Yujia Zhang1
Dong Hu
Xiumin Zhang- 1College of Life Sciences, Hebei University, Baoding, Hebei, China
- 2Institute of Agro−Resources and Environment, Hebei Academy of Agriculture and Forestry Sciences, Shijiazhuang, Hebei, China
- 3Key Laboratory of Microbial Diversity Research and Application of Hebei Province, Baoding, Hebei, China
- 4Engineering Research Center of Microbial Breeding and Conservation, Baoding, Hebei, China
As the core protein source in livestock and poultry feed, soybean meal has its nutritional value limited by anti-nutritional factors (ANFs). This study innovatively combined the Resuscitation Promoting Factor (Rpf) of Micrococcus luteus with Bacillus velezensis and B. amyloliquefaciens in a solid-state fermentation. The aim was to activate functional microbial communities within the system, enabling synergistic degradation of ANFs and improve the nutritional quality of soybean meal. Results demonstrated that the Rpf-treated group significantly accelerated the degradation of glycinin, β-conglycinin, and trypsin inhibitor (the degradation efficiency within 48 hours was 1.28-2.0 times of the control group), while enhancing the enzyme activities of acid protease, amylase, and cellulase. Microbial diversity analysis revealed that Rpf promoted the proliferation of Bacillota by 18.91 percentage points and concurrently suppressed potential pathogenic bacteria, reducing their genus diversity from 8 to 3 and their cumulative relative abundance from 18.5% to 4.88%. Metabolomics profiling indicated that the Rpf-treated group upregulated key metabolites, including phospholipids, amino acids (e.g., serine and arginine), and vitamins. This was accompanied by enhanced glycerophospholipid metabolism and isoflavone biosynthesis pathways, which collectively improved the antioxidant and antibacterial properties of soybean meal. The study confirms that Rpf optimizes the fermentation system through the “functional bacterial resuscitation-Bacillus synergy” mechanism, providing a theoretical foundation and technical strategy for efficient soybean meal biotransformation and feed quality enhancement.
1 Introduction
Soybean Meal (SBM) is the core protein source of livestock and poultry feed and has an important strategic position in China’s livestock and poultry breeding industry. Its nutritional advantages are significant: the crude protein content of dry matter is 44%-50%, and the amino acid spectrum is balanced, especially rich in lysine (2.5%-2.8%) that is easily lacking in grain feed (Karr-Lilienthal et al., 2005; Kiarie et al., 2020), which can effectively make up for the amino acid deficiency of grains. However, the existence of a variety of ANFs in soybean meal seriously restricts its nutritional value. These ANFs, including glycinin, β-conglycinin, trypsin inhibitor (TI), saponin, phytoestrogen, soybean agglutinin, etc., significantly affect the digestion and absorption of nutrients by animals through complex biochemical mechanisms (Choct et al., 2010; Gilani et al., 2005; Liener, 1994; Lu et al., 2025; Medeiros et al., 2018). Glycinin and β-conglycinin are the main soybean antigenic proteins. The latter is composed of three subunits of α (68 kDa), α’ (72 kDa) and β (52 kDa) to form a heterotrimer (Gonzalez de Mejia et al., 2010; Sun et al., 2008; Zhao et al., 2008), accounting for 30% of the total soybean meal protein. Both can cause intestinal allergic immune responses, leading to intestinal mucosal damage, inflammation and diarrhea (Chen et al., 2011; Han et al., 2010; Li et al., 1990). Trypsin Inhibitor (TI) specifically binds to the catalytic site of trypsin through the active center to form an inactive enzyme-inhibitor complex, inhibiting the digestion and absorption of protein in the intestine (Mukherjee et al., 2016). Long-term feeding of feed containing TI can lead to compensatory hypertrophy of the pancreas and amino acid metabolism disorders (Chi and Cho, 2016). Other ANFs, including saponin (interfering with lipid metabolism), soybean agglutinin (destroying the balance of intestinal flora), phytoestrogen (interfering with the endocrine system), etc (Choct et al., 2010; Liener, 1994; Medeiros et al., 2018), synergistically aggravate nutritional and metabolic disorders.
Conventional methods for ANF reduction, such as thermal or enzymatic processing, often incur high costs, risk nutrient damage, or leave residual components (Enneking and Wink, 2000). While solid-state fermentation (SSF) with Bacillus strains presents a promising alternative, its efficacy largely depends on the metabolic potential of the inoculated strains, frequently overlooking the functional capacity of the indigenous microbiota within the soybean meal substrate.
In order to break through this technical bottleneck, solid-state fermentation (SSF), as an economical and efficient biotechnology, has been widely applied (Dai et al., 2022; Wang et al., 2021b; Yao et al., 2021). Fermented soybean meal, a product obtained from traditional SBM through microbial fermentation, has richer nutritional components compared to ordinary SBM and has become a more commonly used high-quality plant protein source in modern animal husbandry. SSF utilizes the metabolic activity of microorganisms in a solid-state matrix to break down large molecules in SBM into smaller molecules that are easier to absorb, while effectively degrading various ANFs, which enhances the nutritional quality of the feed (Dai et al., 2022; Rigo et al., 2010; Yao et al., 2021). During the fermentation process, the protease system secreted by microorganisms degrades soybean meal protein into polypeptides and oligopeptides, increasing the protein content, having a more balanced amino acid composition, and producing abundant organic acids, B-complex vitamins, and unknown growth factors and other bioactive substances. These changes not only improve the bioavailability of nutrients but also enhance the unique probiotic function of fermented SBM.
Resuscitation-promoting factor (Rpf) is a kind of key secreted protein that regulates bacterial dormancy and resuscitation. Its function was first discovered in 1998 in the fermentation supernatant of Micrococcus luteus (Mukamolova et al., 1998). Mukamolova et al. isolated a protein fraction (Rpf) with an apparent molecular weight of 16–17 kDa via chromatography and used this protein to treat dormant M. luteus cells that had been washed with a low-nutrient medium (LMM). They found that the protein could activate the dormant cells at picomolar (pM) concentrations, induce their resuscitation, and promote proliferation (Mukamolova et al., 1999, 2002). Studies have shown that Rpf can selectively enrich functional flora, and activate the metabolic activity of degrading bacteria for organic pollutants such as polychlorinated biphenyls, anthraquinone dyes, polycyclic aromatic hydrocarbons, and phenols, significantly improving their degradation efficiency (Su et al., 2021, 2018a, 2018c; Ye et al., 2020). Adding the culture supernatant of M. luteus containing natural Rpf can significantly increase the abundance of culturable bacteria in environmental samples (such as water samples), promote the isolation of new species, and enhance the activity of functional strains (Liu et al., 2016; Lopez Marin et al., 2021; Su et al., 2013, 2015a, 2015b; Tahmasbizadeh et al., 2025; Wang et al., 2021b). Rpf can also drive the community succession of denitrifying bacteria in sediments and enhance the denitrification capacity of the system (Su et al., 2019). Notably, these documented applications of Rpf are predominantly in environmental bioremediation, with no prior reports on its use in the solid-state fermentation of food or feed substrates. In addition, studies have confirmed that Rpf can activate the activity of cellulose-degrading bacteria (such as Bacillus) in the composting system, improve the expression level of cellulase, and accelerate the transformation of lignocellulose (Su et al., 2018b).
Therefore, this study focuses on the problem of limited bioavailability of nutrients and host metabolic disorders caused by ANFs in SBM. It innovatively introduces the Rpf of M. luteus as an activator of functional flora. Through the dual strategies of “artificial inoculation + directional resuscitation of functional bacteria”, a multi-strain cooperative fermentation system is constructed. Specifically, Rpf is used to selectively activate the indigenous dormant functional flora in SBM (such as cellulose-degrading bacteria and ANF-decomposing bacteria), and form a functional complement with artificially inoculated B. velezensis and B. amyloliquefaciens, which synergistically degrade ANFs such as glycinin and β-conglycinin, and at the same time promote macromolecular proteins to be converted into small peptides and free amino acids. By integrating microbial diversity analysis and non-targeted metabolomics technology, the regulation mechanism of the ternary interaction of “Rpf-Bacillus-functional flora” on ANF degradation and metabolite synthesis is mainly analyzed. This strategy breaks through the technical bottleneck of insufficient functional development of indigenous flora in traditional fermentation systems and provides new ideas for improving the nutritional quality of SBM. Based on this, we hypothesized that Rpf orchestrates a synergistic cooperation between the resuscitated indigenous functional flora and the inoculated Bacillus consortia, which drives the efficient degradation of ANFs and promotes the generation of beneficial metabolites.
2 Materials and methods
2.1 Strains and reagents
Micrococcus luteus (CGMCC1.2299, China General Microbiological Culture Collection Center (CGMCC), Beijing, China), B. amyloliquefaciens YB-1 and B. velezensis 9–1 were isolated from commercially available feed additives. Soybean meal (SBM) and Corn flour were purchased from Sinograin Zhenjiang Grain and Oil Co., Ltd. (Zhenjiang, China).
2.2 Seed culture medium and culture
B. velezensis 9–1 and B. amyloliquefaciens YB-1 were inoculated into 10 mL of NA liquid medium and cultured at 37 °C for 18 hours as the seed liquid. Then, 2% of the inoculum was transferred into the enzyme-producing fermentation medium (yeast extract 1.5 g/L, glucose 10 g/L, peptone 5 g/L, CaCO3–6 g/L, MgSO4·7H2O 0.05g/L, NaH2PO4·2H2O 2 g/L, Na2HPO4·2H2O 4 g/L) and cultured at 30 °C for 48 hours.
M. luteus CGMCC1.2299 was inoculated into YPG liquid medium and cultured at 28 °C for 24 hours as the seed liquid. Then, 5% of the inoculum was transferred into LMM medium (NH4Cl 4 g/L, KH2PO4 1.4 g/L, biotin 0.005 g/L, L-methionine 0.02 g/L, thiamine 0.04 g, inosine 1 g, MgSO4 0.07 g, CuSO4 0.000024 g, MnCl2 0.0005 g, FeSO4 0.001 g, NaMoO4 0.000025 g, ZnSO4 0.00005 g, L-lithium lactate 10 g) and cultured at 30 °C for 48 hours. The culture solution was centrifuged at 5000 rpm for 10 min, and the supernatant containing Rpf protein was collected and filtered for sterilization.
2.3 Solid-state fermentation of SBM
SBM and corn flour (4:1) were used as the dry feed substrate. 1% of B. velezensis 9–1 and B. amyloliquefaciens YB - 1 were added, with the ratio of the two strains being 1:1, and the material-to-water ratio was 1:0.5. 2% ammonium sulfate (not sterilized) was added as a nitrogen source. After mixing evenly, the mixture was packed into plastic sealed small bottles (each bottle with a capacity of about 0.4 kg). To minimize oxygen exposure and create a low-oxygen environment for anaerobic phases, the substrate was tightly compacted to reduce air pockets, and the bottles were filled nearly to the top before the caps were securely sealed. The fermentation was then carried out at 28 °C for 7 days. In the treatment group (T), 50% of the water was replaced by the fermentation supernatant containing Rpf protein of M. luteus, while the control group (C) did not add the fermentation supernatant of M. luteus.
Each treatment (C and T) included three independent biological replicates (separate fermentation bottles). Samples were collected from each replicate at designated time points (0, 12, 24, 48, 72, 96, 120, 144, and 168 h). Each sample was subdivided for immediate microbiological analysis, physicochemical analysis, and storage at -80 °C for omics analysis. Data for physicochemical indices represent the mean of the three biological replicates.
2.4 pH detection
Take 5 g of the above samples respectively, add 45 mL of distilled water, shake at 150 rpm for 1 hour, and then centrifuge at 5000 rpm at 4 °C for 10 min, and determine the pH value of fermentation supernatant using a pH meter (e.g., METTLER TOLEDO FE28).
2.5 Aerobic total colony counting
Fresh samples taken at 12 h, 24 h, 48 h, 72 h, 96 h, 120 h, 144 h and 168 h of fermentation were serially diluted in sterile water, spread on Plate Count Agar (PCA; Beijing biotopped Technology Co., Ltd), and incubated at 37 °C for 24 h for colony enumeration (detection limit: 10 CFU/g). Representative colonies of distinct morphologies were subsequently identified by 16S rRNA gene sequencing to correlate cultivable counts with key microbial groups.
2.6 Determination of antinutritional factors
0.3 g of fermented feed was weighed, mixed with 30 mL of extraction buffer (0.03 mol/L Tris-HCl, pH 8.0), and incubated at 25 °C for 16 h. Then the mixture was centrifuged at 4000 rpm for 10 min, and the supernatant was collected to determine glycinin and β-conglycinin. For trypsin inhibitor, use a different extraction solution (0.01 mol/L NaOH, pH 9.5 ± 0.1), allow it stand at 4 °C for 16–24 h, and then centrifuge the mixture at 4000 rpm for 10 min, and collect the supernatant. According to the manufacturer’s instructions, the concentration was determined using a commercial enzyme-linked immunosorbent assay (ELISA) kit (Shanghai Yuanju Biotechnology Co.; Lot No. 2025054151441A). The detection range and limit of detection (LOD) of the trypsin inhibitor kit were 1.0 - 30.0 ng/mL and 0.5 ng/mL, respectively. To account for potential matrix effects in fermented soybean meal, a spike recovery test was conducted, and the recovery rate was between 92% - 105%, which was acceptable. All sample extracts were appropriately diluted (usually 1.2-fold) to fall within the range of the standard curve.
2.7 Determination of enzyme production capacity
Enzymatic activities were determined according to Chinese national standards under specified conditions: neutral protease (GB/T 28715-2012; pH 7.5, 40 °C; U = 1 μg tyrosine/min from casein), acid protease (reference method; pH 3.0, 40 °C; U = 1 μg tyrosine/min from hemoglobin), amylase (NY/T 912-2020; pH 6.9, 60 °C; U = 1 mg starch hydrolyzed/min), cellulase (NY/T 911-2020; pH 4.8, 50 °C; U = 1 μmol reducing sugar/min from CMC-Na), and glucanase (reference method; pH 5.5, 50 °C; U = 1 μmol reducing sugar/min from barley β-glucan). All activities are expressed as U per gram of dry matter (U/g DW).
2.8 Microbial diversity analysis
For the unfermented SBM (control, labeled K) and the SBM samples fermented for 24 h, 48 h, 72 h, 96 h, 144 h, and 168 h, total microbial genomic DNA was extracted using the E.Z.N.A.® soil DNA kit (Omega Bio-tek, Norcross, GA, U.S.). The V3-V4 hypervariable region of the bacterial 16S rRNA gene was amplified by PCR using barcoded primers 338F (5’-ACTCCTACGGGAGGCAGCAG-3’) and 806R (5’-GGACTACHVGGGTWTCTAAT-3’) (Liu et al., 2016). Purified amplicons were pooled in equimolar amounts and paired-end sequenced on an Illumina Nextseq2000 platform (Illumina, San Diego, USA) according to the standard protocols by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). The raw sequencing reads were deposited into the NCBI Sequence Read Archive (SRA) database (Accession Number: PRJNA1204194).
Raw paired-end reads were primarily processed using the QIIME2 platform (version 2024.5). Briefly, sequence quality control, denoising, merging, and chimera removal were performed using the DADA2 plugin within QIIME2, which infers amplicon sequence variants (ASVs). This process resulted in a feature table of ASVs and their frequencies. Taxonomic assignment of ASVs was conducted using a pre-trained Naive Bayes classifier based on the SILVA 138.1 database (99% OTU full-length sequences), with a confidence threshold of 80%. All ASVs classified as Chloroplast, Mitochondria, or unassigned at the Kingdom level (Archaea, Eukaryota) were filtered out from the feature table prior to downstream analysis. The final ASV table, taxonomy information, and sample metadata were used for all subsequent analyses.
Raw sequencing data were subjected to quality control using the fastp (v0.19.6) (Chen et al., 2018), followed by read merging with FLASH (v1.2.11) (Magoč and Salzberg, 2011). UPARSE (v7.1) (Edgar, 2013; Stackebrandt and Goebel, 1994) clustered OTUs at 97% sequence similarity and to removed chimeric sequences. Taxonomic annotation of OTUs was performed using the RDP classifier (v2.11) (Wang et al., 2007) against the SILVA 16S rRNA gene database (v138) with a 70% confidence threshold. Microbial phenotype prediction was conducted through Bugbase software (v1.0) after OTU table normalization.
The mothur (Schloss et al., 2009) software (v1.30.1) was used to calculate alpha diversity indices (e.g., Chao 1, Shannon index), with group differences analyzed via the Wilcoxon rank- sum test. Microbial community similarity was assessed through principal coordinate analysis (PCoA) based on the Bray-Curtis distance, paired with PERMANOVA testing(999 permutations, P<0.05). The linear discriminant analysis (LDA) effect size (LEfSe) (http://huttenhower.sph.harvard.edu/LEfSe) (Segata et al., 2011) identified differentially abundant taxa (LDA score >2, P<0.05), while a correlation between two nodes was considered to be statistically robust if the spearman’s correlation coefficient over 0.6 or less than -0.6, and the P-value less than 0.01.
Regarding functional prediction, we acknowledge the limitations of phenotypic prediction tools like BugBase. Therefore, we have exercised caution in our interpretation and focused our discussion primarily on the observed taxonomic shifts and their potential metabolic implications, as supported by our concurrent metabolomics data. We have removed the standalone BugBase analysis and note that future studies could employ more robust methods like PICRUSt2 or metagenomic sequencing for deeper functional insights.
2.9 Non-targeted metabolomics analysis
Take 100 mg of unfermented SBM sample (K) and SBM samples fermented for 48 h, 96 h, and 144 h. Add grinding beads (6 mm in diameter) and 800 μL of extraction solution (methanol:water = 4:1, containing a mixture of four internal standards: L-2-chlorophenylalanine (as a chemical internal standard) and three proprietary stable isotope-labeled compounds (covering a range of physicochemical properties)), then grind for 6 min (-10 °C, 50 Hz). Perform low-temperature ultrasonic extraction for 30 minutes (5 °C, 40 kHz). After allowing the sample to stand at -20 °C for 30 min, centrifuge it for 15 min (4 °C, 13000 g). Collect the supernatant for LC-MS analysis. A quality control (QC) sample was prepared by combining equal aliquots of all experimental samples. The QC sample was injected at the beginning of the analytical sequence for column conditioning and then repeatedly after every 10 experimental samples throughout the entire run to monitor the stability and reproducibility of the analytical system.
Chromatographic conditions: A 3 μL aliquot of the sample was separated using a BEH C18 column (100 mm × 2.1 mm, 1.7 µm). Mobile phase A consisted of 2% acetonitrile water (containing 0.1% formic acid), and mobile phase B was acetonitrile (containing 0.1% formic acid). The flow rate was 0.40 mL/min, and the column temperature was maintained at 40 °C.
Mass spectrometry conditions: Analysis is performed in both positive and negative ion scanning mode. The mass scan range was m/z 70-1050. The ion spray voltage was set to +3500 V (positive ion mode) and -3000 V (negative ion mode). Sheath gas and auxiliary heating gas pressures were 50 psi 13 psi, respectively. The ion source temperature was 450 °C, and collision energy followed a stepped gradient of 20-40–60 V per cycle. MS1 and MS2 resolution were 70,000 and 17,500, respectively.
The raw data were uploaded to the Metabolights database (MTBLS12408) and imported into Progenesis QI software (Waters Corporation, Milford, USA). The data preprocessing included peak alignment, peak picking, and normalization. Signal drift correction was performed using the LOESS (Locally Estimated Scatterplot Smoothing) algorithm based on the QC samples. Metabolite features with a relative standard deviation (RSD) > 30% in the QC samples were removed to ensure data quality.
Metabolite identification was performed by matching against the MJDBPM database. The confidence of identification was assigned according to the Metabolomics Standards Initiative (MSI) levels: Level 1 (confirmed by reference standard); Level 2 (putatively annotated based on MS/MS spectral library match); Level 3 (putatively characterized based on exact mass and isotopic pattern). The differentially abundant metabolites reported in this study were primarily at Levels 2 and 3.
Subsequently, the data matrix was uploaded to the Majorbio Cloud platform for preprocessing. Multivariate statistical analyses (PCA and OPLS-DA) were conducted using the R package “ropls” (Version 1.6.2), and model stability was evaluated through 7-cycle interactive validation. Significantly different metabolites were screened based on VIP > 1 and P < 0.05. Significantly different metabolites were screened based on VIP > 1 from the OPLS-DA model and p-value < 0.05 from Student’s t-test, followed by false discovery rate (FDR) correction using the Benjamini-Hochberg method. Metabolites with VIP > 1 and FDR < 0.05 were considered statistically significant. Pathway annotation of these metabolites was performed using the KEGG database (http://www.genome.jp/kegg/), and pathway enrichment analysis was carried out with the Python software scipy.stats package (https://docs.scipy.org/doc/scipy/). Biological pathways associated with the experimental treatment were identified via by Fisher’s exact test.
3 Results
3.1 Changes in pH value and culturable bacterial count
The initial pH value of SBM before fermentation was 6.60 - 6.70. As fermentation progressed, the pH value of C in the natural fermentation state gradually decreased, while the pH value of T with Rpf protein added is relatively higher. At 96 h of fermentation, the pH value of T dropped to 6.37, which was significantly higher than that of C (5.87, with a difference of 0.5) (Figure 1A).
Figure 1. Changes in pH and the number of culturable bacteria during fermentation. (A) Changing trend of pH value; (B) Changing trend of culturable bacteria (p ≤ 0.05). “a” and “b” in the Fig. indicate significant differences between the control group and the treatment group at the same time (p< 0.05); “a”, “a” indicates that it is not significant.
Figure 1B shows the dynamic changes in the number of culturable microorganisms of SBM during the natural fermentation process. In the initial stage of fermentation, the total number of bacteria steadily increased within 96 h and then tends to stabilize. the total aerobic bacterial count in T supplemented with M. luteus fermentation supernatant was significantly higher than in C during the initial fermentation stage. After 96 h, the log CFU value remains at around 4.5.
3.2 Anti-nutritional factors
To verify whether the fermentation supernatant of M. luteus enhances the recovery of functional strains in SBM and effectively eliminates anti-nutritional factors to improve SBM nutritional value, we measured the contents of glycinin, β-conglycinin, and trypsin inhibitor in C and T throughout fermentation. As shown in Table 1, within 48 hours of fermentation, the elimination effect of the T group on the three anti-nutritional factors was significantly better than that of the C group (Figure 2). Specifically, at 48 hours, the contents of glycinin, β-conglycinin, and trypsin inhibitor in the T group decreased to 12.41 μg/g, 7.04 μg/g, and 8.81*10-6 μg/g, respectively. The degradation amounts reached 16.73 μg/g, 2.74 μg/mL, and 9.77*10-6 μg/g, respectively, and the degradation rates were 57.41% (28.72% in the C group), 27.98% (21.78% in the C group), and 52.57% (27.98% in the C group), which were 2.0 times, 1.28 times, and 1.88 times those of the C group, respectively.
Table 1. Statistics on the changes in the content of different anti-nutritional factors.
Figure 2. Changes in anti-nutritional factors during fermentation. (A) Glycinin; (B) β-conglycinin;(C) soybean trypsin inhibitor (p ≤ 0.05). “a” and “b” in the figure indicate significant differences between the control group and the treatment group at the same time (p< 0.05); “a”, “a” indicates that it is not significant.
3.3 Enzyme-producing ability
Protease generate amino acids by degrading polypeptide chains (Cheng et al., 2019; Mótyán et al., 2013), and their activity patterns show significant differences between T and C. The activity of neutral protease (Figure 3A) increased gradually during 0–48 h, but between 72–96 h, the neutral protease activity in T was lower than that in C. Both groups reached peak neutral protease activities at 96 h (999 U/g) and 144 h (1024 U/g) respectively, followed by a decline. For acid protease (Figure 3B), T showed lower activity than C in the early fermentation stage (within 24 h). However, between 72–120 h, the acid protease activity in T was significantly higher, peaking at 144 h (729 U/g). Amylase (Ali et al., 2025) and cellulase decompose SBM by hydrolyzing carbohydrates to release crude protein, which is further utilized by proteases to generate water-soluble protein (Salim et al., 2019; Wu et al., 2019). The amylase activity (Figure 3C) increased continuously throughout fermentation, with slow growth within 72 h. After 72 h, amylase activity in T was significantly higher than in C. Cellulase activity (Figure 3E) was detectable only in C within the first 24 h, peaking at 12 h (2.06 U/g). In contrast, T maintained cellulase activity, peaking at 24 h (2.47 U/g), decreasing at 48 h, and then rising again. Similarly, glucanase activity (Figure 3D) was detected in C only within the first 12 h of fermentation, while T exhibited sustained activity until 72 h, peaking at 24 h (5.67 U/g).
Figure 3. Changes in different enzyme activities during fermentation. (A) neutral protease ;(B) acid protease; (C) amylase; (D) Cellulase; (E) Glucanase (p ≤ 0.05). “a” and “b” in the figure indicate significant differences between the control group and the treatment group at the same time (p< 0.05); “a”, “a” indicates that it is not significant; “*” indicates that there was a significant difference between the control group and the treatment group at the same time (p< 0.05).
3.4 Dynamic changes of bacterial communities during the fermentation process
Figures 4A–C illustrate the changes in bacterial community composition abundance during SSF. In unfermented soybean meal (K), the relative abundances of bacterial phyla were as follows: Cyanobacteriota (60.41%), Pseudomonadota (21.15%), Bacillota (9.97%), and Actinobacteriota (8.21%), with the remaining phyla distributed uniformly. As fermentation progressed, significant shifts occurred in both bacterial community structure and abundance. Notably, the abundance of Bacillota increased sharply. In C (Figure 4A), Bacillota abundance reached 68.95% within 24 h, whereas in T supplemented with M. luteus fermentation supernatant (Figure 4B), its abundance rose to 87.86% during the same period—an increase of 18.91 percentage points compared to C. Additionally, T achieved a stable community structure 24 h earlier than C, with a simpler phylum-level composition dominated by Bacillota, Cyanobacteriota, and Pseudomonadota, exhibiting minimal abundance fluctuations.
Figure 4. Community structure changes during fermentation. (A) Community structure of C at the phylum level; (B) community structure of T at the phylum level; (C) community structure at the genus level between C and T; (D) Gram-positive; (E) Gram-negative; (F) Facultative anaerobic.
At the genus level (Figure 4C), unfermented SBM was primarily colonized by norank_o:Chloroplast (59.72%), norank_f:Mitochondria (19.86%), Carnobacterium (3.85%), Bacillus (3.78%), Corynebacterium (2.89%), and Paeniglutamicibacter (2.30%). Following fermentation, the community shifted dramatically, with dominance by Staphylococcus, Bacillus, Enterococcus, Pediococcus, and Weissella. Specifically, Staphylococcus and Bacillus proliferated rapidly during early fermentation, with relative abundances in T exceeding those in C by 5.34-38.38%, aligning with colony count results. Conversely, Enterococcus abundance in T was 2.05-15.75% lower than C during early fermentation (before 96 h) but surpassed C by 13.43-25.99% in later stages (after 96 h). Meanwhile, Pediococcus and Weissella abundances remained consistently lower in the T throughout fermentation. In the later stages of fermentation (after 96 h), Staphylococcus and Bacillus enter a decline phase. In C, Pediococcus emerges as the dominant anaerobic microflora. In contrast, T exhibits dominance of Enterococcus (Liu et al., 2023).
Finally, BugBase phenotypic prediction analysis of Gram-negative bacteria, Gram-positive bacteria, and facultative anaerobes in SBM (Figure 4) revealed that after SSF, T exhibited lower abundance of Gram-negative bacteria compared to C (Figure 4D). In contrast, Gram-positive bacteria including Staphylococcus, Bacillus, and Enterococcus showed significantly higher abundance in T (Figure 4E). Additionally, most bacterial taxa in T were facultatively anaerobic (Figure 4F).
3.5 Pathogenicity prediction
Alpha diversity (α-diversity) is used to study microbial diversity in the environment, reflecting the richness and evenness of microbial communities through single-sample analysis. The results showed that α-diversity was ranked as K (unfermented soybean meal) > C (control group) > T (treatment group) (Supplementary Figure 1A), indicating a gradual decrease in microbial diversity and stabilization of community structure during SSF of SBM (Wang et al., 2021a). After adding the fermentation supernatant of M. luteus, T exhibited lower community diversity than C but a more stable structure. Additionally, Supplementary Figure 1B revealed that both C and T detected new bacterial genera post-fermentation compared to unfermented samples, with 35 unique genera in C and 26 in T. However, the abundance of these unique genera was low, and no significant quantitative differences were observed.
Further analysis of the pathogenicity of bacteria in SSF process was conducted, and the results are shown in Supplementary Table 1. Eight potential pathogenic bacteria were detected in C, including Arcobacter (g:Arcobacter, 3.70%), Vibrio (g:Vibrio, 3.70%), Tsukamurella (g:Tsukamurella, 2.47%), Pantoea (g:Pantoea, 2.47%), Providencia (g:Providencia, 2.47%), Tissierella (g:Tissierella, 1.23%), Escherichia-Shigella (g:Escherichia-Shigella, 1.23%), and Sodalis (g:Sodalis, 1.23%). In contrast, only three potential pathogenic bacteria were detected in T, including Myroides (g:Myroides, 2.44%), Facklamia (g:Facklamia, 1.22%), and Comamonas (g:Comamonas, 1.22%). It can be seen that the types (3) and abundance (4.88%) of potential pathogenic bacteria in T are lower than those in C (8, 18.5%).
Through phenotypic prediction analysis using BugBase, the pathogenicity of microorganisms was evaluated. The results in Supplementary Table 2 show that in T, the predicted pathogenicity of Staphylococcus (g:Staphylococcus) and Mammaliicoccus (g:Mammaliicoccus) was relatively high. However, existing studies indicate that Staphylococcus species can be safely utilized in fermented foods (Li et al., 2023), while Mammaliicoccus is recognized for its antibacterial properties (Lienen et al., 2022), with limited evidence confirming its pathogenicity. For other bacterial genera in T, phenotypic pathogenicity was significantly reduced compared to C, and C itself showed lower pathogenicity than unfermented samples.
3.6 Changes in metabolite species diversity based on untargeted metabolomics
To evaluate the metabolomic changes during the SSF process of SBM, we employed a non-targeted metabolomics approach for analysis. In all samples, a total of 1977 metabolites were detected. Through chemical classification annotation, Figure 5 demonstrates the quantity and diversity of metabolites detected during the SSF process of SBM. Lipid metabolites constituted the largest proportion (51.19%), predominantly comprising fatty acyls and glycerophospholipids. Amino acid metabolites represented the second major category (27.80%), with α-amino acids and their derivatives along with dipeptides being the principal components. Carbohydrates and derivatives accounted for 15.25%, mainly consisting of O-glycosyl compounds and glycosides. The remaining fractions included nucleotides and derivatives (4.41%) and vitamin metabolites (1.36%), which showed the lowest abundance.
Figure 5. Pie chart of fermented soybean meal primary taxonomic content.
Principal Component Analysis (PCA) effectively captured intra- and inter-group metabolite variations. The cumulative variance contributions of PC1 and PC2 reached 51.4% and 19.3%, respectively, demonstrating significant metabolic divergence among samples (Figure 6A). At 48 h of fermentation (Figure 6B), metabolite profiles between the control (C48) and treatment (T48) groups showed minimal differences, with phospholipid classes exhibiting the most pronounced changes: phosphatidylserine was elevated, whereas phosphatidylcholine was reduced. By 96 h (Figure 6C), the metabolic disparity between C96 and T96 groups peaked, primarily driven by alterations in phospholipids and amino acids. Notably, serine and arginine levels increased, while lysine, γ-aminobutyric acid, and asparagine decreased. At the terminal phase (144 h, Figure 6D), metabolite differences between C and T were no longer evident.
Figure 6. Principal component analysis diagram and KEGG compounds were classified and summarized at different times. (A) Principal component analysis diagram; (B) Classification and summary of KEGG compounds for 48 h; (C) Classification and summary of KEGG compounds for 96 h; (D) Classification and summary of KEGG compounds for 144 h.
Differential metabolites were identified and temporally classified during fermentation by integrating Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) with metabolomic feature selection criteria: Variable Importance in Projection (VIP) scores, fold change (FC), and univariate analysis p-values (statistical significance threshold). The results are systematically presented in Table 2. At 48 h of fermentation, upregulated metabolites of T were predominantly terpenoids, flavonoids, and coumarins (including derivatives), whereas downregulated metabolites encompassed major categories such as lipids, amino acids (and derivatives), organic acids, carbohydrates, and phenolic acids. At 96 h of fermentation, upregulated metabolites of T were predominantly flavonoids, terpenoids, carbohydrates (and derivatives), nucleotides (and derivatives), alkaloids (and derivatives), vitamins, and coumarins (and derivatives). Downregulated metabolites primarily comprised lipids, amino acids (and derivatives), organic acids (and derivatives), phenolic acids (and derivatives), and steroids (and derivatives). At 144 h of fermentation, most metabolites of T exhibited a declining trend, with phenolic acids (and derivatives) being the sole category showing significant accumulation. The volcano plot of differential metabolites at 144 hours of fermentation is shown in Supplementary Figure 3.
Table 2. Classification of metabolites at different time differences.
3.7 Analysis of key metabolic pathways at different times
Through KEGG pathway enrichment analysis of metabolites at different time points (Figure 7) between T and C, we observed primary activity in glycerophospholipid metabolism, glycine/serine/threonine metabolism, and isoflavone biosynthesis before 96 h (Figures 7A, B). Metabolic profiling at 96 h (Figure 7C) revealed an accumulation of zeatin-related metabolites with concomitant changes associated with lipid anabolism, notably affecting fatty acid and phospholipid biosynthesis.
Figure 7. KEGG pathway enrichment bubble diagram in the control group and treatment group at different times. (A) 48 h; (B) 96 h; (C) 144 h. Bubble diagram showing metabolic pathways that were significantly enriched in the treatment group compared to the control at different fermentation time points. The enrichment analysis was performed using a Fisher’s exact test, and the significance of enrichment was determined by the FDR-adjusted q-value (< 0.05). The size of the bubble represents the rich factor, and the color represents the -log10(q-value). The analysis was conducted against the KEGG database as the background reference set.
3.8 Combined analysis of microorganisms and untargeted metabolism
The metabolic data were dimensionally reduced using HCLUST (hierarchical cluster analysis), and all metabolites were grouped into distinct clusters. A correlation analysis was then performed between these clusters and microbial flora. The results (Figure 8) revealed that Enterococcus, Loigolactobacillus, Weissella, Latilactobacillus, Pediococcus, and unclassified_f_Lactobacillaceae exhibit positive correlations with metabolic clusters cluster_5 and cluster_8. These metabolic clusters contain metabolites related to fat metabolism, such as choline, 1-palmitoyl-sn-glycero-3-phosphocholine, Fa(18:1 + 2O), and Pip(18:1(9Z)/18:1(11Z)); metabolites related to isoflavone metabolism, such as daidzein, genistein, and genistein 7-O-glucoside; and metabolites related to glucose metabolism, such as D-Gal alpha 1->6D-Gal alpha 1->6D-glucose.
Figure 8. HCLUST correlation analysis.
On the other hand, Staphylococcus, unclassified_f_Staphylococcaceae, Bacillus, and Mammaliicoccus are positively correlated with metabolic clusters cluster_1, cluster_2, cluster_4, cluster_7, and cluster_10. These metabolic clusters contain metabolites related to amino acids and their derivatives, such as L-histidine, L-proline, D-pyroglutamic acid, and DL-o-tyrosine; metabolites related to vitamins, such as nicotinic acid (Vitamin B3) and 5’-O-beta-D-glucosylpyridoxine (Vitamin B6); and metabolites related to nucleotides and their derivatives, such as uracil.
By calculating the Pearson correlation coefficient, the correlations between metabolites and species within the same sample group were quantified and analyzed. The results are shown in Figure 9. In the Bacillota phylum, genera such as Lactiplantibacillus, Tetragenococcus, Enterococcus, Loigolactobacillus, unclassified_o_Lactobacillales, Latilactobacillus, unclassified_f_Lactobacillaceae, Pediococcus, Leuconostoc, Lactococcus, Weissella, and Ligilactobacillus show positive correlations with linoleic acid metabolism, anthraquinone compounds, and flavone/isoflavone metabolism. In contrast, Staphylococcus and Bacillus are positively correlated with fatty alcohol metabolism. Lipid metabolites predominantly include stearidonic acid, 9(S)-HODE, and oleoylglycerone phosphate. Anthraquinone metabolites are represented by emodin and obtusifolin, while flavonoid metabolites consist of daidzein, genistein, and 7,4-dihydroxyflavone.
Figure 9. Bacteria-metabolite correlation heat map.
4 Discussion
Before fermentation, the initial pH of SBM was 6.60–6.70, indicating a weakly acidic nature. This initial pH difference might be attributed to the protease activity of the YB-1 and 9–1 strains. Studies have shown that Bacillus species generate amines and ammonia through proteolysis during SSF (Nualkul et al., 2022), which could explain the pH increase in T. In the early fermentation stage, Bacillus proliferated rapidly by utilizing nutrients, oxygen, and moisture in the SBM substrate (Philavong et al., 2017). After 96 hours of fermentation, oxygen depletion suppressed the growth of aerobic bacteria such as Bacillus, marking the transition to the anaerobic phase.
Notably, T supplemented with M. luteus fermentation supernatant, exhibited a significantly higher total aerobic bacterial count than C during the initial fermentation stage, along with a more pronounced reduction in ANFs. This suggests that the supernatant may stimulate the short-term proliferation of aerobic bacteria. Colony morphology analysis revealed that the increased aerobic bacteria primarily consisted of Bacillus (Figure 10). Under aerobic conditions, Bacillus degrades ANFs into peptides (Lu et al., 2022), thereby improving their elimination efficiency. By 96 hours, the log colony-forming units (CFU) stabilized at approximately 4.5, and the difference in ANF content between C and T diminished, indicating the onset of the anaerobic phase, during which pH declined. Although Bacillus populations decreased overall, T maintained significantly higher counts than C, implying that the supernatant might extend Bacillus survival. However, Bacillus was no longer the dominant microbiota in the SSF system.
Figure 10. Colonies isolated from samples at different times of fermentation. Note:1–4 were the dilution coating plates of the control group for 12 h (dilution of 10-3), 24 h (dilution of 10-4), 48 h (dilution of 10-5) and 72 h (dilution of 10-5). 5–8 were the dilution coating plates of the treatment group for 12 h (dilution of 10-3), 24 h (dilution of 10-4), 48 h (dilution of 10-5) and 72 h (dilution of 10-5). 9–12 were diluted coating plates of the control group at 96 h, 120 h, 144 h, and 168 h (all dilutions were 10-4). 13–16 were diluted coating plates (all with dilution levels of 10-4) for 96 h, 120 h, 144 h, and 168 h in the treatment group.
During SSF, enzyme activity dynamics are driven by fluctuating environmental conditions (e.g., pH and oxygen availability) and shifts in microbial composition. These factors work synergistically to enhance the nutritional value of fermented SBM. At 48 h, T exhibited higher acid protease activity but lower neutral protease activity compared to the C group. This divergence was pH-dependent: acid accumulation lowered the pH, creating favorable conditions for acid protease activity while suppressing neutral proteases. Consequently, T maintained elevated acid protease levels from 48 to 120 h. Amylase activity trends correlated with Bacillus population dynamics. T showed significantly increased Bacillus abundance and prolonged survival, potentially attributable to starch-degrading strains YB-1 and 9-1 (Wang et al., 2023, 2024). Furthermore, the M. luteus fermentation supernatant added to T might have activated functional amylolytic strains, consistent with the observed increase in amylase activity. Cellulase and glucanase break down complex carbohydrates into monosaccharides. The sustained activity of these enzymes in T was likely caused by the M. luteus supernatant stimulating enzyme-producing strains. However, as oxygen depletion shifted the microbiota toward anaerobes, glucanase activity decreased. Notably, certain anaerobes (e.g., Lactobacillus spp.) exhibit cellulolytic activity (Nurliana et al., 2022), potentially explaining the recovery in cellulase activity after 48 hours.
The observed shifts in microbial community composition suggest that the addition of M. luteus fermentation supernatant, containing Rpfs, likely stimulated Staphylococcus and Bacillus species to rapidly utilize residual oxygen and nutrients in the substrate. This accelerated proliferation enabled these genera to enter the logarithmic growth phase earlier, explaining their significantly higher abundance in T compared to C during initial fermentation. Supporting this finding, plate colony counts revealed a rapid increase in aerobic bacteria in T, with Bacillus population dynamics closely matching relative abundance data.
Previous studies have demonstrated the capacity of Bacillus to degrade ANFs including glycinin, β-conglycinin, and trypsin inhibitors (Lu et al., 2022). The rapid increase in Bacillus abundance likely contributed to the accelerated reduction of these compounds, establishing a functional link between microbial community structure and substrate biochemical properties.
During early fermentation (<96 hours), aerobic metabolism by Bacillus and Staphylococcus dominated the bacterial community in SBM. In later stages (>96 hours), microbial activity shifted toward anaerobic metabolism as oxygen depletion caused Staphylococcus and Bacillus populations to decline, while facultative anaerobes thrived. The supernatant of M. luteus culture containing Rpf protein promoted Enterococcus dominance in T, whereas C was dominated by the acid-producing Pediococcus, resulting in significant pH reduction. This microbial transition may enhance intestinal digestibility and nutrient absorption efficiency of the fermented SBM (Liu et al., 2023). This microbial transition may improve intestinal digestibility and nutrient absorption efficiency of the fermented SBM.
BugBase predictive analysis indicated that Rpf proteins preferentially resuscitate Gram-positive bacteria and enrich facultative anaerobes, thereby improving SSF efficiency in SBM. These findings align with previous studies showing that Rpfs from M. luteus fermentation supernatant selectively revive high-affinity Gram-positive bacteria (Mukamolova et al., 1998), providing a plausible explanation for the observed changes in bacterial community abundance.
This study, through microbial diversity analysis, found that the microbial diversity in the SSF of SBM showed a natural decreasing trend, which is consistent with the community stabilization pattern reported by Wang et al (Wang et al., 2021a). Notably, T with the addition of M. luteus fermentation supernatant containing Rpf exhibited more significant community simplification. This change in community structure may be attributed to the regulatory effect of Rpf on the microbial interaction network, promoting the targeted enrichment of specific functional strains.
In terms of pathogenicity, the study confirmed that the SSF process itself has the potential to inhibit pathogenic bacteria, and the addition of M. luteus fermentation supernatant further enhanced this effect through multiple synergistic mechanisms. We speculate that Rpf may act through the following pathways: first, selectively promoting the growth of beneficial bacteria and optimizing the community composition; second, activating the antibacterial functions of specific strains; and finally, reconstructing the microbial interaction network to form a microenvironment unfavorable for the survival of pathogenic bacteria. These findings provide a new theoretical basis for understanding the function of Rpf in fermentation systems.
Non-targeted metabolomics analysis of fermented SBM revealed dynamic changes in its key functional components, with phospholipids and amino acids showing the most significant differences. Phospholipids, as functional nutrients in animal feed, not only exhibit liver-protective effects but also improve feed conversion efficiency and promote nutrient digestion and absorption, thereby reducing production costs (Yin et al., 2021). Research shows that serine, as a key differential amino acid, plays a crucial role in regulating intestinal health and optimizing production performance by participating in physiological processes such as lipid metabolism, fatty acid synthesis, and muscle development, making it an important dietary supplement for enhancing feed nutritional value (Zhou et al., 2021).
In T, metabolites significantly upregulated within the first 96 hours primarily included coumarins, terpenoids, flavonoid glycosides, certain carbohydrates, and alkaloids. Coumarins and their derivatives (Menezes and Diederich, 2019) possess dual functions as flavoring agents and antibacterial compounds, while flavonoid glycosides (Ding and Yu, 2025) modulate intestinal immunity through plant-microbe interactions. Terpenoids (Wang et al., 2025), as active components of traditional Chinese medicine, exhibit diverse pharmacological properties and serve as important precursors for natural flavors and food additives. Carbohydrates (Tiwari et al., 2020), as the primary energy source, not only supply ATP but also contribute to tissue structure formation. Alkaloids (Matsuura et al., 2014) demonstrate broad-spectrum biological activities, including antitumor, antibacterial, and antiviral effects.
Upon prolonged fermentation (>96 hours), phenolic acids and their derivatives (Alvarado-Martinez et al., 2024) became the dominant upregulated components. These plant-derived bioactive compounds are renowned for their antioxidant, anti-inflammatory, and metabolic regulatory functions, holding great promise for improving animal health and feed nutritional value. This time-dependent metabolite profile provides a scientific basis for optimizing the functionality and application of fermented SBM.
To analyze differences in metabolite categories between C and T, we selected the top 20 metabolites with VIP values for differential analysis. The results are shown in Supplementary Figure 2. Nucleotides and their derivatives (Supplementary Figure 2D) and vitamin metabolites (Supplementary Figure 2E) showed significant increases during fermentation in T. Among these, Inosine (Mian et al., 2023) exhibits high cell membrane permeability and can promote hepatocyte repair, prevent fatty liver, and enhance enzymatic activity. Its decomposition product, inosinic acid (Wang et al., 2014), serves as a key indicator for evaluating umami flavor in poultry meat. Studies indicate that inosinic acid, as a food flavor enhancer, exhibits an umami intensity several-fold higher than that of monosodium glutamate (MSG). Uridine 5′-monophosphate acts as a mitochondrial ATP-dependent potassium channel activator, promoting choline synthesis and inducing intestinal epithelial cell apoptosis, which enhances intestinal development and reduces diarrhea (Uauy et al., 1990). Calcidiol regulates calcium-phosphorus metabolism, improves calcium homeostasis, increases bone mass, and enhances reproductive performance in multiparous animals (Weaver et al., 2024).
Among amino acids and their derivatives (Supplementary Figure 2A), metabolites including N-acetyl-L-glutamic acid, N-acetyltryptophan, S-(1-propenyl) cysteine sulfoxide, 5-aminolevulinic acid, and margaroylglycine were significantly upregulated in T. In contrast, 2-aminomuconic acid semialdehyde, Phe-Leu, Phe-Gly-Ile, Ile-Ala, and lycoperodic acid were downregulated. N-Acetyl-L-Glutamic Acid (Wang et al., 2020), margaroylglycine (Han and Thacker, 2011), and S-(1-Propenyl)-cysteine sulfoxide (Singh et al., 2015) have been reported to promote early intestinal function and morphological development in broilers, thereby improving feed efficiency; these compounds can also act as food additives or nutritional enhancers to improve flavor. Additionally, margaroylglycine supplementation enhances production performance and egg quality in laying hens. N-Acetyltryptophan (Malhotra et al., 2019) contributes to immune system enhancement, and 5-aminolevulinic acid (Mao et al., 2020), a precursor of vitamin B12, improves immune function and intestinal morphology.
In lipid metabolites (Supplementary Figure 2B), tetradecanedioic acid (Werner and Zibek, 2017) not only serves as a spice but also enhances flavor, while myristoleic acid (Nurul Huda et al., 2022) promotes the growth of beneficial bacteria, inhibits harmful bacterial proliferation, and maintains intestinal microbiota balance, leading to significant improvements in production performance. In carbohydrates and their derivatives (Supplementary Figure 2C), the increased glucose content suggests that after adding the M. luteus fermentation supernatant, Bacillus species can hydrolyze carbohydrate compounds more efficiently via glucanase and cellulase activity during SSF process. Consequently, the SSF of Bacillus with M. luteus supernatant supplementation significantly elevates metabolite diversity and concentration, thereby improving SBM flavor, enhancing intestinal digestion, and boosting nutritional value.
Through KEGG metabolic pathway analysis, it was found that the addition of Rpf-containing supernatant significantly activated phosphatidylglycerol metabolism, as well as tryptophan, serine, and threonine metabolic pathways (Xue et al., 2023). This enhancement improved immune system function, stress resistance, and disease resilience, while also strengthening intestinal barrier integrity and immune regulation, thereby promoting gut homeostasis. Furthermore, this treatment significantly stimulated isoflavone biosynthesis (Peirotén et al., 2019), particularly enhancing growth performance in male animals, reducing feed costs, and improving reproductive efficiency. Subsequent studies revealed that after 96 hours of treatment, the activation of the zeatin metabolic pathway, combined with enhanced vitamin production, synergistically improved intestinal nutrient absorption and alleviated gastrointestinal burden. Collectively, these metabolic adaptations comprehensively enhanced animal health and productivity, demonstrating the strong potential of this formulation as a novel feed additive.
Based on the results of the combined analysis, we speculate that certain bacterial flora such as Enterococcus may be involved in the metabolism of fat, isoflavones and glucose, while other flora such as Staphylococcus may participate in the metabolism of amino acids, vitamins and nucleotides. Notably, the upregulation of alkaline ammonium salts may account for the observed increase in pH. After adding the fermentation supernatant of M. luteus to assist Bacillus in SSF, the diversity and abundance of the flora such as Enterococcus and Staphylococcus significantly increased. This enhancement may improve the metabolic processes related to fat, isoflavones, glucose, amino acids, vitamins, and nucleotides. Additionally, these microbial communities may produce antibiotic metabolites, thereby inhibiting the proliferation of pathogenic bacteria during SSF.
The abundance variation of the Bacillota phylum significantly influences metabolite dynamics during SSF, with its effects manifested in the following aspects. Anthraquinone metabolism enhances antibacterial properties and suppresses pathogenic bacterial proliferation; flavonoid metabolites exhibit diverse biological activities, including antioxidant effects, inhibition of cancer cell proliferation, anti-inflammatory functions, and antimicrobial action. Consequently, Bacillota-associated microbiota critically shapes metabolite composition and functionality by regulating lipid, anthraquinone, and flavonoid metabolic pathways. The elevated levels of these metabolites not only strengthen the antibacterial capacity of fermentation products but also confer multifunctional bioactivities such as antioxidant and anti-inflammatory properties, thereby enhancing the functional value of the fermented materials.
Finally, it should be pointed out that this study verified the synergistic effect of Rpf in a laboratory-scale solid-state fermentation system. Its stability and economic benefits in large-scale industrial production remain to be further evaluated. In addition, the actual feeding effect of fermented soybean meal in animals needs to be directly verified through subsequent animal experiments. Nevertheless, the results of this study provide an innovative technical strategy for the development of new functional protein feeds. If successfully applied, this technology is expected to significantly improve the intestinal health of livestock and poultry, reduce the incidence of diarrhea, and thus reduce the demand for antibiotics by efficiently degrading anti-nutritional factors and enhancing the nutritional value and absorption efficiency of feeds. Meanwhile, the optimization of the metabolome can also enhance the immune function and antioxidant capacity of animals, ultimately providing strong support for improving the production performance, health status, and breeding efficiency of farmed animals.
The observed proliferation of bacterial genera such as Staphylococcus and Enterococcus necessitates a discussion on biosafety, as some species within these genera can be opportunistic pathogens or harbor toxin genes. However, it is crucial to emphasize that safety is strain-specific, and several species within these genera are recognized as safe and are commonly utilized in food fermentations. For instance, Staphylococcus succinus has been recently reported as a functional endogenous microbe that positively influences the flavor profile of fermented chili pepper without raising safety concerns (Li et al., 2023). In this study, the functional role of these bacteria in the synergistic degradation of ANFs is evident. The safety of these taxa is strain-specific, and their presence in our system can be attributed to two main factors: firstly, the artificially inoculated Bacillus strains were isolated from commercial feed additives, which implies a prior screening for safety. Secondly, the functional microbiota resuscitated by Rpf evolved within the solid-state fermentation environment, which is highly competitive and may inherently suppress genuine pathogens. Nevertheless, for any direct feed application, it is imperative to conduct strain-level identification and specific screening for critical virulence factors (e.g., staphylococcal enterotoxins) in the final product to conclusively ensure its biosafety.”
5 Conclusion
In this study, a solid-state fermentation system with multi-bacterial synergy was successfully constructed by combining the resuscitation-promoting factor (Rpf) from M. luteus with B. velezensis and B. amyloliquefaciens. The results showed that the addition of Rpf significantly accelerated the degradation of major anti-nutritional factors in soybean meal (such as glycinin, β-conglycinin, and trypsin inhibitor) in the early stage of fermentation (within 48 hours), and its efficiency was 1.28 - 2.0 times that of the control group. Meanwhile, the Rpf treatment optimized the structure of the fermentation microbial community, increasing the abundance of dominant functional bacteria such as Staphylococcus and Bacillus by 5.34 - 38.38%, and reducing the types and abundance of potential pathogenic bacteria by 62.5% and 13.62% respectively. Metabolomic analysis further indicated that this treatment enhanced the metabolic pathways of lipids, amino acids, and zeatin, and upregulated the synthesis of isoflavones, thereby endowing soybean meal with potential antioxidant and immunomodulatory functions.
The results of this study suggest that Rpf may optimize the fermentation system through a potential mechanism of “functional bacteria resuscitation - Bacillus synergy”. However, the final establishment of this working hypothesis still depends on future controlled experiments using purified Rpf protein for direct verification, and systematic biosafety assessment is required to confirm its application potential. Despite these aspects awaiting confirmation in future research, this study undoubtedly provides an innovative strategy for the green degradation of anti-nutritional factors in soybean meal and the efficient utilization of protein resources.
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.
Author contributions
YZ: Data curation, Formal analysis, Methodology, Software, Writing – original draft. YX: Data curation, Methodology, Writing – review & editing. JM: Data curation, Writing – review & editing. DH: Methodology, Resources, Supervision, Writing – review & editing. XZ: Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Writing – review & editing.
Funding
The author(s) declare financial support was received for the research and/or publication of this article. This research was funded by the Key R&D project of Hebei Province (No. 20326614D), the National Natural Science Foundation of China (NSFC) (No. 32270020), and the Natural Science Foundation of Hebei Province (No. C2025201059).
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that no Generative AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fanim.2025.1707154/full#supplementary-material
Abbreviations
Rpf, Resuscitation Promoting Factor; ANF, anti-nutritional factor; SBM, Soybean Meal; TI, trypsin inhibitor; SSF, solid-state fermentation; LMM, low-nutrient medium; pM, picomolar; T, the treatment group; C, the control group; K, the unfermented SBM; SRA, Sequence Read Archive; PCoA, principal coordinate analysis; LDA, linear discriminant analysis; OPLS-DA, Orthogonal Partial Least Squares Discriminant Analysis; VIP, Variable Importance in Projection; FC, fold change; HCLUST hierarchical cluster analysis; CFU, colony-forming units; MSG, monosodium glutamate.
References
Ali Z., Abdullah M., Yasin M. T., Amanat K., Sultan M., Rahim A., et al. (2025). Recent trends in production and potential applications of microbial amylases: A comprehensive review. Protein Expr Purif 227, 106640. doi: 10.1016/j.pep.2024.106640
Alvarado-Martinez Z., Tabashsum Z., Aditya A., Hshieh K., Suh G., Wall M., et al. (2024). Plant-derived phenolic acids limit the pathogenesis of salmonella typhimurium and protect intestinal epithelial cells during their interactions. Molecules. 29, 1364. doi: 10.3390/molecules29061364
Chen F., Hao Y., Piao X. S., Ma X., Wu G. Y., Qiao S. Y., et al. (2011). Soybean-derived beta-conglycinin affects proteome expression in pig intestinal cells in vivo and in vitro. J. Anim. Sci. 89, 743–753. doi: 10.2527/jas.2010-3146
Chen S., Zhou Y., Chen Y., and Gu J. (2018). fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 34, i884–i890. doi: 10.1093/bioinformatics/bty560
Cheng Y.-H., Hsiao F. S.-H., Wen C. M., Wu C.-Y., Dybus A., and Yu Y.-H. (2019). Mixed fermentation of soybean meal by protease and probiotics and its effects on the growth performance and immune response in broilers. J. Appl. Anim. Res. 47, 339–348. doi: 10.1080/09712119.2019.1637344
Chi C.-H. and Cho S.-J. (2016). Improvement of bioactivity of soybean meal by solid-state fermentation with Bacillus amyloliquefaciens versus Lactobacillus spp. and Saccharomyces cerevisiae. LWT - Food Sci. Technol. 68, 619–625. doi: 10.1016/j.lwt.2015.12.002
Choct M., Dersjant-Li Y., McLeish J., and Peisker M. (2010). Soy oligosaccharides and soluble non-starch polysaccharides: a review of digestion, nutritive and anti-nutritive effects in pigs and poultry. Asian-Australasian J. Anim. Sci. 23, 1386–1398. doi: 10.5713/ajas.2010.90222
Dai C., Hou Y., Xu H., Umego E. C., Huang L., He R., et al. (2022). Identification of a thermophilic protease-producing strain and its application in solid-state fermentation of soybean meal. J. Sci. Food Agric. 102, 2359–2370. doi: 10.1002/jsfa.11574
Ding Y. and Yu Y. (2025). Therapeutic potential of flavonoids in gastrointestinal cancer: Focus on signaling pathways and improvement strategies (Review). Mol. Med. Rep. 31, 109. doi: 10.3892/mmr.2025.13474
Edgar R. C. (2013). UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat. Methods 10, 996–998. doi: 10.1038/nmeth.2604
Enneking D. and Wink M. (2000). “Towards the elimination of anti-nutritional factors in grain legumes,” in Linking Research and Marketing Opportunities for Pulses in the 21st Century: Proceedings of the Third International Food Legumes Research Conference. Ed. Knight R. (Springer Netherlands, Dordrecht), 671–683.
Gilani G. S., Cockell K. A., and Sepehr E. (2005). Effects of antinutritional factors on protein digestibility and amino acid availability in foods. J. AOAC Int. 88, 967–987. doi: 10.1093/jaoac/88.3.967
Gonzalez de Mejia E., Martinez-Villaluenga C., Roman M., and Bringe N. A. (2010). Fatty acid synthase and in vitro adipogenic response of human adipocytes inhibited by α and α′ subunits of soybean β-conglycinin hydrolysates. Food Chem. 119, 1571–1577. doi: 10.1016/j.foodchem.2009.09.044
Han P., Ma X., and Yin J. (2010). The effects of lipoic acid on soybean beta-conglycinin-induced anaphylactic reactions in a rat model. Arch. Anim. Nutr. 64, 254–264. doi: 10.1080/17450391003625003
Han Y.-K. and Thacker P. A. (2011). Influence of energy level and glycine supplementation on performance, nutrient digestibility and egg quality in laying hens. Asian-Australasian J. Anim. Sci. 24, 1447–1455. doi: 10.5713/ajas.2011.11123
Karr-Lilienthal L. K., Kadzere C. T., Grieshop C. M., and Fahey G. C. (2005). Chemical and nutritional properties of soybean carbohydrates as related to nonruminants: A review. Livestock Production Sci. 97, 1–12. doi: 10.1016/j.livprodsci.2005.01.015
Kiarie E. G., Parenteau I. A., Zhu C., Ward N. E., and Cowieson A. J. (2020). Digestibility of amino acids, energy, and minerals in roasted full-fat soybean and expelled-extruded soybean meal fed to growing pigs without or with multienzyme supplement containing fiber-degrading enzymes, protease, and phytase. J. Anim. Sci. 98, skaa174. doi: 10.1093/jas/skaa174
Li D. F., Nelssen J. L., Reddy P. G., Blecha F., Hancock J. D., Allee G. L., et al. (1990). Transient hypersensitivity to soybean meal in the early-weaned pig. J. Anim. Sci. 68, 1790–1799. doi: 10.2527/1990.6861790x
Li Y., Luo X., Guo H., Bai J., Xiao Y., Fu Y., et al. (2023). Metabolomics and metatranscriptomics reveal the influence mechanism of endogenous microbe (Staphylococcus succinus) inoculation on the flavor of fermented chili pepper. Int. J. Food Microbiol. 406, 110371. doi: 10.1016/j.ijfoodmicro.2023.110371
Lienen T., Schnitt A., Hammerl J. A., Maurischat S., and Tenhagen B. A. (2022). Mammaliicoccus spp. from german dairy farms exhibit a wide range of antimicrobial resistance genes and non-wildtype phenotypes to several antibiotic classes. Biol. (Basel) 11, 152. doi: 10.3390/biology11020152
Liener I. E. (1994). Implications of antinutritional components in soybean foods. Crit. Rev. Food Sci. Nutr. 34, 31–67. doi: 10.1080/10408399409527649
Liu Z. L., Chen Y. J., Meng Q. L., Zhang X., and Wang X. L. (2023). Progress in the application of Enterococcus faecium in animal husbandry. Front. Cell Infect. Microbiol. 13. doi: 10.3389/fcimb.2023.1168189
Liu C., Zhao D., Ma W., Guo Y., Wang A., Wang Q., et al. (2016). Denitrifying sulfide removal process on high-salinity wastewaters in the presence of Halomonas sp. Appl. Microbiol. Biotechnol. 100, 1421–1426. doi: 10.1007/s00253-015-7039-6
Lopez Marin M. A., Strejcek M., Junkova P., Suman J., Santrucek J., and Uhlik O. (2021). Exploring the potential of Micrococcus luteus culture supernatant with resuscitation-promoting factor for enhancing the culturability of soil bacteria. Front. Microbiol. 12. doi: 10.3389/fmicb.2021.685263
Lu F., Alenyorege E. A., Ouyang N., Zhou A., and Ma H. (2022). Simulated natural and high temperature solid-state fermentation of soybean meal: A comparative study regarding microorganisms, functional properties and structural characteristics. LWT 159, 113125. doi: 10.1016/j.lwt.2022.113125
Lu D. L., Zhang M. S., Wang F. B., Dai Z. J., Li Z. W., Ni J. T., et al. (2025). Nutritional value improvement of soybean meal through solid-state fermentation by proteases-enhanced Streptomyces sp. SCUT-3. Int. J. Biol. Macromol 298, 140035. doi: 10.1016/j.ijbiomac.2025.140035
Magoč T. and Salzberg S. L. (2011). FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics 27, 2957–2963. doi: 10.1093/bioinformatics/btr507
Malhotra P., Gupta A. K., Singh D., Mishra S., Singh S. K., and Kumar R. (2019). Protection to immune system of mice by N-acetyl tryptophan glucoside (NATG) against gamma radiation induced immune suppression. Mol. Immunol. 114, 578–590. doi: 10.1016/j.molimm.2019.09.003
Mao Y., Chen Z., Lu L., Jin B., Ma H., Pan Y., et al. (2020). Efficient solid-state fermentation for the production of 5-aminolevulinic acid enriched feed using recombinant Saccharomyces cerevisiae. J. Biotechnol. 322, 29–32. doi: 10.1016/j.jbiotec.2020.06.001
Matsuura H. N., Rau M. R., and Fett-Neto A. G. (2014). Oxidative stress and production of bioactive monoterpene indole alkaloids: biotechnological implications. Biotechnol. Lett. 36, 191–200. doi: 10.1007/s10529-013-1348-6
Medeiros S., Xie J., Dyce P. W., Cai H. Y., DeLange K., Zhang H., et al. (2018). Isolation of bacteria from fermented food and grass carp intestine and their efficiencies in improving nutrient value of soybean meal in solid state fermentation. J. Anim. Sci. Biotechnol. 9, 29. doi: 10.1186/s40104-018-0245-1
Menezes J. C. and Diederich M. (2019). Translational role of natural coumarins and their derivatives as anticancer agents. Future Med. Chem. 11, 1057–1082. doi: 10.4155/fmc-2018-0375
Mian S., Saha S., Rabbani M. G., Hossain M. A., Dey T., Nasren S., et al. (2023). Dietary inosine monophosphate improved growth, feed utilization, blood biochemical characteristics, and intestinal histo-morphology of slow growing golden mahseer (Tor putitora). Anim. Feed Sci. Technol. 295, 115545. doi: 10.1016/j.anifeedsci.2022.115545
Mótyán J. A., Tóth F., and Tőzsér J. (2013). Research applications of proteolytic enzymes in molecular biology. Biomolecules 3, 923–942. doi: 10.3390/biom3040923
Mukamolova G. V., Kaprelyants A. S., Young D. I., Young M., and Kell D. B. (1998). A bacterial cytokine. Proc. Natl. Acad. Sci. 95, 8916–8921. doi: 10.1073/pnas.95.15.8916
Mukamolova G. V., Kormer S. S., Kell D. B., and Kaprelyants A. S. (1999). Stimulation of the multiplication of Micrococcus luteus by an autocrine growth factor. Arch. Microbiol. 172, 9–14. doi: 10.1007/s002030050733
Mukamolova G. V., Turapov O. A., Kazarian K., Telkov M., Kaprelyants A. S., Kell D. B., et al. (2002). The rpf gene of Micrococcus luteus encodes an essential secreted growth factor. Mol. Microbiol. 46, 611–621. doi: 10.1046/j.1365-2958.2002.03183.x
Mukherjee R., Chakraborty R., and Dutta A. (2016). Role of fermentation in improving nutritional quality of soybean meal - A Review. Asian-Australas J. Anim. Sci. 29, 1523–1529. doi: 10.5713/ajas.15.0627
Nualkul M., Yuangsoi B., Hongoh Y., Yamada A., and Deevong P. (2022). Improving the nutritional value and bioactivity of soybean meal in solid-state fermentation using Bacillus strains newly isolated from the gut of the termite Termes propinquus. FEMS Microbiol. Lett. 369, fnac044. doi: 10.1093/femsle/fnac044
Nurliana N., Siregar B. H., Sari W. E., Helmi T. Z., and Sugito S. (2022). Identification of cellulolytic lactic acid bacteria from the intestines of laying hens given AKBISprob based on 16S ribosomal ribonucleic acid gene analysis. Vet. World 15, 1650–1656. doi: 10.14202/vetworld.2022.1650-1656
Nurul Huda A., Chuzaemi S., Mashudi M., Ndaru P., and Khoirunisa K. (2022). Nutrient Content and Total VFA Concentration Evaluation by Addition of Condensed Tannin and Myristic Acid in Complete Feed Through In Vitro Method. Proceedings of the 6th International Seminar of Animal Nutrition and Feed Science (ISANFS 2021), vol. 21. Indonesia: Advances in Biological Sciences Research.
Peirotén A., Álvarez I., and Landete J. M. (2020). Production of flavonoid and lignan aglycones from flaxseed and soy extracts by Bifidobacterium strains. Int. J. Food Sci. Technol. 55, 2122–2131. doi: 10.1111/ijfs.14459
Philavong S., Preston T. R., and Leng R. A. (2017). Biochar improves the protein-enrichment of cassava pulp by yeast fermentation. Livestock Res. Rural Dev. 29.
Rigo E., Ninow J. L., Luccio M. D., Oliveira J. V., Polloni A. E., Remonatto D., et al. (2010). Lipase production by solid fermentation of soybean meal with different supplements. LWT - Food Sci. Technol. 43, 1132–1137. doi: 10.1016/j.lwt.2010.03.002
Salim A. A., Grbavčić S., Šekuljica N., Vukašinović-Sekulić M., Jovanović J., Jakovetić Tanasković S., et al. (2019). Enzyme production by solid-state fermentation on soybean meal: A comparative study of conventional and ultrasound-assisted extraction methods. Biotechnol. Appl. Biochem. 66, 361–368. doi: 10.1002/bab.1732
Schloss P. D., Westcott S. L., Ryabin T., Hall J. R., Hartmann M., Hollister E. B., et al. (2009). Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl. Environ. Microbiol. 75, 7537–7541. doi: 10.1128/aem.01541-09
Segata N., Izard J., Waldron L., Gevers D., Miropolsky L., Garrett W. S., et al. (2011). Metagenomic biomarker discovery and explanation. Genome Biol. 12, R60. doi: 10.1186/gb-2011-12-6-r60
Singh M., Singh D., and Srivastava S. (2015). Formulation and evaluation of NLCS encapsulated with S-Allyl-L-Cysteine Sulfoxide for treatment of inflammatory bowel disease. Planta Med. 81. doi: 10.1055/s-0035-1545234
Stackebrandt E. and Goebel B. M. (1994). Taxonomic Note: A place for DNA-DNA reassociation and 16S rRNA sequence analysis in the present species definition in bacteriology. MicroSoc. 44, 846–849. doi: 10.1099/00207713-44-4-846
Su X. M., Bamba A. M., Zhang S., Zhang Y. G., Hashmi M. Z., Lin H. J., et al. (2018c). Revealing potential functions of VBNC bacteria in polycyclic aromatic hydrocarbons biodegradation. Lett. Appl. Microbiol. 66, 277–283. doi: 10.1111/lam.12853
Su X., Li S., Xie M., Tao L., Zhou Y., Xiao Y., et al. (2021). Enhancement of polychlorinated biphenyl biodegradation by resuscitation promoting factor (Rpf) and Rpf-responsive bacterial community. Chemosphere 263, 128283. doi: 10.1016/j.chemosphere.2020.128283
Su X. M., Liu Y. D., Hashmi M. Z., Ding L. X., and Shen C. F. (2015b). Culture-dependent and culture-independent characterization of potentially functional biphenyl-degrading bacterial community in response to extracellular organic matter from Micrococcus luteus. Microb. Biotechnol. 8, 569–578. doi: 10.1111/1751-7915.12266
Su X., Shen H., Yao X., Ding L., Yu C., and Shen C. (2013). A novel approach to stimulate the biphenyl-degrading potential of bacterial community from PCBs-contaminated soil of e-waste recycling sites. Bioresour Technol. 146, 27–34. doi: 10.1016/j.biortech.2013.07.028
Su X., Wang Y., Xue B., Zhang Y., Mei R., Zhang Y., et al. (2018a). Resuscitation of functional bacterial community for enhancing biodegradation of phenol under high salinity conditions based on Rpf. Bioresour Technol. 261, 394–402. doi: 10.1016/j.biortech.2018.04.048
Su X., Xue B., Wang Y., Hashmi M. Z., Lin H., Chen J., et al. (2019). Bacterial community shifts evaluation in the sediments of Puyang River and its nitrogen removal capabilities exploration by resuscitation promoting factor. Ecotoxicol Environ. Saf. 179, 188–197. doi: 10.1016/j.ecoenv.2019.04.067
Su X., Zhang Q., Hu J., Hashmi M. Z., Ding L., and Shen C. (2015a). Enhanced degradation of biphenyl from PCB-contaminated sediments: the impact of extracellular organic matter from Micrococcus luteus. Appl. Microbiol. Biotechnol. 99, 1989–2000. doi: 10.1007/s00253-014-6108-6
Su X., Zhang S., Mei R., Zhang Y., Hashmi M. Z., Liu J., et al. (2018b). Resuscitation of viable but non-culturable bacteria to enhance the cellulose-degrading capability of bacterial community in composting. Microb. Biotechnol. 11, 527–536. doi: 10.1111/1751-7915.13256
Sun P., Li D., Dong B., Qiao S., and Ma X. (2008). Effects of soybean glycinin on performance and immune function in early weaned pigs. Arch. Anim. Nutr. 62, 313–321. doi: 10.1080/17450390802066419
Tahmasbizadeh M., Nikaeen M., Movahedian Attar H., Khanahmad H., and Khodadadi M. (2025). Resuscitation-promoting factors: Novel strategies for the bioremediation of crude oil-contaminated soils. Environ. Res. 271, 121085. doi: 10.1016/j.envres.2025.121085
Tiwari O. N., Sasmal S., Kataria A. K., and Devi I. (2020). Application of microbial extracellular carbohydrate polymeric substances in food and allied industries. 3 Biotech. 10, 221. doi: 10.1007/s13205-020-02200-w
Uauy R., Stringel G., Thomas R., and Quan R. (1990). Effect of dietary nucleosides on growth and maturation of the developing gut in the rat. J. Pediatr. Gastroenterol. Nutr. 10, 497–503. doi: 10.1097/00005176-199005000-00014
Wang Y., Dong J., Jin Z., and Bai Y. (2023). Analysis of the action pattern of sequential α-amylases from B. stearothermophilus and B. amyloliquefaciens on highly concentrated soluble starch. Carbohydr Polym. 320, 121190. doi: 10.1016/j.carbpol.2023.121190
Wang R., Dong P., Zhu Y., Yan M., Liu W., Zhao Y., et al. (2021a). Bacterial community dynamics reveal its key bacterium, Bacillus amyloliquefaciens ZB, involved in soybean meal fermentation for efficient water-soluble protein production. LWT 135, 110068. doi: 10.1016/j.lwt.2020.110068
Wang Q., Garrity G. M., Tiedje J. M., and Cole J. R. (2007). Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl. Environ. Microbiol. 73, 5261–5267. doi: 10.1128/aem.00062-07
Wang J., Lin J., Wang J., Wu S., Qi G., Zhang H., et al. (2020). Effects of in ovo feeding of N-acetyl-L-glutamate on early intestinal development and growth performance in broiler chickens. Poult Sci. 99, 3583–3593. doi: 10.1016/j.psj.2020.04.003
Wang X. F., Liu G. H., Cai H. Y., Chang W. H., Ma J. S., Zheng A. J., et al. (2014). Attempts to increase inosinic acid in broiler meat by using feed additives. Poult Sci. 93, 2802–2808. doi: 10.3382/ps.2013-03815
Wang Y., Shi J., Tang L., Zhang Y., Zhang Y., Wang X., et al. (2021b). Evaluation of Rpf protein of Micrococcus luteus for cultivation of soil actinobacteria. Syst. Appl. Microbiol. 44, 126234. doi: 10.1016/j.syapm.2021.126234
Wang Z. M., Wang S., Bai H., Zhu L. L., Yan H. B., Peng L., et al. (2024). Characterization and application of Bacillus velezensis D6 co-producing α-amylase and protease. J. Sci. Food Agric. 104, 9617–9629. doi: 10.1002/jsfa.13786
Wang M., Zhang Z., Liu X., Liu Z., and Liu R. (2025). Biosynthesis of edible terpenoids: hosts and applications. Foods 14, 673–673. doi: 10.3390/foods14040673
Weaver A. C., Braun T. C., Braun J. A., Golder H. M., Block E., and Lean I. J. (2024). Effects of negative dietary cation-anion difference and calcidiol supplementation in transition diets fed to sows on piglet survival, piglet weight, and sow metabolism. J. Anim. Sci. 102, skae027. doi: 10.1093/jas/skae027
Werner N. and Zibek S. (2017). Biotechnological production of bio-based long-chain dicarboxylic acids with oleogenious yeasts. World J. Microbiol. Biotechnol. 33, 194. doi: 10.1007/s11274-017-2360-0
Wu P., Guo Y., Golly M. K., Ma H., He R., Luo S., et al. (2019). Feasibility study on direct fermentation of soybean meal by Bacillus stearothermophilus under non-sterile conditions. J. Sci. Food Agric. 99, 3291–3298. doi: 10.1002/jsfa.9542
Xue C., Li G., Zheng Q., Gu X., Shi Q., Su Y., et al. (2023). Tryptophan metabolism in health and disease. Cell Metab. 35, 1304–1326. doi: 10.1016/j.cmet.2023.06.004
Yao Y., Li H., Li J., Zhu B., and Gao T. (2021). Anaerobic solid-state fermentation of soybean meal with Bacillus sp. to Improve Nutritional Quality. Front. Nutr. 8. doi: 10.3389/fnut.2021.706977
Ye Z., Li H., Jia Y., Fan J., Wan J., Guo L., et al. (2020). Supplementing resuscitation-promoting factor (Rpf) enhanced biodegradation of polychlorinated biphenyls (PCBs) by Rhodococcus biphenylivorans strain TG9(T). Environ. pollut. 263, 114488. doi: 10.1016/j.envpol.2020.114488
Yin M., Matsuoka R., Xi Y., and Wang X. (2021). Comparison of egg yolk and soybean phospholipids on hepatic fatty acid profile and liver protection in rats fed a high-fructose diet. Foods. 10, 1569. doi: 10.3390/foods10071569
Zhao Y., Qin G., Sun Z., Zhang X., Bao N., Wang T., et al. (2008). Disappearance of immunoreactive glycinin and beta-conglycinin in the digestive tract of piglets. Arch. Anim. Nutr. 62, 322–330. doi: 10.1080/17450390802190318
Keywords: soybean meal, solid-state fermentation, resuscitation promoting factor (Rpf), Bacillus, anti-nutritional factor degradation, microbial flora synergy
Citation: Zhang Y, Xie Y, Ma J, Hu D and Zhang X (2025) Mechanisms of anti-nutritional factor degradation and quality enhancement in soybean meal via Rpf-mediated microbiota synergy and Bacillus reinforcement during solid-state fermentation. Front. Anim. Sci. 6:1707154. doi: 10.3389/fanim.2025.1707154
Received: 24 September 2025; Accepted: 10 November 2025; Revised: 07 November 2025;
Published: 25 November 2025.
Edited by:
Anusorn Cherdthong, Khon Kaen University, ThailandReviewed by:
Rangsun Charoensook, Naresuan University, ThailandRifqi Ahmad Riyanto, Sultan Ageng Tirtayasa University, Indonesia
Copyright © 2025 Zhang, Xie, Ma, Hu and Zhang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Xiumin Zhang, emh4aXVtaW4xMTA2QDEyNi5jb20=; Dong Hu, ZG9uZ2h1MTk4M0AxNjMuY29t