Total amino acid composition of P. citrinopileatus
The basic compositions of P. citrinopileatus powder were assayed before enzymatic hydrolysis. The result of chemical compositions showed that there were high content of crude protein for P. citrinopileatus (34.41 ± 0.04 g/100 g), crude fiber (33.58 ± 0.48 g/100 g), and carbohydrate (12.46 ± 0.29 g/100 g), as well as low content of fat (3.46 ± 0.22 g/100 g) and ash (8.57 ± 0.07 g/100 g). Considering that there was usually a range of protein content from 12.0% to 47.21% in fungus (Zhang et al. 2013; Wang et al. 2014), P. citrinopileatus had a relatively high protein content. Meanwhile, Asp and Glu were monosodium glutamate-like (MSG-like) components that gave the most typical mushroom taste, umami taste, or palatable taste in mushrooms, which earned them the name of umami amino acids. The ratios of umami amino acids to total amino acids generally varied among mushrooms from 21 to 32% (Wang et al. 2014). According to Table 1, umami amino acids of P. citrinopileatus accounted for 42% of the total amino acids, indicating the level of umami-flavored amino acids was relatively high in P. citrinopileatus. In conclusion, P. citrinopileatus could be considered as a potential material to be hydrolyzed by proteases for natural flavoring.
Physical and chemical properties of PCHs
DH was used as a true reflection of the enzymatic hydrolysis process, which influenced product yield, bitterness, and functional properties. As shown in Fig. 1a, the DH of all PCHs increased significantly with time, while after 2 h of hydrolysis, the DH changes of the Papain and AECP became relatively stable, and the DH values of the Alcalase, Neutrase, and Protamex groups were no longer change after 3 h. This result might be due to the specific catalytic sites of the proteases, which gradually reduced as the hydrolysis continued (Ktari et al. 2013). On the basis of these results, 2 h was selected as the optimal duration for the preparation of PCHs by AECP and Papain, whereas 3 hours was chosen as the optimum duration to produce PCHs using Alcalase, Neutrase, and Protamex. Under these reaction conditions, the DH of AECP hydrolysate had peaked at 35.91%, which was markedly higher than other proteases (P < 0.05) (Fig. 1b). In contrast, the DHs of Alcalase, Neutrase, Papain, and Protamex groups were low, ranging from 15.29% to 22.21%, with no significant difference among these four proteases was observed (P > 0.05) (Fig. 1b). Since the proteases varied in cleavage sites on the polypeptide chain, the peptide chain length and amino acid sequence might differ in the hydrolysates prepared from various proteases, which could result in markedly different DH (Himonides et al. 2011).
PR was also used to quantify the utilization rate of raw protein during the preparation of protein hydrolysates. As indicated in Fig. 1b, the highest PR of 81.46% was recorded for AECP hydrolysate, followed by the Papain, Neutrase and Alcalase hydrolysates at 59.13%, 58.71% and 58.05% respectively. The lowest percent PR values were observed as Protamex hydrolysate (55.85%), and no statistically significant PR changes were perceived among the four commercial protease hydrolysates (P > 0.05).
Comparing with commercial proteases, although AECP had the lowest enzymatic activity, it showed highest DH and PR (Fig. 1). The possible explanation for the inconsistent results might be the AECP preparation process, where A. elegans could not directly absorb or utilize the proteins of P. citrinopileatus as the nitrogen source and must secrete protease to break them down into peptides and amino acids. Therefore, the proteases secreted by A. elegans under the induction of P. citrinopileatus could specifically hydrolyze P. citrinopileatus proteins, resulting in higher DH and PR. Depending on the assay performed, P. citrinopileatus also contained carbohydrates (12.46%) and crude fiber (33.58%), in which carbohydrates were the main components of the P. citrinopileatus cell wall, and fiber could easily absorb water and swell, which increased the viscosity of the hydrolysis system and hindered the contact between protease and protein, causing the reducing enzymatic hydrolysis efficiency and decreasing DH. The results, depicted in Fig. 1b, showed the TSC of the AECP hydrolysate was 5.09%, which was relatively higher than the other hydrolysates. It could be inferred that except for protease, A. elegans also secreted glucoamylase and glucanase to decompose carbohydrates (mainly chitin, β-glucan, and mannan) and crude fiber (mainly β-glucan) (Yin et al. 2020), which improved the permeability of the cell wall and reduced the viscosity of the system. Thereby, the progress of proteolysis was accelerated increasing the protein extraction rate. In another study, the protease secreted by Aspergillus oryzae under the induction of the peanut meals was used to hydrolyze peanut meal and obtained high DH and PR (Su et al. 2011), which was consistent with the results of this study.
Molecular weight (MW) distribution analysis of PCHs
To certain extent, the peptide MW distribution of the protein hydrolysates reflected the enzymatic properties of the proteases. Molecular weight distribution of PCHs treated by various proteases were shown in Table 2. It could be seen that 79.91% of the peptides in the Control were below 500 Da, as a result of the heating during the drying process of P. citrinopileatus powders, which degraded part of the protein into small-molecule peptides or amino acids (Li et al. 2015). By comparison, after hydrolysis with different proteases, the number of higher MW peptides (above 1,000 Da) presented a dramatically downward trend (P < 0.05), while those with lower MW (below 500 Da) were found to increase obviously (P < 0.05). Comparing to commercial protease hydrolysates, the content of peptides (500 to 3,000 Da) in AECP hydrolysate was significantly reduced (P < 0.05). Besides, the AECP hydrolysate was rich in fractions with smaller molecular weight (below 500 Da) at 89.34%, which was possibly attributed to endoproteases and exoprotease that A. elegans proteases produced (Sousa et al. 2002; Fu et al. 2011). As could be guessed, the proteins were degraded into peptides by endoproteinases and then further disassembled into smaller peptide segments or free amino acids by exoprotease. It was reported that the use of Umamizyme from A. oryzae for proteolysis could significantly raise the content of small peptides (below 500 Da) (Guerard et al. 2002). Furthermore, peptides produced by proteolysis exhibited unique taste properties that influenced the flavor characteristics of protein hydrolysates, in which, importantly, the low molecular weight peptide displayed a much higher taste-active than the large one. A previous study indicated that the peptides below 500 Da could significantly increase the umami and salty intensity of the system (Su et al. 2011).
Free amino acids, 5'-nucleotide content, and Equivalent umami concentration analysis of PCHs
The free amino acids were divided into several classes based on their taste characteristics in edible mushrooms. The compositions and contents of free amino acids in PCHs were presented in Table 3. There was a significant difference (P < 0.05) in the free amino acid contents among different PCHs. The underlying cause was that enzymes had specific cleavage positions on the polypeptide chain, leading to different amino acids contents of protein hydrolysates obtained from P. citrinopileatus using the various enzymes (Ktari et al. 2013). What's more, after hydrolysis, the content of total free amino acids was significantly increased, while that of His and Arg were greatly lower than the Control, in which the AECP hydrolysates showed the highest production of total free amino acids (120.00 ± 1.45 mg/g), indicating that AECP displayed enhanced proteplysis to release more free peptides and amino acids, which was consistent with the DH results. MSG-like and sweet amino acids could be responsible for the pleasant taste of mushrooms (Sun et al. 2017). Additionally, the bitterness from the bitter amino acids could probably be masked by the sweetness from the sweet components. Remarkably, the MSG-like amino acids content of AECP hydrolysates reached the peak (20.23 ± 0.16 mg/g) and also sweet amino acids (36.40 ± 0.45 mg/g) compared with PCHs by the four commercial proteases. Therefore, the AECP hydrolysate could be regarded as a potential raw material for developing umami ingredients.
Flavor 5'-nucleotides were discovered to be 5'-inosine monophosphate (5'-IMP), 5'-guanosine monophosphate (5'-GMP), and 5'-xanthosine monophosphate (5'-XMP) (Zhang et al. 2013), which also endowed the umami or palatable taste. On the basis of Table 3, content of flavor 5'-nucleotides ranged from 2.68 ± 0.02 mg/g in Neutrase to 4.30 ± 0.07 mg/g in AECP. 5'-GMP was a flavor enhancer with meaty flavor affecting equivalent umami concentration (EUC) that was used for calculating the umami taste of many foods (Beluhan & Ranogajec 2011). The amount of 5'-GMP in Alcalase hydrolysates was markedly highest (2.65 ± 0.06 mg/g), comparing to the Control at 2.52 ± 0.02 mg/g and the other PCHs. Moreover, 5'-IMP was another typical taste-active component in mushrooms, which also had the function to strengthen the flavor (Yin et al. 2019). Besides, 5'-XMP could also provide the umami taste and be transformed into 5'-GMP by the 5'-XMP aminase. Among all PCHs, the content of 5'-XMP was the highest in AECP hydrolysate (2.50 ± 0.05 mg/g), while the lowest content of 5'-IMP was found in Protamex hydrolysate (0.23 ± 0.01 mg/g). It was noticed that the total amount of 5'-nucleotides in all hydrolysates was lower than that in the Control, which was possibly related to the thermal degradation during enzyme inactivation, resulting in the decrease in nucleotides levels (Gao et al. 2021).
In accordance with the preceding report, the mixture of MSG-like components and flavor 5'-nucleotides showed a synergistic effect on the umami taste (Beluhan & Ranogajec 2011). According to Eq. (2), the EUC level of Alcalase hydrolysate (10.25 ± 0.22 gMSG/g) was higher than that of the other PCHs. The relatively high EUC level was observed in AECP hydrolysate (8.74 ± 0.11 gMSG/g), followed by Control (7.68 ± 0.09 gMSG/g), Protamex (7.61 ± 0.15 gMSG/g) and Neutrase (6.82 ± 0.14 gMSG/g), as well as the lowest leve was found in Papain (6.42 ± 0.02 gMSG/g). Mau divided EUC values into four levels: level 1 (> 10 gMSG/g dry matter), level 2 (1-10 gMSG/g), level 3 (0.1-1 gMSG/g), and level 4 (< 0.1 gMSG/g).(Mau 2005) In this case, the EUC value of Alcalase hydrolysate was at level 1, whereas others were at level 2. These results indicated that PCHs might be served as food seasoning components to enhance the umami taste.
Electronic tongue sensory evaluation of PCHs
The electronic tongue determination obtains satisfactory taste outcomes that were approximated by human sensory evaluation, among which the taste sensors could differentiate hydrolysates produced using various proteases. From Fig. 2a, PCHs derived from Neutrase, Papain, Protamex and AECP showed more sourness, bitterness and astringency. This could be interpreted that after enzymatic hydrolysis, the Asp and Glu contents of PCHs, which was in a free form and dissociated state and provided sour taste, were increased significantly, making those samples on response values of sour intensity raise. The ratio of hydrophobic amino acids to total amino acids in P. citrinopileatus protein was 0.29 (Table 1). The possible reason was that the more the hydrolysis, the higher the extent of degradation of native protein structure, and the more exposure of hidden hydrophobic peptides causing bitterness (Fu et al. 2018; Idowu & Benjakul 2019). The umami substances are originally acids, and they exist as salt form at neutral pH, such as monosodium glutamate, disodium 5'-inosinate and disodium 5'-guanylate (Zhang et al. 2017). Since the best enzymolysis pH of Alcalase was alkaline, the pH was adjusted to achieve its optimal reaction pH of 8 using NaOH, which promoted the formation of umami substances. Consequently, Alcalase hydrolysate showed the minimum sourness and the highest umami and saltiness. Notably, the umami taste of the PCHs correlated positively with saltiness and negatively with sourness and bitterness. This might be attributable to the interaction between taste attributes, that is, umami could enhance salty taste and diminished bitterness, whilst sourness mark umami taste (Kim et al. 2015). Meanwhile, there was the best richness (aftertaste-umami) in AECP hydrolysate, presumably because the umami peptides were contained. The umami taste was primarily taken from amino acids, 5’-nucleotides and peptides, however, the retention time of the two previous in the mouth was shorter, bring about the worse aftertaste.
Principal component analysis (PCA) was a multivariate statistical analysis to simply analyze the similarities and differences among samples by reducing the number of dimensions without much loss of information. In this study, PCA was conducted in order to evaluate the differences in sensory attributes (taste) between different PCHs. As shown in the Fig. 2b, the principal component one (PC1) accounted for 95.41% of the total variance whilst the principal component two (PC2) accounted for 3.13% of the total variance. The results illustrated that the E-tongue could be used to distinguish the taste characteristics of all samples, which could be explained that the smaller the distance between the samples, the closer the comprehensive taste and vice versa. According to Fig. 2b, AECP hydrolysates and Alcalase hydrolysates (with lying in the first quadrant and the second quadrant respectively) were different from other PCHs which were located in the fourth quadrant. Therefore, the comprehensive taste of the AECP and Alcalase hydrolysates was quite distinct from that of other PCHs, which were similar. The results revealed that the diverse influences on the types and the content of taste compounds of hydrolysis might result from the different impacts of various proteases.
Correlation between non-volatile components and taste characteristics of PCHs
The sensory distinctiveness of protein hydrolysates was influenced by molecular weight distribution, free amino acid, and 5'-nucleotide composition of hydrolysates (Gao et al. 2021). So PLSR was conducted to give a visual overview of the correlation between molecular weight distribution, free amino acids, 5'-nucleotides, and taste attributes. The X-matrix depicted the molecular weight, free amino acids, and 5'-nucleotides, while the Y-matrix denoted different PCHs and the taste evaluation. Outer and inner ellipses indicate 100% and 50% explained variance, respectively. Results revealed cumulative R2X, R2Y, and Q2 of 0.637, 0.978, and 0.809, respectively, which indicated that the model had great stability and predictive ability. As seen in Fig. 3, the PCHs of Protamex, Neutrase and free amino acids (Arg, Pro and Val) were lying inside the inner ellipse (r2 = 0.5). Additionally, except for the above-mentioned, all other variables were found between the outer and inner ellipses, which demonstrated that the assocations among the variables could be well accounted for by the model. That meant the Protamex, Neutrase, and Papain hydrolysates were not associated with any taste characteristic. The result suggested that those three PCHs had no particularly obvious taste attribute, which corresponded to the electronic tongue test results (Fig. 2). The Alcalase hydrolysate was located in the left-hand higher quadrant, showing umami and salty taste, which were positively influenced by 5'-GMP, 5'-AMP, Lys and His. It had been reported that 5'-GMP and 5'-AMP gave umami or palatable taste (Yin et al. 2019). The pH of the Alcalase hydrolysate was adjusted to 8.0 using NaOH, so that the intensity of saltiness increased as the sodium ion concentration was ramped up higher. Toelstede et al. reported that peptides containing Lys or His had a salty taste (Toelstede et al. 2009), which would potentially lead to increase Alcalase hydrolysate saltiness.
The AECP hydrolysate showed the best richness (aftertaste-umami) but unflavorable mouthfeel features such as sourness, bitterness, and astringency, which had a positive correlation with Phe and 5'-UMP. Furthermore, AECP hydrolysate was similarly found to be distributed in the first quadrants where below 500 Da peptides, free amino acids (Met, Thr, Glu, Tyr, Ile, Leu, Asp, Ser, Ala and Gly) and 5'-nucleotides (5'-IMP and 5'-XMP) were located. Therefore, the above substances also affected the flavor of AECP hydrolysate, just as a previous study showed that a higher proportion of low MW peptides below 500 Da in the protein hydrolysates resulted in a stronger umami taste or umami-enhancing effect (Fu et al. 2018). The AECP hydrolysate could contain umami peptides with glutamate residues or (and) aspartate residues, which elicited umami taste, a broth-like or savory taste, and had an amicable after-taste (Rhyu & Kim 2011; Zhang et al. 2017). Kokumi peptides, that normally showed weak acidity, astringency, and (or) bitterness in plain water (Li et al. 2020), might be included in the AECP hydrolysate, characterizing by taste-enhancing, complex and a long-lasting impression. Studies also reported that glutamyl peptides and leucine peptides had kokumi activity (Liu et al. 2015; Hillmann & Hofmann 2016) .The contents of Glu, Asp, Leu, Gly and Ala in the AECP hydrolysate were higher than other PCHs (Table 3), indicating that AECP possessed specific enzyme cleavage sites pointing at above amino acids. Due to this, the AECP hydrolysate probably included umami peptides and kokumi peptides, causing the best richness (aftertaste-umami). Peptides could also exhibit a bitter and sour tastes. For instance, most peptides composed of hydrophobic amino acids coupled with Phe, Tyr, Trp, Leu, Val, Pro, Ala, Trp, Gly, Met and Ile as well as their respective free forms could display bitterness. Some peptides containing N-terminal Glu, or the presence of Glu and Asp could release sourness (Kong et al. 2017). For this reason, the AECP hydrolysate was characterized by sour, bitter and other unpleasant tastes.