan). All the chemicals, standards, and solvents used in HPLC analyses were HPLC grade. LC-MS grade acetonitrile, ammonium formate, formic acid, and methanol were supplied by Thermo Fisher Scientific (Waltham, MA, USA).

2.2. Mycelial culture preparation

The mycelium strain was supplied by Royal Flush Mushrooms (British Columbia, Canada) and was grown on sterilized potato dextrose agar (PDA) following incubation at 70% humidity and at 25 °C for 14 days. After a single subculturing step, the cultures were maintained at 4 °C.

2.3. Substrate preparation

Sawdust was sieved (mesh size ∼2 mm). Okara was thawed at room temperature for about 4–6 h before use or drying. Basal media of the mixed substrates were prepared using either sawdust or wheat bran with different okara compositions (0%, 25%, 50%, 75%, and 100%) as shown in Table 1 to systematically evaluate the concentration-dependent effects of okara incorporation on fungal growth and taste-related properties. For those media, okara was dried in a hot air oven at 60 °C for 22 h. Then, hot air-dried okara was ground and sieved (mesh size ∼2 mm). For example, the basal medium with 0% okara and 100% wheat bran was labelled as OW0, while the basal medium with 75% okara and 25% sawdust was labelled as OS75. pH of the substrates ranged between 5 and 6.6 after mixing 1% CaCO₃ (Ma et al., 2020), and 50% deionized water was added to each substrate.

Table 1.

The composition and details of the substrate formulations used in the fermentation.

Medium	Okara (%)	Wheat bran (%)	Sawdust (%)	C/N	Details of ingredients
OW0	0	100	0	12.99	Wheat bran (MC: 10.82 ± 0.03%)
OW25	25	75	0	11.47	Wheat bran (MC: 10.82 ± 0.03%) Hot air-dried okara (MC:7.49 ± 0.01%)
OW50	50	50	0	10.33	Same as above
OW75	75	25	0	9.44	Same as above
OW100	100	0	0	8.73	Hot air-dried okara (MC: 7.49 ± 0.01%)
DE	100	0	0	8.87	Dehydrated okara (MC: 13.26 ± 0.29%)
FD	100	0	0	8.19	Freeze-dried okara (MC:14.48 ± 0.01%)
FO	100	0	0	8.47	Undried/raw okara (MC: 78.59 ± 0.02%)
OS75	75	0	25	10.39	Sawdust (MC: 11.27 ± 0.08%) Hot air-dried okara (MC: 7.49 ± 0.01%)
OS50	50	0	50	12.92	Same as above
OS25	25	0	75	17.19	Same as above
OS0	0	0	100	25.99	Same as above

OW - Okara and wheat bran mixed media; DE - Dehydrated okara; FD - Freeze-dried okara; FO - Fresh okara; OS - Okara and sawdust mixed media; MC - Moisture content. Five replicates from each substrate.

Additionally, there were three other media containing okara that were pre-treated in different ways. They were fresh okara, freeze-dried okara, and dehydrated okara. Dehydrated okara (dried at 60 °C for 12 h using a domestic dehydrator) and freeze-dried okara (freeze-dried for 48 h) were ground and sieved as mentioned earlier. 50% deionized water was mixed with dried okara. Fresh okara was only subjected to thawing. The pH of the okara media was adjusted similarly. All the substrates were autoclaved at 121 °C and 15 psi for 20 min.

2.4. Solid-state fermentation

16 g of each substrate was placed in each petri dish. Inoculation was performed using one piece of mycelial culture (∼ 25 mm × 15 mm) per petri dish. All the plates were incubated at 70–80% humidity and 25 °C in an enclosed dark environment. Fermentation was terminated once the mycelium fully colonised the entire substrate surface. The radial growth (extension) of mycelium and the weight reductions of each petri dish were measured daily. The average extension from two perpendicular directions was considered. The mycelial growth rate (mm/day) was determined from the slope of the linear regression of the growth curve (Guadarrama-Mendoza et al., 2014). Following mycelial growth completion, moisture levels were determined at 110 °C using an MB120 moisture analyser (Ohaus Corporation, USA). All samples were stored at 4 °C before lyophilization and grinding for subsequent analysis.

2.5. Ultimate analysis

Carbon (C), nitrogen (N), and hydrogen (H) contents of all fermented samples and raw materials were determined using a CHNS elemental analyser (EA3100, Eurovector, Pavia, Italy) with 0.5–1.5 mg of dried sample (Samarasiri & Chen, 2025a).

2.6. Taste component analyses

Concentrations of non-volatile taste components, including soluble sugars, organic acids, free amino acids, and 5′-nucleotides, were determined as described by Samarasiri and Chen (2025a). The detailed analytical procedures are provided in the Supplementary Material. Briefly, soluble sugars were extracted following Giannoccaro, Wang, and Chen (2006) and analysed using HPLC equipped with a Supelcosil LC-NH2 HPLC Column (250 × 4.6 mm, 5 μm, Supelco) and Shimadzu RID-20 A refractive index detector (X. Yu, Yuan, Fu, & Zhu, 2016). Organic acids and 5′-nucleotides were extracted according to Wu, Tang, Pei, and Wang (2015) and Schmidt, Olsen, and Mouritsen (2020), respectively, and analysed by HPLC using a Shimadzu UV detector with an Inertsil ODS-3 column (250 × 4.6 mm, 5 μm, GL Sciences) and a Shim-pack GIST C18 column (250 × 4.6 mm, 5 μm, Shimadzu), respectively. Free amino acids (FAAs) were extracted and analysed as described by Samarasiri and Chen (2025a). FAAs were separated using a Zorbax HILIC Plus column (3.0 × 100 mm, 1.8 μm, Agilent) on an Agilent 1290 Infinity II LC system and detected using an Agilent 6546 LC/Q-TOF mass spectrometer.

2.7. Electronic tongue measurement

Samples were prepared through hot extraction of dried powder at 2.5% w/v as given by Samarasiri and Chen (2025a). The mixture (dried powder in deionized water) was stirred, heated at 40 °C for 15 min, allowed to stand at ambient conditions for 60 min, and then centrifuged at 4000 rpm and 10 °C for 10 min. The strained supernatant was allowed to reach ambient temperature before taste measurement. The sample conductivity and pH were maintained within 1–10 mS/cm and 4–6, respectively.

Electronic tongue taste measurement was carried out with the Taste Sensing System TS-5000Z (Intelligent Sensor Technology, Inc., Kanagawa, Japan) as mentioned by Samarasiri and Chen (2025b). The sensors used were C00 for bitterness and aftertaste-bitterness, AAE for umami and richness, AE1 for astringency and aftertaste-astringency, CA0 for sourness, and CT0 for saltiness. Preconditioning of two reference electrodes and sensors for 24 h was conducted with the preconditioning solution and reference solution, respectively. After configuring the measurement parameters (sensor set, measurement method, and sample set), assigning the appropriate solutions (negatively/positively charged membrane washing solutions and reference solution), and mounting the sensors, sensor checks were conducted. Taste intensity was determined from the potentiometric difference between each sensor and a reference electrode. Before this sample measurement, a maintenance measurement was executed to ensure a reliable sensor performance (Supplementary Material).

2.8. Statistical analysis

Experiments were performed in triplicate. OriginPro 2024b (OriginLab Corporation, Northampton, MA, USA) was used to execute statistical analyses, graph plotting, correlation analysis (Pearson's correlation coefficient), and principal component analysis (PCA). Significant differences were assessed using one-way analysis of variance (ANOVA), and mean comparisons were performed using Tukey's post hoc test (p < 0.05). Unfermented raw materials were used as control samples to represent baseline substrate composition prior to fermentation.

3. Results and discussion

3.1. Mycelial growth

Okara or mixed substrates of okara and either sawdust or wheat bran supported the growth of P. ostreatus (Fig. 1A). Generally, each substrate showed increasing growth with an approximately constant expansion rate over most of the colonization phase (Fig. 1B), consistent with typical apical growth dynamics in filamentous fungi, where peripheral hyphal tips maintain near-steady elongation through nutrient and energy translocation from older mycelium (Boswell, Jacobs, Davidson, Gadd, & Ritz, 2002; Trinci, 1971). This behaviour reflects the lignocellulose-degrading capacity of Pleurotus spp., in which extracellular enzymes facilitate the breakdown of structural polysaccharides into utilizable carbon sources. The growth substrate is an essential factor as it supplies nutrients, moisture, and structure required for Pleurotus spp. growth (Bellettini et al., 2019). Differences in growth rates among substrates (Table 2) likely reflect variations in substrate composition and physical properties, including C/N ratio, lignin accessibility, and porosity (Osunde, Olayinka, Fashina, & Torimiro, 2019). Similarly, Krupodorova, Barshteyn, Tsygankova, Sevindik, and Blume (2024) reported growth rates ranging from 9.0 to 15.0 mm/day, which may be attributed to strain-specific differences and substrate composition. Mechanical properties and material constitution of the lignocellulose biomass can significantly influence the mycelial growth. The highest mycelial growth rate was observed on FO, followed by OW0, while sawdust-only medium experienced the lowest growth rate, a longer lag phase, and the lowest mycelial density (Table 2 and Fig. 1B). Nashiruddin et al. (2022) also obtained a slower growth of P. ostreatus on sawdust with a lag phase of 2 days. Poor growth could be caused by nutrient unavailability and the lack of air voids (Elsacker, Vandelook, Brancart, Peeters, & De Laet, 2019). Particularly, on sawdust, more energy could be directed towards enzyme production for lignin degradation while limiting the resource availability for hyphal extension (Suwandecha & Pisuchpen, 2024).

Fig. 1.

P. ostreatus growth on okara and okara-derived substrates. (A) Mycelial growth progression from inoculation to full colonization. (B) Growth curves of mycelium over the incubation period. OW - Okara and wheat bran mixed media; DE - Dehydrated okara; FD - Freeze-dried okara; FO - Fresh okara; OS - Okara and sawdust mixed media.

Table 2.

Mycelial growth performance and moisture content of okara and okara-derived substrates fermented by P. ostreatus.

Medium	Mycelial growth rate (mm/day)	Average weight reduction rate (g/day)	Net weight reduction (g/16 g substrate)	Growth period	Growth density	Moisture content (%) before drying
OW0	7.12 ± 0.45b	0.05 ± 0.01bc	0.55 ± 0.10c	12 ± 0fg	++++	61.21 ± 2.87ab
OW25	5.81 ± 0.18bcd	0.04 ± 0.01abc	0.58 ± 0.10c	14 ± 1efg	+++	59.01 ± 5.42abc
OW50	4.90 ± 0.53cdef	0.05 ± 0.02c	0.83 ± 0.31bc	16 ± 0def	+++	55.29 ± 4.49bcd
OW75	4.22 ± 0.10def	0.04 ± 0.02abc	0.67 ± 0.27c	20 ± 2cd	+++	53.92 ± 0.81bcd
OW100	3.38 ± 0.54ef	0.04 ± 0.01bc	1.19 ± 0.18b	27 ± 1b	+++	47.77 ± 2.30d
DE	5.54 ± 1.24bcde	0.02 ± 0.01ab	0.40 ± 0.10c	18 ± 2de	+++	55.31 ± 2.86bcd
FD	4.05 ± 1.21def	0.02 ± 0.01a	0.36 ± 0.19c	23 ± 3c	+++	50.31 ± 3.33cd
FO	15.48 ± 1.08a	0.04 ± 0.01abc	0.40 ± 0.11c	11 ± 0g	++++	70.85 ± 0.18a
OS75	5.01 ± 0.69bcdef	0.03 ± 0.01abc	0.47 ± 0.07c	16 ± 3de	+++	63.72 ± 1.02ab
OS50	6.09 ± 1.20bcd	0.05 ± 0.02bc	0.66 ± 0.22c	14 ± 1efg	++	64.62 ± 1.43ab
OS25	6.98 ± 1.12bc	0.03 ± 0.01abc	0.43 ± 0.11c	14 ± 0efg	++	60.31 ± 0.57abc
OS0	3.25 ± 0.27f	0.04 ± 0.01bc	1.94 ± 0.37a	44 ± 1a	+	56.27 ± 1.70bcd

Data presented as means ± standard deviation (n = 5 for all except moisture content; n = 3 for moisture content%). Values in the same column with different superscripts are significantly different based on the Tukey test (p < 0.05). The apparent density of growth was classified as ++++ (highly dense, excellent), +++ (medium dense, good), ++ (soft, satisfactory), or + (faint, poor) based on Postemsky et al. (2019) and Postemsky, Bidegain, González-Matute, Figlas, and Cubitto (2017) with some modifications.

Favorable particle size distribution and composition are necessary for better colonization and binding of the medium. After about 30 days, tiny threads of mycelium reached the edges of the plate only in certain areas in the OS0, and the growth was not uniform. Suwandecha and Pisuchpen (2024) mentioned that a larger particle size of core wood sawdust allowed mycelium to spread over a larger surface area and bridge gaps between larger particles, leading to uneven and slow growth, while a smaller particle size led to a more compact and consistent arrangement in the structures. Coarser particle sizes may be desirable to prevent substrate compaction and maintain efficient oxygen diffusion, especially under high-humidity conditions. Advocate void spaces in the medium, also necessary for metabolic heat dissipation, whereas sufficient structural support for mycelial growth cannot be attained with excessive porosity (Shakir & Ahmad, 2024). Apart from the nutrient availability in the medium, mycelial density can also be influenced by oxygen exchange and heat buildup inside the substrate, substrate microstructure, pH, and incubation conditions (Magaña Amaya & Shimizu, 2025; Schritt, Vidi, & Pleissner, 2021). Despite the capability of P. ostreatus to degrade nitrogen-deficient lignocellulosic substrates (e.g., sawdust), optimal mycelial development can be achieved with adequate nitrogen supplementation, which supports enzyme production, biomass formation, and metabolic activity. In the present work, the addition of okara into sawdust significantly improved the mycelial growth, likely due to nitrogen enrichment from okara, which contained higher nitrogen (5.81 ± 0.12%) than sawdust (1.86 ± 0.26%) (Fig. 2).

Fig. 2.

Ultimate analysis results of raw materials, fermented okara, and fermented okara-derived substrates. Values of the columns in the same color with different letters are significantly different based on the Tukey test (p < 0.05). Fermented samples (OW - Okara and wheat bran mixed media; DE - Dehydrated okara; FD - Freeze-dried okara; FO - Fresh okara; OS - Okara and sawdust mixed media). Unfermented raw materials (WB-UF - Unfermented wheat bran; OK-UF - Unfermented okara; SD-UF - Unfermented sawdust).

Mycelial growth performance varied depending on the drying method of okara, as different drying parameters and durations may influence the nutrient composition and physical properties of the materials. Fresh okara provided the fastest mycelial growth and exhibited a great mycelial apparent density. The growth period on fresh okara was comparable to that reported by Samarasiri and Chen (2025a). Even though substrates of dried okara had comparatively lower growth rates and higher colonization time, their mycelial density was comparable to that in wheat bran-okara mixed media. The growth differences may be owing to the possible impacts of pre-drying treatments on nutrient retention, water holding capacity, porosity, and the degree of rehydration of okara (Asghar et al., 2023; Tian, Zhao, Huang, Zeng, & Zheng, 2016). These structural modifications can limit the enzyme accessibility to cellulose and lead to a lower enzymatic digestibility in dried substrate compared to the non-dried form (Koo, Jo, & Cho, 2020; Luo & Zhu, 2011). Prolonged colonization phase can also be discussed with the differences in air spaces. Flaky textures of sawdust and wheat bran could have facilitated better oxygen access for mycelial growth.

The mass changes during fermentation are attributable to the formation of mycelial biomass, the loss of the substrate, and the metabolic conversion of biomass to H₂O and CO₂ (W. Sun, Tajvidi, Howell, & Hunt, 2022). In the present study, all media exhibited weight reductions during SSF by P. ostreatus. The weight reduction of the growing medium represents the degradation rate of organic materials in the medium (Khoo et al., 2022). The weight loss of the medium during SSF is directly correlated to fungal metabolism, including oxygen consumption, carbon dioxide release, volatile products, water consumption in polysaccharide hydrolysis, and evaporation of water generated from carbohydrate metabolism (Dorta, Bosch, Arcas, & Ertola, 1994). Total weight reduction during the fermentation was highest in the sawdust-only medium, followed by hot air-dried okara. Compared to sawdust, which has higher levels of lignin, hemicellulose, and cellulose, wheat bran may have higher contents of easily assimilable nutrients. In addition, a limitation of this study is the absence of a non-inoculated incubation control, which could further distinguish microbial-driven changes from intrinsic substrate changes. However, unfermented raw materials were used as baseline controls, and observed changes are primarily attributed to P. ostreatus fermentation.

Different media could have different moisture levels and water uptake properties due to variations in their physicochemical and structural characteristics. The increase in fermentation time could increase dry matter weight because of higher microbial proliferation (Appels et al., 2019; Desta et al., 2021). The moisture content is important in fungal fermentation as it affects microbial growth (water activity and nutrient diffusion), substrate swelling, and enzyme stability (Borkertas et al., 2025). Even though most media initially had moisture contents within 54–57%, moisture levels mainly ranged between 47.77% and 64.62% after SSF (Table 2), reflecting changes driven by microbial growth and heat transfer (Jin et al., 2019). Moreover, the water-holding capacity of mycelium-based composites was found to rely on the growing medium (Camilleri, Narayan, Lingam, & Blundell, 2025). Freeze drying for 48 h decreased the moisture contents of all the fermented samples to below 10%, which may be beneficial for prolonged storage. However, future studies should include water activity (a_w) measurements together with microbiological stability evaluation, as a_w more directly reflects microbial stability and storage safety.

3.2. Elemental composition

During SSF, fungal assimilation of nutrients into biomass, enzymatic degradation of lignocellulosic components, and respiratory carbon loss as CO₂ together change the overall composition of the substrate and thereby alter the relative proportions of its elemental constituents (Wang et al., 2023). Therefore, mycelial growth and the breakdown of the substrate materials could have caused changes in C, H, and N contents of the media (Fig. 2). Overall, the current results corroborate the H% (6.99%) and C% (46.34%) reported for freeze-dried okara (S. Li et al., 2013). Additionally, C% of okara, as 50.2% mentioned by Taokaew, Nakson, Zhang, Kongklieng, and Kobayashi (2022), corresponded to the current result of unfermented okara.

Okara has been recognized as a protein-rich source (18.1–25%) (B. Li, Qiao, & Lu, 2012; Lian, Luo, Gong, Zhang, & Serventi, 2020). Increasing the amount of okara added led to a rise in the nitrogen level of fermented mixed substrates. It may also be caused by improved biomass production, as a proper C/N ratio is one of the crucial parameters for enhanced mycelial growth. Although higher C/N ratios (28–55) may be favorable during the fruiting stage of P. ostreatus (Kong et al., 2020; Samarasiri & Chen, 2025b), enhanced mycelial growth and biomass accumulation have been reported on substrates with lower C/N ratios (10–10.5), such as spent mushroom substrate and okara (Economou, Diamantopoulou, & Philippoussis, 2017; Samarasiri & Chen, 2025a). In this work, cultivation media had C/N ratios spanning from 8.19 to 25.99 (Table 1). Since OS0 demonstrated a nitrogen deficiency leading to a higher C/N ratio, it suggests the possibility of limited mycelial expansion and branching rate (Enriquez-Medina et al., 2024), which was confirmed with the findings discussed earlier.

Compared to unfermented okara, fermented okara (OW100) gained an improved level of nitrogen, probably owing to mycelial development. This nitrogen increase could be caused by the generation of proteins, amino acids, and nucleic acids, which are all essential for the synthesis of cell structures and the production of enzymes during fungal metabolism (J. Zhang et al., 2023). Similarly, when okara was fermented by R. oligosporus or A. oryzae, N% increased from 4.196% to 6.168% and 7.200%, respectively (Sitanggang, Sinaga, Wie, Fernando, & Krusong, 2020). In previous investigations, an increasing pattern was also observed in wheat bran and wheat bran-mixed media, as reflected by increases in crude protein content after fermentation (Gu et al., 2024; Mohammady, Aboseif, Soaudy, Ramadan, & Hassaan, 2023). Similar upward trends in nitrogen levels were identified in paddy straw, coffee husk, and areca husk fermented using F. verticillioides and M. circinelloides (Geethanjali, Gowtham, & Jayashankar, 2020) and mixtures containing hemp shives or birch sawdust with wheat bran and birch bark valorised by T. versicolor (Irbe, Loris, Filipova, Andze, & Skute, 2022). Taokaew et al. (2022) obtained N% of 2.86% and 10.4% in okara and okara-extracted protein, respectively. In the present study, four types of fermented okara had N levels between 6.14% and 6.97%, which were mostly higher than those of fermented mixed media. OW100 had the maximum N%, followed by FO and OW75. In terms of achieving a protein-enriched fermented product (∼27.73–30.24% dry weight; Supplementary Material), these substrate formulations would be desirable.

On the other hand, the okara-only media reduced their C content by 4.17–11.39% after SSF, while wheat bran and sawdust-containing media showed a decrease in C level of 0.15–9.14%. White-rot fungi consume organic matter in agricultural residues as an energy and carbon source during lignocellulose degradation. The carbon% drop was ascribed to the fungal metabolism as fungi convert carbon to CO2 and energy (Geethanjali et al., 2020; Wang et al., 2023). A higher decrease of C% was achieved with the increase of okara% in the medium, which could be attributed to the increased mycelial growth and substrate utilization. Furthermore, most fermented media had a lower C/N ratio than the C/N ratio of their unfermented forms, indicating a release of gaseous metabolites and metabolic changes with SSF.

3.3. Soluble sugars

Differences in substrates can result in different C/N ratios for mushroom development and lead to variations in soluble sugar concentrations of mushrooms (Samarasiri & Chen, 2022), contributing to the sweet taste and also suppressing bitter taste. For example, the addition of rice straw into the substrate reduced the level of trehalose, whereas enhancing the concentrations of arabitol and mannitol (Gao et al., 2020). Drying techniques can also influence soluble sugars owing to thermal degradation, enzymatic reactions, and Maillard reactions (Samarasiri & Chen, 2022). Accordingly, the concentrations of soluble sugars in fermented media (Table 3) were significantly affected (p < 0.05) by the substrate composition and okara pre-treatment method. Glucose was the predominant sugar in all fermented samples, consistent with the findings of Samarasiri and Chen (2025a).

Table 3.

Contents of soluble sugars, organic acids, and 5′-nucleotides in raw materials, fermented okara, and fermented okara-derived substrates.

Medium	Content (mg/g dry weight)
	Glucose	Fructose	Sucrose	Trehalose	Citric acid	Fumaric acid	Malic acid
OW0	23.65 ± 0.95a	1.63 ± 0.16c	4.34 ± 0.39c	14.06 ± 2.07a	5.20 ± 0.94e	0.59 ± 0.08ef	10.62 ± 1.80bc
OW25	23.55 ± 0.55a	1.83 ± 0.35c	7.67 ± 0.45b	14.12 ± 1.28a	7.85 ± 0.74cde	0.69 ± 0.04de	13.52 ± 1.70ab
OW50	24.75 ± 1.03a	2.29 ± 0.29c	4.15 ± 0.15c	8.97 ± 1.30bc	11.10 ± 1.89bcd	0.56 ± 0.08ef	7.35 ± 0.97cd
OW75	18.82 ± 1.23b	2.51 ± 0.66c	3.63 ± 0.09cd	5.72 ± 0.72cd	14.33 ± 1.28b	0.96 ± 0.10cd	5.87 ± 0.96cd
OW100	27.81 ± 2.69a	3.15 ± 1.05c	5.03 ± 0.63c	4.28 ± 0.17d	ND	2.20 ± 0.11a	13.11 ± 1.43ab
DE	16.94 ± 1.41b	8.45 ± 1.25b	4.70 ± 0.27c	4.53 ± 1.12d	6.18 ± 1.19de	1.19 ± 0.06c	16.49 ± 1.67a
FD	17.34 ± 0.74b	6.75 ± 0.30b	8.21 ± 0.77b	4.35 ± 0.25d	14.00 ± 0.08b	1.75 ± 0.13b	13.32 ± 0.69ab
FO	14.95 ± 0.73b	1.81 ± 0.31c	2.25 ± 0.82d	9.62 ± 0.72b	ND	0.75 ± 0.11de	16.12 ± 1.85a
OK-UF	4.35 ± 0.40c	16.93 ± 0.97a	3.51 ± 0.74cd	3.39 ± 0.07d	19.75 ± 3.26a	0.24 ± 0.00g	ND
WB-UF	8.50 ± 0.90c	8.18 ± 1.32b	13.52 ± 0.28a	4.09 ± 0.91d	11.49 ± 0.95bc	0.32 ± 0.00fg	4.01 ± 0.33d

Medium	Content (mg/g dry weight)
	Oxalic acid	Succinic acid	Tartaric acid	5′-AMP	5′-CMP	5′-GMP	5′-UMP
OW0	2.43 ± 0.70ab	33.01 ± 4.70d	8.16 ± 1.45de	0.37 ± 0.03cde	0.37 ± 0.08fg	0.32 ± 0.00def	0.84 ± 0.09d
OW25	2.46 ± 0.50ab	54.46 ± 2.50c	9.24 ± 1.77de	0.53 ± 0.08bcd	0.51 ± 0.13ef	0.47 ± 0.06cde	1.05 ± 0.12bcd
OW50	3.17 ± 0.29a	55.28 ± 5.58c	11.22 ± 1.00cd	0.58 ± 0.13bc	0.61 ± 0.12def	0.61 ± 0.12bc	1.32 ± 0.03b
OW75	1.28 ± 0.20b	46.01 ± 4.66cd	16.73 ± 1.70c	0.23 ± 0.05ef	0.81 ± 0.08cde	0.33 ± 0.03de	0.99 ± 0.05cd
OW100	1.91 ± 0.73ab	82.58 ± 2.57b	63.51 ± 3.38a	0.71 ± 0.15b	1.37 ± 0.08a	0.69 ± 0.03ab	1.81 ± 0.07a
DE	2.34 ± 0.66ab	102.62 ± 6.41a	23.75 ± 2.24b	0.51 ± 0.09bcd	1.03 ± 0.08bc	0.51 ± 0.02bcd	1.19 ± 0.05bc
FD	1.29 ± 0.06b	78.68 ± 7.23b	29.15 ± 1.11b	0.31 ± 0.04def	0.91 ± 0.12bcd	0.32 ± 0.02e	1.18 ± 0.12bc
FO	2.61 ± 0.82ab	41.37 ± 6.23cd	7.94 ± 1.99de	1.13 ± 0.02a	1.21 ± 0.12ab	0.85 ± 0.07a	0.85 ± 0.06d
OK-UF	1.46 ± 0.21ab	30.77 ± 4.98d	ND	0.06 ± 0.00f	0.11 ± 0.02g	0.11 ± 0.01f	0.44 ± 0.02e
WB-UF	1.87 ± 0.15ab	2.21 ± 0.19e	4.62 ± 0.73e	ND	0.15 ± 0.02g	ND	0.43 ± 0.13e

Data presented as means ± standard deviation (n = 3). Values in the same column with different superscripts are significantly different based on the Tukey test (p < 0.05). ND - Not detected. Fermented samples (OW - Okara and wheat bran mixed media; DE - Dehydrated okara; FD - Freeze-dried okara; FO - Fresh okara). Unfermented raw materials (OK-UF - unfermented okara; WB-UF - unfermented wheat bran). Acetic acid and 5′-IMP were not detected in all samples. Umami flavor nucleotides were considered as 5’-AMP + 5’-GMP.

The cell wall of lignocellulosic substrates is mainly composed of polysaccharides, predominantly cellulose, together with hemicellulose and lignin. Nemes et al. (2025) observed variations in soluble sugar levels in the substrate due to enzymatic hydrolysis of lignocellulosic components and microbial metabolic activity during fermentation. Glucose concentration was significantly enhanced in all the media after SSF (∼2.78–6.39-fold improvements). This could be attributed to the ability of P. ostreatus to hydrolyze the lignocellulosic components in okara and wheat bran and convert them into various sugars. This capability is associated with the production of cellulolytic enzymes by P. ostreatus when cultivated on different agro-industrial residues (Dedousi, Melanouri, & Diamantopoulou, 2023; Melanouri, Dedousi, & Diamantopoulou, 2022). Moreover, Hu et al. (2024) reported the ability of beta-glucosidase to convert cellobiose and oligosaccharides into glucose for fungal consumption. The highest glucose content was achieved from OW100, which was not statistically different from the treatments with 50–100% wheat bran. Both okara and wheat bran could contain a high level of insoluble dietary fiber (35–55.63%). The present findings implied that the oyster mushroom mycelium could perform the hydrolysis of their insoluble dietary fiber, resulting in the liberation of glucose (Lu, Liu, & Li, 2013; Stevenson, Phillips, O'sullivan, & Walton, 2012). Comparable to the trends in the present work, Nemes et al. (2025) showed a rise in glucose level from 5.167 mg/g to 17.728 mg/g and a reduction of fructose level from 7.179 mg/g to an undetectable level in wheat bran after SSF via A. niger.

The changes in sugar profile after fermentation indicate the fungus's preferential sugar consumption and enzymatic breakdown of polysaccharides (Nemes et al., 2025). All the fermented media had significantly lower fructose contents than their respective unfermented raw materials. These levels of fructose were within the low fructose levels achieved in several mushroom-related past works (W. Li et al., 2014; F. Wu et al., 2015). On the other hand, the sucrose level was highest in unfermented wheat bran. Most fermented samples did not comprise significantly distinct concentrations of sucrose relative to unfermented okara. Based on past metabolomics analysis, using B. subtilis increased the sucrose content of fermented okara (Zhan et al., 2023), whereas Vong, Hua, and Liu (2018) revealed a reduction in sucrose level after okara fermentation using Y. lipolytica and R. oligosporus, indicating a dependence on the strain and fermentation parameters used in fermenting okara.

Trehalose, which functions as an energy reserve supporting fungal growth and reproduction, and as a protective agent against stress (Zhou, Ma, Zhang, & Zhang, 2016), had significantly higher levels in the fermented mixed media and FO than in unfermented raw materials. Trehalose was shown to be one of the dominant soluble sugars in mushrooms (Samarasiri & Chen, 2022), and thus, the production of trehalose by P. ostreatus could occur during the fermentation of okara and wheat bran. During P. ostreatus mycelial expansion, a portion of the glucose formed through cellulose breakdown was converted into trehalose (Zhou et al., 2016).

3.4. Organic acids

Variations in organic acid levels during SSF are closely associated with carbohydrate metabolism, tricarboxylic acid (TCA) cycle activity, and the bioconversion of lignocellulosic-derived sugars released during substrate degradation (Y. Wu et al., 2021; Zhou, Ding, Han, & Deng, 2023). Organic acids, generated as metabolic byproducts, can accumulate during fungal fermentation (Chai, Ng, Samarasiri, & Chen, 2022; Chin et al., 2025). Overall, fermentation enhanced the levels of organic acids in substrate formulations (Table 3), in agreement with Samarasiri and Chen (2025a). Regarding the development of organic acids in relation to drying techniques of okara, the highest was in the dried okara medium. With respect to OW0, mixing okara with wheat bran enhanced concentrations of organic acids. It can be due to their variations in initial level of organic acids, substrate composition, and C/N ratio, which could influence microbial metabolism (Zwinkels, Van Oorschot, Van Mastrigt, & Smid, 2026). Previous investigations also observed the generation of organic acids in wheat bran and okara through fermentation (Ghamry, Zhao, & Li, 2023; Nemes et al., 2025; Vong et al., 2018). Even though the variations of concentrations of organic acids in mushrooms with different drying methods (improved release or decarboxylation) or substrate compositions have been investigated in previous studies (Gao et al., 2020; Li et al., 2015; Zhang et al., 2021), such evaluations could not be retrieved for fermented okara from the available studies.

Succinic acid, contributing to umami flavor development, was the predominant organic acid in all fermented media and unfermented okara, while citric acid represented the major organic acid in raw unfermented wheat bran. Following the fermentation, succinic acid contents showed 1.34- to 14.92-fold elevations. Citric acid can provide refreshing and mild sourness. However, compared to citric acid levels of HongJun Tofu fermented by multiple strains (21.84–45.36 mg/g) (Qiu et al., 2023), all the present samples gained lower concentrations. Differences could be attributed to the okara cultivar, fermentation parameters, or extraction techniques. In the current work, fermentation via P. ostreatus led to a 1.23–3.20-fold reduction in citric acid concentrations. Reflecting a similar trend, a reduction of citric acid content was also observed during the stationary phase of okara biotransformation via Y. lipolytica, where the observed reduction in citric acid levels was associated with metabolism via the glyoxylate pathway and/or TCA cycle (Vong, Au Yang, & Liu, 2016). As citrate can be converted to succinate through a series of intermediates in the TCA cycle, this metabolic activity may account for the observed increase in succinic acid and decreased citric acid concentrations following SSF.

Fumaric acid had the lowest concentration in all the samples, and it is consistent with several past works relevant to okara and mushrooms (Qiu et al., 2023; J. Wang, Li, Li, Wu, & Tang, 2018; F. Wu et al., 2015). Most of the present levels of fumaric acid are consistent with observations by Qiu et al. (2023) for HongJun Tofu, which was made using multiple strains. In most current media, fumaric acid levels were significantly enhanced after SSF, and a similar tendency was also noticeable in soybean residue fermented by R. oligosporus or L. plantarum (Gupta & Chen, 2021). Malic acid contains a refreshing acidity with a marginal bitterness (W. Chen et al., 2015). Concentrations of malic acid were also considerably enhanced after SSF by P. ostreatus. Past studies have similarly reported increases of malic, succinic, and fumaric acids in fermented okara, which may result from the accumulation of TCA cycle intermediates (Gupta, Lee, & Chen, 2018; Qiu et al., 2023; Vong et al., 2018). The differences in malic acid contents after fermentation were also reported for other fermented foods (J. Yu et al., 2023; Zhu et al., 2021).

On the other hand, concentrations of oxalic acid in fermented samples were not significantly different from those of unfermented raw materials, and the present data approximate the oxalic acid levels given for fermented okara via Y. lipolytica and R. oligosporus (Vong et al., 2018). Except for the sour taste, tartaric acid can impact the perception of astringency. Tartaric acid was only detected in fermented samples and raw wheat bran, and it had a maximum level in OW100. Correspondingly, Wang et al. (2016) determined that P. eryngii mycelium grown in potato dextrose liquid had a 72.56 mg/g tartaric acid concentration. This was a deviation from the findings of Gupta and Chen (2021), in which tartaric acid was not detected in okara after performing fermentation using either R. oligosporus or L. plantarum. In addition, as observed for fresh okara and okara fermented by R. oligosporus (Vong et al., 2018), acetic acid was not detected in the present samples.

3.5. Free amino acids

Free amino acids have been recognized as one of the key contributors to food flavor. Concentrations of FAAs in fermented samples and raw materials were significantly changed (p < 0.05) after fermentation (Table 4) and influenced by both substrate composition and okara pre-drying technique. Variations in free amino acid profiles during SSF may arise from enzymatic hydrolysis of substrate proteins, along with fungal biosynthesis, biomass formation, amino acid utilization during metabolic activity, and their conversion into flavor-related compounds (Das et al., 2023; J. Sun et al., 2025). All the fermented substrates obtained significantly higher FAA levels compared to raw materials, suggesting that mycelial proteolysis was more profound than amino acid catabolism. Extracellular proteases formed by P. ostreatus may have hydrolyzed proteins in okara and wheat bran and led to the release of FAAs. Similar net increases of amino acids or FAAs were also observed in several past works relevant to the biotransformation of okara with Y. lipolytica, R. oligosporus, B. subtilis, and Aspergillus spp. (Ichikawa et al., 2022; Keong, Toh, Lu, & Liu, 2023; Vong et al., 2018).

Table 4.

Free amino acid (FAA) contents of raw materials, fermented okara, and fermented okara-derived substrates.

FAA	Content (mg/g dry weight)
	WB-UF	OK-UF	OW0	OW25	OW50
Phe	0.04 ± 0.00d	0.89 ± 0.03bc	0.80 ± 0.04c	0.91 ± 0.03b	1.14 ± 0.01a
Leu	0.05 ± 0.00g	1.34 ± 0.03a	0.51 ± 0.03f	0.55 ± 0.02ef	0.91 ± 0.02d
Tyr	0.04 ± 0.00f	0.61 ± 0.02e	0.80 ± 0.04d	0.94 ± 0.04cd	1.19 ± 0.03ab
Ile	0.05 ± 0.00h	0.37 ± 0.02g	1.41 ± 0.07bc	1.52 ± 0.04ab	1.62 ± 0.06a
Met	0.01 ± 0.00g	0.26 ± 0.01b	0.10 ± 0.01f	0.11 ± 0.01ef	0.21 ± 0.00c
Val	0.13 ± 0.01f	0.48 ± 0.02e	1.79 ± 0.06a	1.96 ± 0.05a	1.90 ± 0.03a
Ala	0.29 ± 0.00f	0.90 ± 0.02f	2.52 ± 0.31de	2.74 ± 0.21cd	3.34 ± 0.16bc
Thr	0.08 ± 0.00g	0.69 ± 0.02f	1.80 ± 0.06bc	1.89 ± 0.05b	1.92 ± 0.03b
Pro	0.11 ± 0.01g	0.47 ± 0.01f	1.71 ± 0.07a	1.59 ± 0.06a	1.32 ± 0.04b
Ser	0.09 ± 0.00g	0.36 ± 0.01f	1.90 ± 0.04bc	2.04 ± 0.10ab	2.00 ± 0.06abc
Gly	0.08 ± 0.01f	0.22 ± 0.01e	0.79 ± 0.05bc	0.89 ± 0.03ab	0.81 ± 0.03bc
Glu	0.78 ± 0.00h	1.09 ± 0.04g	1.85 ± 0.04f	1.81 ± 0.04f	2.23 ± 0.03e
Arg	0.34 ± 0.05e	0.78 ± 0.06e	2.03 ± 0.08d	2.24 ± 0.12d	3.47 ± 0.33a
Lys	0.07 ± 0.00f	0.40 ± 0.06bc	0.26 ± 0.03e	0.28 ± 0.02de	0.39 ± 0.03cd
Cys	0.01 ± 0.00d	0.01 ± 0.00d	0.03 ± 0.00c	0.03 ± 0.00 c	0.03 ± 0.00c
Asp	0.76 ± 0.02b	0.46 ± 0.03cd	0.51 ± 0.07cd	0.57 ± 0.02cd	0.63 ± 0.00bc
His	0.04 ± 0.00f	0.14 ± 0.01f	0.62 ± 0.05de	0.71 ± 0.01bcd	0.66 ± 0.05cde
Bitter	0.65 ± 0.06f	4.25 ± 0.14e	7.27 ± 0.25cd	8.00 ± 0.03bc	9.91 ± 0.49a
Umami	1.54 ± 0.02g	1.54 ± 0.06g	2.37 ± 0.06f	2.38 ± 0.05f	2.86 ± 0.03e
Sweet	0.66 ± 0.02f	2.64 ± 0.07e	8.72 ± 0.45c	9.15 ± 0.33bc	9.39 ± 0.01bc
Tasteless	0.11 ± 0.01g	1.02 ± 0.07f	1.09 ± 0.04ef	1.25 ± 0.02de	1.61 ± 0.02b
Total	2.96 ± 0.10g	9.46 ± 0.31f	19.45 ± 0.76d	20.78 ± 0.37cd	23.78 ± 0.53b

FAA	Content (mg/g dry weight)
	OW75	OW100	DE	FD	FO
Phe	0.99 ± 0.01b	0.92 ± 0.01b	1.11 ± 0.05a	1.09 ± 0.04a	1.12 ± 0.03a
Leu	0.96 ± 0.02cd	0.71 ± 0.01e	1.10 ± 0.06bc	0.97 ± 0.04cd	1.19 ± 0.12ab
Tyr	0.90 ± 0.02d	0.96 ± 0.00cd	1.28 ± 0.11a	1.05 ± 0.05bc	1.22 ± 0.03a
Ile	0.71 ± 0.03f	1.31 ± 0.03c	1.26 ± 0.09cd	1.01 ± 0.06e	1.08 ± 0.07de
Met	0.13 ± 0.00e	0.18 ± 0.00d	0.26 ± 0.02b	0.16 ± 0.01d	0.31 ± 0.01a
Val	0.69 ± 0.03e	1.52 ± 0.03b	1.31 ± 0.12bc	0.97 ± 0.06d	1.12 ± 0.10cd
Ala	1.95 ± 0.08e	5.19 ± 0.04a	3.90 ± 0.24b	3.87 ± 0.19b	4.69 ± 0.36a
Thr	1.23 ± 0.01e	2.14 ± 0.04a	1.71 ± 0.09c	1.47 ± 0.05d	1.72 ± 0.03c
Pro	0.97 ± 0.07de	1.15 ± 0.05bc	1.26 ± 0.06b	1.07 ± 0.04cd	0.89 ± 0.06e
Ser	1.17 ± 0.03e	2.16 ± 0.06a	1.78 ± 0.12c	1.51 ± 0.06d	1.89 ± 0.04bc
Gly	0.58 ± 0.02d	0.88 ± 0.02b	1.00 ± 0.05a	0.84 ± 0.04b	0.71 ± 0.03c
Glu	2.38 ± 0.02d	2.94 ± 0.05c	3.30 ± 0.08b	2.85 ± 0.06c	3.95 ± 0.02a
Arg	2.92 ± 0.12bc	2.83 ± 0.08c	3.87 ± 0.09a	3.45 ± 0.25ab	3.85 ± 0.08a
Lys	0.37 ± 0.02cd	0.40 ± 0.03bc	0.50 ± 0.03b	0.42 ± 0.03bc	0.64 ± 0.04a
Cys	0.03 ± 0.00c	0.03 ± 0.00bc	0.04 ± 0.00b	0.03 ± 0.00c	0.10 ± 0.01a
Asp	0.39 ± 0.01d	0.79 ± 0.04b	0.75 ± 0.03b	0.43 ± 0.01d	2.41 ± 0.13a
His	0.57 ± 0.03e	0.75 ± 0.05bc	0.93 ± 0.06a	0.79 ± 0.02b	0.97 ± 0.03a
Bitter	6.97 ± 0.09d	8.22 ± 0.14b	9.84 ± 0.30a	8.44 ± 0.39b	9.63 ± 0.23a
Umami	2.77 ± 0.03e	3.73 ± 0.09c	4.05 ± 0.10b	3.28 ± 0.05d	6.37 ± 0.11a
Sweet	5.91 ± 0.15d	11.51 ± 0.19a	9.65 ± 0.10bc	8.76 ± 0.36c	9.90 ± 0.48b
Tasteless	1.30 ± 0.01d	1.39 ± 0.03cd	1.82 ± 0.08a	1.50 ± 0.07bc	1.96 ± 0.03a
Total	16.95 ± 0.22e	24.86 ± 0.23b	25.36 ± 0.46b	21.98 ± 0.77c	27.85 ± 0.60a

Data presented as means ± standard deviation (n = 3). Values in the same row with different superscripts are significantly different based on the Tukey test (p < 0.05). FAA categorization was based on Samarasiri and Chen (2022): Umami taste FAAs (Glu, Asp); Sweet taste FAAs (Ala, Gly, Pro, Ser, Thr); Bitter taste FAAs (His, Ile, Leu, Met, Phe, Val, Arg); Tasteless FAAs (Lys, Tyr, Cys).

After SSF, wheat bran-only medium gained about a 6.58-fold increase in total FAA content. Mao et al. (2020) also showed a rise in total FAA levels of sterilized wheat bran after SSF, as hydrolysis was supported by proteases synthesized by E. faecalis M2 used. Total FAA content was further increased in wheat bran mix of 25–50% okara, whereas okara-only substrates had 2.32–2.95-fold enhancements. Nevertheless, the highest content of total FAA was achieved in fermented okara-only substrates, FO, followed by DE and OW100. Samarasiri and Chen (2025a) also observed an increase in FAA concentrations following SSF by P. ostreatus, and fermented okara exhibited total FAA levels ranging between 14.66 and 23.29 mg/g, depending on the strain.

According to Komata in 1969, FAAs can be categorized into umami (MSG-like), bitter, tasteless, and sweet amino acids. The mushroom's characteristic flavor is substantially supplied by glutamic and aspartic acids, which are recognized for their MSG-like flavor-enhancing properties (Samarasiri & Chen, 2022). Relative to respective raw materials, 2.13–4.13-fold improvements of umami taste amino acids could be seen in fermented okara, while it was only a 1.54-fold increase in fermented wheat bran, suggesting a possibility of better meaty taste development in okara fermented by P. ostreatus. Similarly, bioconversion via strain B. subtilis R0179 elevated umami amino acids in soybean residue from 42.92 to 112.73 mg/g, indicating glutaminase activity (Keong et al., 2023). Mao et al. (2020) presented a concentration of umami taste amino acids in raw wheat bran similar to the present study's data (1.54 mg/g). Moreover, thermally treated okara (DE and OW100) enhanced the production of FAAs (including umami taste amino acids), compared to freeze-dried okara. In mushrooms, drying was identified as promoting the formation of FAAs from proteins and Maillard reactions involving sugars and amino acids (Samarasiri & Chen, 2022). Nevertheless, FO provided the highest umami taste amino acid concentrations, over the dried media, which can likely be attributed to the earlier observed better mycelial growth performance on fresh okara. Additionally, based on the Taste Activity Values (TAVs) of all the fermented samples (Table S1), two umami amino acids had the greatest taste activity, suggesting that Glu and Asp were the major FAAs responsible for the taste of substrates fermented by P. ostreatus.

According to TAVs of most fermented media, Arg, Val, His, Phe, and Ile were higher in bitter taste contribution, while Ser and Ala were enriched among sweet taste amino acids. Alanine was shown to be the dominant FAA in most of the fermented media. OW100 had the maximum level of sweet taste amino acids, whereas bitter taste amino acids were significantly abundant in OW50, DE, and FO. In mushrooms, the perception of bitter taste is often reduced when polyols, soluble sugars, and sweet amino acids are present (Lee, Jian, & Mau, 2009). Thus, there may be a possibility of bitter taste suppression after fermentation using P. ostreatus.

3.6. 5′-nucleotides

Levels of 5′-nucleotides were significantly influenced (p < 0.05) by the variations in the medium (Table 3). There is a significant lack of investigations about 5′-nucleotides in okara or their variations with fermentation; only two available studies could be retrieved (Qiu et al., 2023; Samarasiri & Chen, 2025a). The total contents of 5′-nucleotides (2.44–5.75 mg/g) in HongJun Tofu fermented by multiple strains agree with those of fermented okara in the current work (Qiu et al., 2023). The total concentrations of 5′-nucleotides in substrates fermented by P. ostreatus were significantly greater than those of raw materials. A similar tendency was seen in wheat after fermentation by G. frondosa mycelium with increased 5′-nucleotide level from 0.56 mg/g to 3.59 mg/g (Huang et al., 2011). The levels of 5′-nucleotides are attributable to nucleic acid degradation by nucleases (phosphodiesterases) and microbial metabolic processes (L. Sun et al., 2020; Zhang et al., 2023). The de novo and salvage pathways are the main routes for nucleotide biosynthesis. Particularly, compared with amino acid profiling, 5′-nucleotide profiling in fermented foods remains a relatively underexplored area, requiring further studies on content evaluation and associated metabolic pathways.

Variations in substrate composition and formulation influence the flavor 5′-nucleotide composition of mushrooms (Samarasiri & Chen, 2022). For example, the flavor 5′-nucleotides of T. melanosporum mycelia grown on media with the additions of soybean flour, soybean flour + corn syrup, or corn syrup gained about 6.0-fold, 4.7-fold, and 2.6-fold increases, respectively, compared to the control (Liu, Li, & Tang, 2012). In terms of the fermented mixed media, higher levels of total 5′-nucleotides were achieved in OW25 and OW50. However, the maximum concentrations of total 5′-nucleotides were provided by OW100 and FO, indicating that hot air-dried okara and fresh okara were better as growing media for gaining more 5′-nucleotides than freeze-dried okara. However, after SSF, all samples were subjected to freeze-drying. High porosity and treatments with low temperature could prevent the degradation of 5′-nucleotides (X. Li et al., 2015). 5′-nucleotide degradation could be affected by storage and temperature fluctuations, consumption of 5′-nucleotides for microbial growth and metabolism, and activity of nucleotide-metabolizing enzymes during food processing (Cao, Song, Mu, Sun, & Su, 2025; Yin et al., 2023).

In all the samples, either 5’-UMP or 5’-CMP was the dominating 5′-nucleotide. During the fermentation process, 5′-CMP represented the major 5′-nucleotide in sufu, a Chinese fermented soybean delicacy made via M. racemosus (Xi et al., 2022). 5’-IMP was not detected, as observed by Rotola-pukkila, Yang, and Hopia (2019) and Qiu et al. (2023). Within the de novo purine pathway, 5’-IMP functions as an intermediate and precursor for 5’-AMP and 5’-GMP (Ledesma-Amaro, Buey, & Revuelta, 2015). Therefore, this absence of detection could be due to the conversion or degradation of 5’-IMP via specific enzymatic activities or detection limitations. 5’-GMP levels increased after SSF in all samples. Similarly, when P. fermentans and L. fermentum were co-cultured on soybean protein hydrolysates for 36 h, 5’-GMP concentration elevated from 0.64 to 1.44 mg/100 ml, and the microbial enzymatic activity was reported to promote GMP formation from guanosine precursors and enhance umami flavor (Cao et al., 2025).

5’-GMP, 5’-AMP, 5’-XMP, and 5’-IMP are umami flavor contributors and have synergistic effects with umami taste amino acids (Manninen, Rotola-pukkila, Aisala, Hopia, & Laaksonen, 2018; Schmidt et al., 2020). FO achieved the maximum content of umami flavor 5′-nucleotides (1.97 ± 0.08 mg/g), followed by OW100 (1.41 ± 0.14 mg/g). Considering the flavor 5′-nucleotide classification by Yang, Lin, and Mau (2001), FO, OW100, OW50, and DE were found to belong in the middle range (1–5 mg/g), while the others were in the low range (<1 mg/g). Based on taste thresholds for 5’-AMP (0.125 mg/g) and 5’-GMP (0.255 mg/g) (Duan, Huang, Xiao, Zhang, & Tang, 2020), all the fermented products exceeded the taste thresholds of 5’-AMP and 5’-GMP.

3.7. Electronic tongue taste profile

Variations in taste profiles during fermentation may result from enzymatic degradation of macromolecules and microbial metabolism, which alter the concentrations and interactions of taste-active compounds, thereby influencing overall taste perception (Chen et al., 2021). Changes in the substrate composition and the okara pre-treatment led to significant (p < 0.05) variations in e-tongue taste responses (Fig. 3A). Both bitterness and astringency increased after fermentation. It may be owing to the rise of bitter-taste amino acids (previously observed) and bioactive compounds like polyphenols and alkaloids, which contribute to these two tastes (Wen et al., 2024; X. Zhao, Wei, Gong, Xu, & Xin, 2020). Aftertaste-bitterness and aftertaste-astringency showed only slight variations, as observed for the bioconversion of okara using probiotics and mixed yeast (Shi et al., 2020).

Fig. 3.

Electronic tongue evaluation results of raw materials, fermented okara, and fermented okara-derived substrates. (A) Radar charts. (B) Principal component analysis. Aftertaste-B (Aftertaste-bitterness); Aftertaste-A (Aftertaste-astringency). Taste responses are based on the interpolation difference (taste difference between samples), relative to the control, which was unfermented dried okara.

Unfermented wheat bran and all the fermented samples contained lower sourness compared to unfermented okara. In contrast, Shi et al. (2020) reported a rise in sourness after the biotransformation of okara by S. cerevisiae and Hansenula sp., due to the acid formation by yeast under aerobic conditions in the fermentation tank. The quantities gained for sourness were negative (Table S2), as in the past studies associated with mushroom/fermented okara (Samarasiri & Chen, 2025a; Shi et al., 2020; X. Zhao et al., 2020). OW0, OW25, OW100, and FD did not have a recognizable sour taste as the values of their sourness (Table S2) were lower than the threshold (−13) given by Intelligent Sensor Technology, Inc. Echoing the present pattern, wheat bran subjected to fermentation via yeast and lactic acid bacteria also achieved a sourness drop (H. M. Zhao, Guo, & Zhu, 2017). OW0 and FD, which had a greater umami taste, had relatively lower intensities of sourness. This could be related to a possible inhibitory effect of umami taste on sourness perception (Wang et al., 2023).

The umami taste, richness (umami aftertaste), and saltiness of fermented samples were higher than those of unfermented raw materials. The increase of umami taste through fermentation is consistent with popular fermented products like soy sauce, natto, tempeh, and miso, and fermented soy-related products made using C. glutamicum, L. plantarum, N. crassa, and M. guilliermondii (Koetke, Miller, & Deutsch, 2024; Qiu et al., 2023; Shangguan et al., 2024). The highest umami taste was achieved by OW0, followed by FD and OW100, whereas the maximum richness was provided by OW100. Aspartic acid and glutamic acid are two abundant amino acids in wheat bran and okara proteins, and fermentation could release higher contents of these FAAs through protein hydrolysis, leading to an umami taste increase (Chen et al., 2021 Florence, 2019; Koetke et al., 2024; Y. Li et al., 2023). As previously discussed, the concentrations of these umami FAAs increased after the fermentation. Since mushrooms are valued for their dominant umami taste substances (Phat, Moon, & Lee, 2016), the SSF of okara by P. ostreatus could be a promising path to enhance umami taste. Fermentation caused fresh okara to develop a lower umami taste compared to dried okara, highlighting the effects of fermentation and drying pretreatment on flavor. Conversely, the umami taste of dried shiitake mushrooms was higher than that of their fresh form (Hou et al., 2021).

Fermentation has also been identified as a method for enhancing the richness of food. The lower richness in unfermented okara can be elucidated by the formation of umami taste compounds during fermentation and the probable suppressive effect of okara polysaccharides on richness perception (Imamura & Matsushima, 2013). In contrast to umami taste variations, the richness of the fermented samples was escalated when more okara was integrated into the mixed medium. Richness is enhanced not only by umami taste components but also by the presence of some aroma-active compounds such as phenylacetaldehyde, (E,E)-2,4-decadienal, 1-octen-3-ol, dimethyl trisulfide, eugenol, ethyl hexanoate, and (E,E)-2,4-nonadienal (R. Q. He, Wan, Liu, & Chen, 2020; Inoue et al., 2016). On the other hand, as mentioned by Wang, Zhou, and Liu (2020), umami taste substances could also have enhanced the perceived saltiness. However, saltiness was not much affected by the amount of okara added in mixed media or variable drying techniques.

The cumulative contribution of variance from PC1 and PC2 exceeded 85% (Fig. 3B), and it demonstrates that this PCA chart could reflect most of the taste intensities of okara and wheat bran. A distinct separation in the main PC1 component can be observed between the raw materials and fermented samples, suggesting that SSF by P. ostreatus significantly varied e-tongue taste profiles. Most of the fermented media, except FO and OW0, clustered closely, reflecting a higher similarity in their e-tongue taste patterns. Overall, the outcomes of PCA revealed the development of most taste characteristics of okara and wheat bran after fermentation via P. ostreatus, except for aftertaste-astringency and sourness.

3.8. Principal component analysis and correlation assessment

PC1 and PC2 contributed cumulatively to more than 70% of the variance (Fig. 4A), indicating a representation of the majority of the taste substances of okara. A clear distinction was observed between the taste compound profiles of the raw materials and the fermented samples. Fermented okara, particularly FO, DE, and FD samples, was strongly associated with umami taste substances compared to fermented mixed media. The heat map with dendrogram displays differences in standardized concentrations of each taste substance across samples and hierarchical relationships (Fig. 4B). Compared to raw materials, fermented samples exhibited more taste components, particularly free amino acids, 5′-nucleotides, glucose, succinic, fumaric, malic, and tartaric acids, which could be attributable to the capability of P. ostreatus for biotransformation of lignocellulose biomass (M. He, Peng, Xu, Shi, & Qiao, 2024; Yan et al., 2025). Increases in concentrations of flavor components can also be observed in past studies on SSF of wheat bran and soy-based by-products (Ghamry et al., 2023; Keong et al., 2023; Wang et al., 2023; H. M. Zhao et al., 2017). In comparison, unfermented raw materials had high contents of certain sugars (fructose, sucrose) and citric acid, which could have been consumed during mycelial growth and metabolic activities.

Fig. 4.

Multivariate analysis of taste characteristics of raw materials, fermented okara, and fermented okara-derived substrates. (A) Principal component analysis of non-volatile taste substances. (B) Hierarchical clustering heatmap of taste substances. The color from red to blue represents a positive to negative z-score. The scale indicates the standardized scores of concentrations of each taste substance. (C) Pearson correlation coefficient matrix between electronic tongue taste attributes and concentrations of non-volatile taste substances. The color from red to blue represents positive to negative correlations. Asterisk (*) indicates statistically significant correlations at p < 0.05. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The heat map (Fig. 4C) illustrates correlations between taste properties assessed via the electronic tongue and taste compounds of raw materials and fermented samples. Glutamic acid and flavor nucleotides had weak positive relationships with umami taste and richness, while aspartic acid was negatively correlated with umami taste and richness. In the studies by Ismail, Hwang, and Joo (2020) and Xia et al. (2021), negative associations between umami taste by electronic tongue and umami taste-free amino acids were also observed. In the present study, succinic acid, a known umami taste contributor, indicated a moderate degree of positive association with richness and umami taste. The complexity of the overall taste profile, particularly due to synergistic and inhibitory interactions, could explain these deviations between present results and the anticipated stronger correlations between taste substances and their respective taste attributes (Samarasiri & Chen, 2022; Shangguan et al., 2024; Zhang et al., 2022). When analyzing mixtures of taste substances, oral physiological, cognitive, and chemical interactions that affect taste perception also need to be examined (Keast & Breslin, 2002). Apart from citric acid, all other organic acids had negative associations with sourness as previously observed by Samarasiri and Chen (2025a). Zhang et al. (2022) similarly reported an inconsistency between TAV analysis and sensory evaluation of tea, possibly due to interactions between organic acids and other compounds.

Most of the bitter taste-free amino acids showed positive correlations with both bitterness and aftertaste-bitterness. Strong or moderate positive correlations were recorded between astringency and some umami taste substances, including glutamic acid and flavor nucleotides. It is in line with the findings of the large-leaf yellow tea processed via various roasting methods (Sheng et al., 2024). That study also reported positive associations between some bitter taste components and umami taste, which were evident in the present study. Taste interactions causing promotion and inhibition can exist between various groups of taste compounds: sweet and umami substances; bitter and astringent substances; umami and astringent substances; astringent and sweet substances (Sheng et al., 2024). Based on electronic tongue sensory data, the synergistic effects between umami substances, including 5’-GMP/5’-IMP/Fru-Glu and MSG, were detected in the study by J. Zhang et al. (2023). Taken together, the taste characteristics of complex food matrices are influenced not only by individual taste compounds but also by their receptor-level interactions and matrix effects, underscoring the need for future studies integrating chemical and sensory evaluations. Moreover, since an electronic tongue may not fully represent human sensory perception in complex fermented food systems, future studies should include human sensory evaluation (following safety assessment) for a more comprehensive assessment of flavor quality and consumer acceptability.

4. Conclusions

This work showed the viability of Pleurotus ostreatus for solid-state fermentation of okara or mixtures containing okara and wheat bran or sawdust. Mycelium had a rapid growth and apparent density in fresh okara. Significant improvement in protein levels was achieved in fermented hot-air-dried okara. During the fermentation, all the substrates experienced a net substrate weight loss, possibly caused by fungal metabolic activities. This study investigated the concentrations of non-volatile taste compounds (soluble sugars, free amino acids, organic acids, and 5′-nucleotides), and their levels were significantly influenced by the medium composition, okara pre-treatment, and fermentation. Fermentation by P. ostreatus led the media to have higher contents of succinic acid, glucose, 5′-nucleotides, and total free amino acids, while resulting in a rise in umami taste substances. More umami taste amino acids existed in fermented fresh okara, whereas fermented hot air-dried okara had the maximum level of sweet taste amino acids. Furthermore, electronic tongue test results revealed that there was a reduction in sourness and an increase in umami taste and richness as a result of bioconversion. In conclusion, this work showed the potential of P. ostreatus for the biotransformation of okara and mixed media (okara + wheat bran) to acquire a potential food product with improved protein levels and umami taste. Further studies are warranted, particularly focusing on comprehensive metabolite profiling, elucidation of metabolic pathways, optimization of fermentation conditions, and evaluation of food safety and storage stability, including the assessment of potential contaminants, harmful metabolites, water activity, and microbiological stability, as well as characterization of volatile aroma profiles. Fermented mixtures of sawdust and okara may be further explored in future studies for the development of mycelium-based biocomposites, subject to evaluation of their structural and material properties.

CRediT authorship contribution statement

Malsha Samarasiri: Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation, Visualization. Cherie Chin: Writing – review & editing, Methodology, Investigation, Data curation. Nobuhisa Kawaguchi: Validation, Resources. Taisei Nakaminoto: Validation, Resources. Wei Ning Chen: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.