2.1 Reagents and materials

Methanol, sodium 3-trimethylsilyl propionate-2,2,3,3-d₄ (TSP-d₄), sodium lactate, and deuterium oxide (D₂O) were purchased from Sigma-Aldrich Inc. (St. Louis, MO, USA). Formic acid was obtained from Shanghai Aladdin Bio-Chem Technology Co., Ltd. (Shanghai, China). Water used in the experiment was purified by a Milli-Q system (Millipore, Milford, MA, USA). SAA, SV, HAA, HP, HX, and WB were provided by Hebei Chunkai Pharmaceutical Co., Ltd. (Hebei, China). Flour was bought from Shandong Meilejia Food Co., Ltd. (Shandong, China). Straw was purchased from Lixue Thatch Industry Co., Ltd. (Guangdong, China). Reference standards of AMY, caffeic acid (CA), chlorogenic acid (CGA), and tryptophan (TRP) were obtained from the National Institute for Food and Drug Control (Beijing, China). Neochlorogenic acid (NCA), cryptochlorogenic acid (cCGA), isochlorogenic acid A (iCAA), isochlorogenic acid B (iCAB), isochlorogenic acid C (iCAC), cynarin (CYN), hyperoside (HYP), RUT, protocatechuic acid (PA), and ferulic acid (FA) were obtained from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). The purity of all reference standards was above 98% by UPLC analysis.

2.2 Preparation of MMF

According to the current ministry standard of the People's Republic of China (Chinese Pharmacopoeia Commission of Ministry of Health, 1998; Xu et al., 2019), 750 g WB, 15 g SAA, and 15 g SV were smashed and mesh screened, respectively. 75 g HAA, 75 g HP, and 75 g HX were mixed with 2700 mL water and boiled for 1 h, filtered by four layers of gauze to obtain the decoction. Straw (7.5 g) was soaked in 50 mL water, cultured for 1 h in the constant temperature and humidity chamber (YSEI, Chongqing, China) and filtered. The filtrate was mixed with the decoction, and then concentrated to 600 mL under vacuum at 40 °C. The concentrate was added to the mixture of flour, WB, SAA, and SV, and then thoroughly stirred and mixed with hands. A certain amount of the mixture (100 g) was pressed into a small square mold, and then incubated at 33 °C and 80% moisture for eight days. Samples were collected at 0, 48, 96, 144, and 192 h for determination of weight and moisture content and subsequently stored at –80 °C for further UPLC, HS–GC×GC–MS, and NMR analyses. Six batches MMF samples were prepared in parallel.

2.3 Preparation of standard solution and sample solution

2.4 UPLC–PDA analysis and method validation

For the analysis of 14 substrate components in MMF, a Waters ACQUITY UPLC system (Waters, Milford, MA, USA) was used for detection of the mixed standard and MMF sample solutions. Chromatographic analysis was performed on an Agilent ZORBAX SB-C18 column (4.6 × 100 mm, 1.8 μm, Agilent, Santa Clara, CA, USA) at 40 °C in gradient elution with mobile phase consisting of 0.1% formic acid aqueous solution (A) and methanol (B) at the flow rate of 0.5 mL/min. The following gradient elution was applied: 3–4% B during 0–5 min, 4–17% B during 5–10 min, 17–24% B during 10–16 min, 24–38% B during 16–35 min, 38–41% B during 35–38 min, and 41–95% B during 38–39 min. The detection wavelength was set as follows: 296 nm at 0.00–10.00 min, 254 nm at 10.00–12.45 min, 327 nm at 12.45–31.62 min, 254 nm at 31.62–35.00 min, and 327 nm at 35.00–38.00 min. The injection volume was 2 μL.

The analytical method was systematically implemented to validate linearity, limit of detection (LOD), limit of quantification (LOQ), precision (intra- and inter-day), repeatability, stability, and recovery. The calibration curves were constructed with six different concentrations of the tested references as the abscissa and the peak area as the ordinate. LOD and LOQ were estimated by the signal-to-noise ratios (S/N) at 3 and 10, respectively, by injecting a series of dilut solutions with known concentrations. The analysis of intra- and inter-day precision was conducted by six repetitive injections of the same sample solution on the same day and three consecutive days. Six samples from the same source were processed and measured to verify the repeatability. The stability of the tested compounds was investigated by replicated injection of sample solution at 0, 2, 4, 6, 8, 10, and 12 h, respectively. Recovery test was performed by spiking accurate authentic standards to 0.25 g sample power, then treated as sample preparation procedure, which were repeated six times in parallel.

2.5 HS–GC×GC–MS analysis

The MMF samples were analyzed by HS–GC×GC–MS, which was composed of an Agilent 7890B gas chromatograph, an Agilent 5977A mass spectrometer, an Agilent 7697A headspace sampler (Agilent, USA), and a solid–state modulator SSM1810 (J&X Technologies, China). The column set consisted of an Agilent DB-FFAP capillary column (30.00 m length × 0.25 mm internal diameter × 0.25 μm film thickness) coupled with an Agilent DB-17MS capillary column (1.20 m length × 0.18 mm internal diameter × 0.18 μm film thickness). Vial, transfer line, and loop temperatures were set at 120, 130, and 140 °C, respectively. The vial was heated in equilibrium for 10 min; pressure equilibration time and injection time were set at 30 s, respectively. The primary oven temperature was programmed as follows: the initial temperature 50 °C and held for 5 min, increased to 140 °C at the rate of 3 °C/min, increased to 200 °C at the rate of 8 °C/min, and held for 2 min. The secondary oven offset temperature was +30 °C relative to the GC oven. The total analysis time was 44.5 min and the modulation period was 4 s. Ultrahigh purity helium (≥99.999%) was employed as the carrier gas. The injector temperature was set at 220 °C with a split ratio of 1:1. For electron ionization (EI), we used the ionization voltage at 70 eV. The temperatures used were 150 °C for the MS Quad and 230 °C for the MS Source. Full scan mass spectra were acquired in the mass range of 40–400 m/z. The volatile components in MMF were qualitatively analyzed by using the Canvas 1.1.0 comprehensive two-dimensional chromatographic data processing software and NIST 17 mass spectrum library. In the six batches MMF, the common components with values of both similarity and reverse library match greater than 700 were screened out (Savareear et al., 2017).

2.6 NMR conditions and methodological validation of the quantitative method

NMR spectra were acquired on a 600 MHz Bruker AVANCE III NMR spectrometer (Bruker, Zurich, Switzerland) equipped with a proton excitation frequency at 600.23 MHz and an experimental temperature at 298.7 K. All NMR experiments were performed with 90˚ pulse width of 13–14 μs and zgcppr water suppression sequence for each sample. 65,536 data points were collected by 16 scans using a spectral width of 1233.5 Hz and a relaxation delay of 15 s. Chemical shifts were referenced to the TSP-d₄ signal at 0.0 ppm. The linearity, precision (intra- and inter-day), repeatability, stability, and recovery of analytical method were validated comprehensively.

The characteristic proton signals of lactic acid were detected at δ1.33 (3H, d, J = 6.6 Hz, H-3) and 4.13 (1H, q, J = 6.6 Hz, H-2) for the methyl protons and hydroxymethyne proton, respectively. To ensure accurate determination of lactic acid, the proton signal of H-3, which has no interference and symmetry, was selected as the quantitative peak (López-Rituerto et al., 2009). The content of lactic acid was calculated according to the following equation:

2.7 Statistical analysis

The line charts were accomplished by GraphPad Prism 8.30 software (GraphPad Software Inc. San Diego, CA, USA). The cluster heat map of volatile compounds and PCA of the tested compounds on different fermentation days were carried out using Origin 2017b software (Originlab Crop., Northampton, MA, USA) and SIMCA-P 14.1 software (Umetrics, Sartorius Stedim Biotech), respectively. All NMR spectra were processed and analyzed using MestReNova V 9.0.1 software (Mestrelab Research, Spain). The statistical package for social science (SPSS) V 17.0 software (SPSS Inc., Chicago, IL, USA) was applied to analyze the data.

3.1 The key variations of physical properties occured in the fermentation process of MMF

By taking the weight, moisture, temperature, and hyphae growth into account, the variations of key physical properties during the fermentation process of MMF were comprehensively and systematically investigated. The fermentation process of MMF is a complex process involving the dynamic changes of microorganisms, enzymes, and chemical components, dominated by aerobic and anaerobic fermentation. Obviously, the microorganisms consumed the carbon source during the fermentation process, converting them into water, carbon dioxide, and energy, as shown in Fig. 1 (Feng et al., 2016). The weight of MMF continued to decrease from 100 g to 88.8 g with a maximum reduction rate of 11.2% on day 8, while the moisture gradually increased from 39.1% and then fluctuated between 44.4% and 46.8% during the fermentation anaphase. Certainly, the increase of moisture was not only caused by the decrease of the MMF's weight, but also the more water produced in the process of fermentation. Also, production of water outweighed evaporation, resulting in the increase amount of water. During the fermentation process, the temperature of the sample firstly increased to 41.1 °C on the 4th day, and then decreased to room temperature. Interestingly, the hyphae gradually covered the surface of MMF, and continually became denser during the fermentation. The hyphae gradually grew to 7 mm in length, whose color turned from pure white to yellow-white or gray-white. This may be due to the variations of species and quantity of microorganisms in different fermentation periods. At the initial stage of fermentation, abundant species were observed in the raw materials of MMF, including fungi and bacteria. With the progress of fermentation process, filamentous fungi such as Aspergillus spp. gradually became the dominant flora, which may be associated with the growth of hyphae (Liu et al., 2017). As one kind of dominant fungi in MMF, the appearance of Rhizopus oryzae changed from white to dark brown during fermentation. It is reasonable to deduce that the growth of Rhizopus oryzae gives rise to the change of appearance (Chen et al., 2020).

Close

Fig. 1

Variations of the hyphae growth (a), hyphae length (b), weight and moisture (c), and temperature (d) during the fermentation process of MMF.

Export to PPT

In our previous fermentation study, it was found that the hyphae easily turned black during the fermentation process of MMF. It has been documented that addition of rice straw to the fermentation of Chungkukjang resulted in transferring microorganisms from the straw into the Chungkukjang fermentation matrix (Heu et al., 1999). Up to 103 species of microorganisms have been isolated from rice straw, including Aspergillus spp., which can produce a variety of enzymes and multiple aromatic components (Kim et al., 2013). It was found that Aspergillus spp. were the dominant fungi at the initial stage of MMF fermentation (Chen et al., 2020). Therefore, the use of straw-covered MMF substrates for fermentation is reasonably feasible, which was also found to be effective in avoiding the black hyphae situation. To make it easier to process the samples in the subsequent steps, the straw soaking water solution was added to the fermentation substrate, which was consistent with the effect of the straw covering on the surface of MMF substrate for fermentation. Studies have shown that the soaked solution with straw can promote the growth of microorganisms and improve fermentation efficiency (Gao et al., 2016).

3.2 Optimization of the quantified method and application in studying dynamic trend of substrate components during fermentation of MMF

During the fermentation process of MMF, microorganisms continued to grow in the medium, which produced a variety of digestive enzymes such as lipase, protease, and amylase (Xu et al., 2013). The enzymes transformed macromolecules into small molecules that could be easily absorbed, for instance, transformation of starches into monosaccharides and oligosaccharides, which regulated the intestinal flora and promoted the growth of beneficial bacteria (Wilson and Whelan, 2017). The interaction of microorganisms and enzymes may involve the decomposition and transformation of chemical components, displaying the effects of invigorating the spleen and promoting digestion after fermentation. Therefore, the fermentation of MMF was studied to focus on the changes of chemical components during the fermentation process.

The chromatographic conditions were optimized to accurately determine the 14 chemical components in MMF with satisfactory peak shape, sensitivity, resolution, and retention time. The optimized mobile phase composed of 0.1% formic acid aqueous solution-methanol showed better effect on separation than 0.1% formic acid solution-acetonitrile, water-acetonitrile, and water-methanol. Agilent ZORBAX SB-C18 column provided superior resolution and peak shape compared to ACQUITYTM UPLC BEH C18 column. After compared with 30 °C and 50 °C, the column temperature at 40 °C performed better effect on separation and peak shape. The key factors that affect the extraction efficiency were optimized for the focused components, including the extraction solvent (25%, 30%, 35%, 40%, 45%, 50%, 75% methanol aqueous solution, and methanol), ratio of material/solvent (1:100, 1:50, and 1:25), and ultrasonic time (20 min, 30 min, and 40 min). In order to concisely and visually evaluate the efficiency of the extraction method, the “spider-web” mode was preferably employed for optimizing the extraction conditions (Li et al., 2019). The content of each compound under all extraction conditions was normalized, and the best extraction condition was selected according to the shaded area of “spider-web” mode, as shown in Fig. 2a. Finally, 45% methanol aqueous solution for 30 min at a ratio of 1:50 material/solvent emerged as the optimal extraction condition, whose “spider-web” area was calculated to be 2.74 (Fig. S1).

Close

Fig. 2

Optimization of the extracting method for MMF by “spider-web” mode (a); UPLC-PDA chromatograms of MMF sample solution acquired by wavelength switching mode (b₁), amygdalin in MMF sample solution at 210 nm (b₂), the mixed standard solution (b₃) acquired by wavelength switching mode, and the standard solution of amygdalin at 210 nm (b₄). 1: PA, 2: TRP, 3: NCA, 4: CGA, 5: cCGA, 6: CA, 7: CYN, 8: FA, 9: iCAB, 10: iCAA, 11: HYP, 12: RUT, 13: iCAC, 14: AMY.

Export to PPT

Based on UPLC-PDA method, a multi-components quantitative analysis of MMF was accomplished by employing 14 compounds as indicators. The typical chromatograms of the sample and mixed standard solution of MMF are displayed in Fig. 2b. The systematical methodology was validated in terms of intra- and inter-day precision, stability, linearity, LOD, LOQ, repeatability, and recovery (Table 1). For the 14 tested compounds, determination coefficient values were above 0.9990, which indicated good linear correlation. The LOD and LOQ values were 0.03175–2.064 μg/mL and 0.06350–4.128 μg/mL, respectively. The RSDs of intra- and inter-day precision, stability, and repeatability were all below 3%. In addition, the mean recovery ranged from 92.36% to 105.2%, with RSD value < 3%. Therefore, it was demonstrated that this method is reasonable and feasible for determining the content of 14 components in MMF.

The variation of the substrate components during the fermentation process is the fundamental problem that is related with the efficacy of fermented Chinese medicine. Therefore, elucidation of the dynamic trend of chemical compositions is the key to reveal the fermentation mechanism of MMF. This study focused on the chemical components present in the original medicinal materials and fermentation substrate, whose content was determined by the established UPLC-PDA method in MMF with the different degrees of fermentation. The variations of the 14 components during the fermentation process are shown in Fig. 3, from which FA and TRP were found to dominate. FA showed an upward trend from 45.53 ± 6.94 to 141.89 ± 78.40 μg/g (n = 6) during the fermentation process, which was released from esterified FA by FA esterase secreted by microorganisms (Oliveira et al., 2020). FA belongs to the phenolic acid group and possesses many physiological functions, such as acceleration of gastrointestinal motility, anti-inflammatory, antioxidant, and antimicrobial activities (Badary et al., 2006). TRP content did not change significantly after fermentation. Studies have shown that TRP as a common essential amino acid is an important regulator of inflammation and immunity, and has a certain therapeutic effect on inflammatory bowel diseases (Tao and Wu, 2020). Also, TRP has been closely related to the management of digestive diseases and has the function of promoting gastric emptying (Tachibana et al., 2018). Therefore, these compounds with higher content in MMF contribute to treating diseases.

Close

Fig. 3

The dynamic curves of index components content during MMF fermentation process, including compounds with stabilized content (a), compounds with increased content (b), glycoside components with reduced content (c), and phenolic acid components with reduced content (d). Comparison of index components content on days 2, 4, 6, and 8 during the fermentation process with that on day 0, respectively (*P < 0.05, **P < 0.01, ***P < 0.001).

Export to PPT

The content of glycosides and most of the phenolic acids were found to decrease obviously. AMY as glycoside is the main pharmacological ingredient in SAA, whose content decreased sharply from 95.80 ± 5.58 to 3.30 ± 0.72 μg/g during the fermentation process. This may be due to the deglycosylation reaction by enzyme. The content of the phenolic acids decreased along with the increased fermentation time, especially cCGA, iCAA, and iCAC, which could not be detected after four days' fermentation. Interestingly, the content of iCAB, CGA, and NCA decreased significantly in two days and tended to stabilize. Hence, it is meaningful to unveil this diversified variations of the tested compounds, which are possibly correlated with the function of MMF.

3.3 Variations of volatile metabolites during MMF fermentation revealed by HS–GC×GC–MS

The traditional fermented foods include vinegar, wine, curd, and dairy products, whose fermentation involves complex microbial systems, produces many active substances beneficial to humans, and enhances the overall flavor (Huang et al., 2018). Volatile components are the compounds that make up the unique flavor of fermented foods, such as alcohols, esters, aldehydes, ketones, acids, furans, and phenols.

By using HS–GC×GC–MS, an investigation was performed to identify and analyze 66 volatile components in the MMF fermentation process, including 22 alcohols, 21 aldehydes and ketones, six acids, three esters, and 14 other compounds, as shown in Table S1. The analytical data showed that the relative content of the 66 volatile components varied significantly along with fermentation time, as shown in Fig. 4a. The heatmap plot indicated the relative abundances of volatile compounds in MMF at the different fermentation times. Moreover, according to the variations on the relative content of the volatile compounds, the tested compounds were divided into three groups. The first group included 13 components, which were consistently detected with comparatively high abundance throughout the fermentation process, such as hexanal, benzaldehyde, and so on. The second group had 28 components, which were presented in the fermentation prophase, including 1-butanol, octanoic acid, nonanal, and so on. The third group comprised 25 components from MMF at the fermentation anaphase, including ethanol, 1-propanol, 2,3-butanediol, etc.

Close

Fig. 4

Heatmap plots (a) and column diagram (b) of relative content of volatile compounds during fermentation process of MMF.

Export to PPT

A total of 21 aldehydes and ketones were detected during the fermentation process, which were the main volatile components on day 0, with hexanal and benzaldehyde dominating. As shown in Fig. 4b, the content of aldehydes and ketones dynamically changed in MMF during the fermentation process, which was observed by a significant reduction at 72.35% throughout the fermentation process, and tended to stabilize after four days. Aldehydes and ketones can be formed by degradation of amino acids or oxidation of unsaturated fatty acids. At the same time, aldehydes can be reduced to alcohol or oxidized to carboxylic acids due to its instability (Wang et al., 2015).

Twenty-two alcohols were detected during the fermentation of MMF, whose total relative content increased gradually and then stabilized on the 4th day. Among them, the relative content of 3-methyl-1-butanol, 2-furanmethanol, benzyl alcohol, and phenylethyl alcohol increased significantly by 99.65%, 98.25%, 97.17%, and 99.35% on day 4 respectively and tended towards stable at the fermentation anaphase. Instead, 1-pentanol and (E)-2-octen-1-ol were in high relative abundance on day 0, which gradually decreased with the progress of fermentation and cannot been detected on days 6 and 8. Ethanol was detected on day 6, which may be caused by the anaerobic environment of MMF at this stage. Two main sources are considered to procduce alcohol from sugar catabolism and the conversion of amino acids through the Ehrlich pathway (Kim et al., 2014).

Acids are secondary metabolites of yeast, which can be formed by oxidation of alcohols (Bao and Zhang, 2012). Six acids were detected during the fermentation process of MMF, namely acetic acid, 2-methyl-propanoic acid, 3-methyl-butanoic acid, hexanoic acid, pentanoic acid, and octanoic acid. Their total relative content was found to increase by 40.64%. Esters are mainly generated by the esterification of organic acids and alcohols under the action of enzymes (Wang et al., 2015). Esters were not detected on the first two days, but observed at the fermentation anaphase, particularly hexadecanoic acid ethyl ester.

3.4 Dynamic trend of lactic acid detected by NMR during the fermentation process of MMF

As an active product of anaerobic fermentation, lactic acid is an important indicator to evaluate the degree of fermentation and its quality. Because of no conjugated group and low volatility, lactic acid cannot be analyzed directly by general methods, such as HPLC-UV and GC. Derivatization of lactic acid can overcome these limitations, but it leads to tedious procedures and employment of toxic reagents (De Baere et al., 2013; Jarukas et al., 2020). With the advantages of simple sample pre-treatment procedure, time-saving, and efficient determination of structural information and quantitative information (Avenoza et al., 2006), NMR was employed as an alternative method to monitor lactic acid during the fermentation process of MMF.

The illumination of lactic acid was conducted by ¹H NMR and ¹H–¹H correlation spectroscopy (¹H–¹H COSY) spectra of MMF samples, which are illustrated in Fig. 5a. The satisfactory validation of the methodology implemented paved the way for NMR analysis of lactic acid in MMF, the detailed results of which are shown in Table 1. Excellent linear relationship was achieved for lactic acid in the tested range. The RSDs of intra- and inter-day precision, stability, and repeatability were all below 2.0%. Besides, the mean recovery was 101.0%, with RSD value < 2.6%. The validated method was successfully applied to detect lactic acid in MMF. Lactic acid was not detected in MMF substrate on day 0, but an obvious increase in lactic acid content was observed along with the increased fermentation days, resulting in 2.54 ± 0.15 mg/g on day 8 (Fig. 5b). This change was possibly derived from the lactic acid bacteria during MMF fermentation, including Lactobacillus, Pediococcus, Weissella, etc. (Wang et al., 2020; Xu et al., 2013). Moreover, lactic acid is known to promote digestion and improve the function of the gastrointestinal tract, which can be originated from glucose through the glycolytic pathway (homolactic acid metabolism) or the PK pathway (heterolactic acid metabolism) (Abdel-Rahman et al., 2013), as shown in Fig. 5c. The illumination on production and dynamic variation of lactic acid sheds the light on the fermentation mechanism and suggests lactic acid as a quality marker of MMF for quality control.

Close

Fig. 5

The ¹H NMR and ¹H–¹H COSY spectra of MMF samples (a); the line chart of lactic acid content during MMF fermentation process, whose content on days 2, 4, 6, and 8 during the fermentation process by comparing with that on day 0, respectively, ***P < 0.001 (b); metabolic pathways for lactic acid production from sugars, including glycolytic pathway (homolactic acid metabolism) and PK pathway (heterolactic acid metabolism) (c).

Export to PPT

3.5 The fermentation characteristics of MMF revealed by PCA

The components from substrates (14 componments), volatile components (66 componments), and lactic acid from MMF show dynamic trends during the fermentation process, which are closely related to the fermentation degree. Thus the fermentation characteristics of MMF can be distinguished according to the characteristics of variations for components. The PCA score scatter plot illustrated that the first principal component (PC1) accounted for 40.6% of the total variation, while PC2 amounted to 11.5% (Fig. 6a). The analytical results of PCA showed that the samples from day 0, the samples from day 2, and the samples from day 4, 6, 8 exhibited the obvious difference, while the samples from day 4, 6, 8 showed the similar characteristics. MMF samples from the different fermentation days were distributed clockwise. As shown in Fig. 6a and b, MMF samples from day 0 were distributed in the third quadrant and mainly characterized by 1-pentanol, octanoic acid, and iCAC. Meanwhile, the samples from day 2 were located in the first and second quadrants and chiefly characterized by 1-hexanol, 1-hydroxy-2-propanone, and 3-ethyl-2-methyl-1,3-hexadiene. Additionally, MMF samples from days 4, 6, 8 were located in the first and fourth quadrants and principally characterized by lactic acid, octanal, 3-octanone, PA, and acetic acid. The orthogonal partial least squares discriminant analysis (OPLS-DA) model was used to analyze the association between the content of the tested components and the fermentation days. It was displayed that R² and Q² in the model were 0.959 and 0.913 respectively, suggesting the OPLS-DA model was well fit for analysis and prediction. The VIP values of analyzed components varied from 0.3999 to 1.4076 (Fig. 6c), among which a total of 32 components, including (E)-2-hexenal, 3-furaldehyde, (Z)-3-hexen-1-ol, ethanol, 1-hydroxy-2-propanone, 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one, acetic acid, lactic acid, and PA (VIP _(pred) > 1.0), had important effects on differentiation of fermentation stages.

Close

Fig. 6

Score plot (a), loading plot (b), and variable importance for predictive omponents (VIP_(pred)) (c) using principal component analysis (PCA) by employing the variations of substrate components, volatile components, and the lactic acid at different fermentation times of MMF.

Export to PPT

During the fermentation process, the fermentation temperature elevated due to the growth of microorganisms, and the enzymatic catalysis promoted the production of metabolites. From this, it is clear that the fermentation of MMF relies on the biotransformation of microorganisms and enzymes. It has been reported that the stages of fermentation were discerned based on the growth status of microorganisms during the fermentation process (Peleg et al., 2011), which also can be categorized into three distinct stages, namely 0–2 days (growth lag phase), 2–4 days (logarithmic phase), and 4–8 days (stationary phase) based on the variations of the tested components. To make the best use of time and fermentation resources, four days for fermentation was proposed to fermentate MMF. Zhang et al. investigated the dynamic changes of amylase, saccharifying enzyme, and protease activity in MMF at the different fermentation times. They concluded that the optimal fermentation time was four days (Zhang et al., 2018), which is consistent with the results of our study.