Reduction in the bitterness of protein hydrolysates by an aminopeptidase from Aspergillus oryzae

Abstract

An aminopeptidase was purified from Proteax, a food-grade commercial enzyme reagent derived from Aspergillus oryzae that showed the highest debittering effect on casein hydrolysates among nine such reagents, and its enzymatic characteristics and debittering activities were investigated. Treatment with the enzyme, identified as leucine aminopeptidase A (LapA), on pepsin-degraded casein and cod protein solutions reduced their bitterness and increased the amount of free hydrophobic amino acids such as leucine, valine, isoleucine, and phenylalanine, as well as arginine, tyrosine, and threonine. These results indicated that LapA preferentially released hydrophobic amino acid residues at the amino terminus of peptides including bitter-tasting peptides and reduced the bitterness of protein hydrolysates.

Introduction

Hydrolysis of proteins with proteolytic enzymes improves solubility, digestibility, and absorption (Adler-Nissen, 1976). In recent years, it has become clear that peptides are more digestible and absorbable than amino acids (Hellier et al., 1972). With advancing age, humans tend to become undernourished because of protein deficiency due to reduced digestion and absorption. Therefore, nutritional supplements in the form of peptides are used for the purpose of efficient protein intake. They are also used as post-exercise protein supplements. However, when proteins are hydrolyzed, bitter-tasting peptides are produced, which can cause an unpleasant bitter taste. Proteins themselves have no taste, but when they are hydrolyzed, the hydrophobic amino acids are exposed and stimulate the bitter taste receptors on the tongue, producing a bitter taste (Saha and Hayashi, 2001). The production of a bitter taste limits the use of protein hydrolysates in the food processing industry. Since bitterness is closely related to the hydrophobicity of peptides, there have been attempts to reduce bitterness by hydrolyzing bitter peptides with exopeptidases (Raksakulthai and Haard, 2003; Komai et al., 2007; Huang et al., 2015; Song et al., 2020). It is also reported that aminopeptidases purified from Aeromonas caviae T-64 (Izawa et al., 1997) and the edible basidiomycete Grifola frondosa (Nishiwaki et al., 2002) preferentially released hydrophobic amino acids at the amino terminus of peptides and reduced the bitterness of soy protein and casein hydrolysates. However, these enzymes are not commercially available, a major impediment to their use to reduce the bitterness of food peptides. Commercial enzyme reagents include endopeptidases and exopeptidases. The former randomly cleaves protein molecules to produce peptides, while the latter releases the terminal amino acids of peptides. In this report, an aminopeptidase, which is an exopeptidase, was purified and showed to clearly demonstrate a bitterness-reducing effect. In addition, there are no studies that have purified aminopeptidases from commercially available enzymes for food use and examined their effects on the bitterness reduction of protein hydrolysates. Soybeans, milk (casein), and whey are relatively pure protein materials, and surimi (cod) was also considered to be a potential material because of its popularity in Japan. Moreover, the muscle protein of Alaska pollock has an excellent amino acid score of 100 and a very high protein utilization rate in the body. It has been reported that muscle growth is increased simply by ingesting the protein (Mizushige et al., 2010). Thus, cod protein was used in this experiment because of its low content of amino acid residues that impart a bitter taste, making it suitable for the preparation of seasoning peptides, and it is also expected to be a good material as a dietary supplement.

In this study, we examined the debittering effect of nine commercial enzyme reagents on protein hydrolysates. Proteax, an enzyme reagent derived from A. oryzae, showed the highest debittering effect, and an aminopeptidase purified from Proteax effectively reduced the bitterness of both casein and cod protein hydrolysates.

Materials and Methods

Debittering activity assay of commercial enzyme reagents    Casein derived from milk (Fujifilm Wako Pure Chemicals, Osaka, Japan) was suspended in distilled water to prepare a 5.0% (w/v) solution, and HCl was used to adjust the pH to 2.0. To the casein solution (2 000 mL), 2.0 g [Enzymes/Substrates (E/S) = 1/50] of pepsin (P7000, Sigma-Aldrich, St. Louis, MO, USA) was added, incubated at 37 °C for 2 h, and then heated at 100 °C for 5 min to inactivate the pepsin. The mixture was then centrifuged at 5 900 × g for 20 min, and the obtained supernatant was used as the casein hydrolysate.

The commercial food-grade enzyme reagents used were Proteax, Protease M “Amano” SD, Protease P “Amano” 3SD, Protease A “Amano” SD, Peptidase R, Protin SD-AY10, Neurase F3G, Protin SD-NY10, and Samoase PC10F. All enzyme reagents were obtained from Amano Enzyme Inc. (Aichi, Japan). To 89.1 mL of the casein hydrolysate solution, 0.90 mL of 500 mM phosphate buffer (pH 7.0) was added, and each enzymatic reagent was added to yield E/S = 1/50. E/S indicates the ratio of enzyme preparation (g) to protein material (g) by weight. The mixture was incubated at the optimum temperature of each enzyme for 2 h, then the mixture was cooled on ice, centrifuged at 5 900 × g for 20 min, and the supernatant was used for the sensory evaluation. Reaction temperatures for each enzyme were as follows: Protease M "Amano” SD, Protease P “Amano” 3SD, and Protease A "Amano” SD at 45 °C; Proteax, Peptidase R, Protin SD-AY10, Neulase F3G, and Protin SD-NY10 at 55 °C; Thermoase PC10F at 65 °C.

Enzyme assay    Casein derived from milk (Fujifilm Wako Pure Chemicals) was used as the substrate for the measurement of endopeptidase activity. Casein solutions (2% w/v) in 0.1 M phosphate buffer (pH 6.2, 7.2, and 8.2) and an equal amount of enzyme solution were reacted, and then the reaction was terminated by the addition of trichloroacetic acid. The absorbance of the supernatant at 280 nm was measured after centrifugation (11 000 × g, 2 min).

l-Leucine p-nitroanilide (Leu-pNA), Ala-pNA, and Glu-pNA were obtained from Peptide Research Institute, Inc. (Osaka, Japan), and used as substrates for the measurement of aminopeptidase activity. The enzyme activity was assessed with 2 mM of Leu-pNA, unless stated otherwise, in 25 mM Tris(hydroxymethyl)aminomethane (Tris)-HCl buffer (pH 7.46), and trichloroacetic acid solution was added to stop the reaction. The absorbance at 405 nm was measured to determine the amount of p-nitroaniline (pNA) released.

Benzyloxycarbonylglycyl-l-phenylalanine (Z-Gly-Phe) obtained from Peptide Research Institute was used as a substrate for the measurement of carboxypeptidase activity. Enzyme regents were reacted with 7.5 mM Z-Gly-Phe in 50 mM Tris-HCl buffer (pH 7.30) at 30 °C for 90 min, and then the released phenylalanine was determined with a ninhydrin solution (Fujifilm Wako Pure Chemicals). Absorbance at 590 nm was measured to determine the amount of free phenylalanine.

Purification of aminopeptidase    Proteins in Proteax, a food-grade enzyme reagent from A. oryzae, were separated by FPLC with an anion exchanger TOYOPEARL GigaCap Q-650M (2.64φ × 33.5 cm; Tosoh Corp., Tokyo, Japan). Solvent A was 25 mM phosphate buffer (pH 7.71), and solvent B was solvent A + 0.5 M NaCl. A 100-mL aliquot of the enzyme reagent solution (50 mg/mL) was introduced into the column and the proteins were eluted with a linear gradient of NaCl from 50 to 200 mM (20 column volumes, 3.7 L) at a flow rate of 6 mL/min. The absorbance of each fraction (12 mL) at 280 nm and the aminopeptidase activity were measured.

Enzyme characterization of the aminopeptidase    The enzyme activity was measured at 30 to 90 °C to estimate the optimum temperature. Temperature stability was determined by holding the samples at each temperature for 30 min and measuring the residual activity. The optimal reaction pH and pH stability of the aminopeptidase were estimated in 2-(N-morpholino) ethanesulfonic acid (MES)-NaOH (pH 5.30–7.24), 2-(N-morpholino)ethanesulfonic acid (MES)-NaOH (pH 6.48–8.18), 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-NaOH (pH 6.54–8.45), Tris-HCl (pH 7.43–9.20), and N-cyclohexyl-3-aminopropanesulfonic acid (CAPS)-NaOH (pH 9.80–11.44). Enzyme stability at various pHs was determined by holding the samples at 60 °C for 30 min in each buffer solution and measuring the residual activity. Since it was inferred that this enzyme is a metalloenzyme that possesses Mg in its active center, the effect of Mg addition was examined. The effect of ethylenediaminetetraacetic acid (EDTA) on enzyme activity was measured at final concentrations of 0 to 10 mM. The effect of 14 metal compounds (PbCl₂, CoCl₂, NiCl₂, CuCl₂, CaCl₂, BaCl₂, FeCl₃, AlCl₃, MnCl₂, FeCl₂, SnCl₂, MgSO₄, ZnCl₂, and LiCl) on enzyme activity was measured in the presence of each metal ion at a final concentration of 1 mM.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed to verify the purity of the enzyme. Equal amounts of a reducing solution containing SDS and 2-mercaptoethanol and the samples were mixed and heated (95 °C, 5 min) prior to the analyses. Samples were run in a 16.5% PAGEL (ATTO Corp., Tokyo, Japan) at 70–200 V, 42 mA for 90 min using an electrophoresis apparatus (WSE-1150PageRun Ace, ATTO Corp.). The molecular weight marker used was EzProtein Ladder (ATTO Corp.).

Identification of the aminopeptidase gene    A shotgun analysis by means of a mass spectrometer was performed to identify the encoding gene of the aminopeptidase. An aliquot of the aminopeptidase was alkylated after reduction, and then digested with trypsin to make samples for LC-MS/MS analyses. The samples were measured using an UltiMate^® 3000 (Thermo Fisher Scientific Inc., Waltham, MA, USA) for LC and a Q-Exactive Plus (Thermo Fisher Scientific Inc.) for MS. The obtained MS/MS data were searched with the NCBI databaseⁱ⁾.

Debittering activity assay of the purified aminopeptidase    To 1 000 mL solutions of 10% (w/v) casein and 10% (w/v) cod protein (kindly supplied by Professor Kazufumi Osako, Tokyo University of Marine Science and Technology), in which the pH was adjusted to 2.0 with HCl, 333 mg of pepsin (P7000, Sigma-Aldrich) was added (E/S = 1/300). The casein solution was incubated at 37 °C for 4 h, then the pH was adjusted to 2.0 with HCl, and the mixtures were further incubated at 37 °C for 16 h. The cod protein solution was incubated at 37 °C for 2 h, then the pH was adjusted to 2.0 with HCl, and the mixtures were further incubated at 37 °C for 5 h. After incubation, the mixtures were cooled in ice, adjusted to pH 7.0 with NaOH, and heated at 100 °C for 15 min to inactivate pepsin. The mixtures were centrifuged at 8 300× g for 15 min, and the supernatants were used as casein hydrolysate and cod protein hydrolysate, respectively.

Prior to the aminopeptidase treatment, the pH of protein hydrolysates was adjusted to 8.0 with NaOH. Appropriate amounts of the purified enzyme solution (0.157 mg/mL) were added to 350 mL of the casein hydrolysate (E/S = 1/24 500) and 350 mL of the cod protein hydrolysate (E/S = 1/80 000), and then incubated at 50 °C. As mentioned above, E/S indicates the enzyme-substrate ratio by weight. The enzyme amount (g) was calculated for E by setting A280 = 1.0 as 1.0 mg/mL. S indicates the amount of protein used (g). Small aliquots were withdrawn after 1, 2, 4, 8, and 21 h for further characterization.

Molecular weight distributions of protein hydrolysates were estimated by gel filtration chromatography, which was performed as follows. Samples were separately applied to a Superose 12 10/300 GL column (Cytiva, Tokyo, Japan), eluted with 100 mM ammonium acetate (pH 7.0) containing 0.02% sodium azide at a flow rate of 0.40 mL/min, and the absorbance at 280 nm was monitored. Carbonic anhydrase (molecular weight 29 000), cytochrome C (12 400), and aprotinin (6 500), purchased from Sigma-Aldrich, and l-Tyr (181), purchased from Fujifilm Wako Pure Chemicals, were used as molecular weight markers.

The free amino acid compositions of protein hydrolysates were analyzed with an L-8900 Rapid Amino Acid Analyzer (Hitachi High-Tech, Tokyo, Japan) according to the manufacturer's protocol after the removal of acid insoluble materials.

The bitterness of protein hydrolysates was evaluated by 9 panelists, consisting of two men and six women in their 20s, and one man in his 60s. The sensory testing was performed using a graded scale in comparison with reference solutions, which were 0.5, 1.0, and 1.5% Gly-Phe (Peptide Research Institute) aqueous solutions with bitterness scores of 1, 2, and 3, respectively. The bitterness score was set at 5 (maximum) to 0 (no bitterness). Casein hydrolysate was provided in a 6-fold dilution due to its strong bitterness. The t-test was conducted to confirm whether bitterness was significantly reduced by AMP treatment. The sensory testing was approved by the Ethics Committee of Toyo University (approval number: TU2021-018).

Results and Discussion

Debittering of casein hydrolysate by commercial enzymes    The results of sensory evaluation are shown in Supplementary Fig. 1. The bitterness of casein hydrolysate (5% w/v) was set at 100%, and when treated with various enzymatic reagents, Thermoase PC10F > Protin SD-NY10 > Neulase F3G > Protin SD-AY10 > Peptidase R > Protease A “Amano” SD > Protease P “Amano” 3S D > Protease M “Amano” SD > Proteax showed bitter taste in this order. Thermoase PC10F did not reduce the bitterness and presented the same bitterness as casein hydrolysate (5% w/v). Among the 9 commercial enzymes, Proteax was found to have the highest debittering effect. The endo- and exo- peptidase activities of the various enzyme reagents were then measured. Proteax, which reduced bitterness the most, had higher aminopeptidase activity (substrate, Leu-pNA) (Supplementary Table 1). This means that the aminopeptidase of Proteax released the N-terminal amino acid of the peptide, which may have reduced the bitterness. However, it also showed activity, albeit weak, against casein, the substrate of endopeptidases, suggesting that Proteax contains an endopeptidase as well as an aminopeptidase. Therefore, the aminopeptidase was isolated and purified from Proteax by ion-exchange chromatography, subsequently characterized, and the effect of the purified aminopeptidase on the debittering of protein hydrolysates was investigated.

Supplementary Fig. 1.

Sensory evaluation of casein hydrolysate (5% w/v) treated with commercial enzymes

*, The bitterness, set at the casein hydrolysate as 100%, was evaluated by five panelists.

Supplementary Table 1. Endo and exo protease activities of each enzyme reagent (U/mg)

Enzyme reagents	Aminopeptidase		Carboxypeptidase	Endopeptidase
	Leu-pNA	Ala-pNA	Z-Gly-Phe	Casein
				pH 6.2	7.2	8.2
Proteax	541	0.500	24.4	20.5	19.0	25.3
Protease M “Amono” SD	545	25.5	163	10.0	11.0	11.2
Protease P “Amano” 3SD	548	N.D.	47.0	15.7	26.2	18.3
Protease A “Amano” SD	32.5	N.D.	42.2	13.7	24.0	22.2
Peptidase R	520	43.0	35.2	10.7	4.33	2.33
Protin SD-AY10	N.D.*1	N.D.	3.80	8.00	9.50	20.3
Neulase F3G	115	41.5	53.6	8.83	3.83	2.67
Protin SD-NY10	N.D.	7.50	1.20	10.3	9.00	10.8
Thermoase PC10F	12.0	N.D.	0.600	3.33	2.83	10.7

*1  N.D. = Not detected.

Purification and characterization of the aminopeptidase    As a possible debittering enzyme, the aminopeptidase was purified from Proteax, an enzyme reagent from A. oryzae, using the anion exchanger TOYOPEARL GigaCap Q-650M. The enzyme eluted at fraction No. 170 to No. 200 seemed to be the major leucine aminopeptidase, as shown in Supplementary Fig. 2. The data from fraction No. 0 to No. 49 in the figure are omitted because the initial stage of purification contains contaminants and is not relevant to the target. The aminopeptidase was purified to homogeneity and its molecular mass was estimated to be 37 kDa by SDS-PAGE (Supplementary Fig. 3). Fractions No. 190 to No. 192 were used as the purified enzyme.

Supplementary Fig. 2.

Purification of aminopeptidase

—, gradient of NaCl (mM); ●, A280; ■, aminopeptidase activity (A405)

Supplementary Fig. 3.

Molecular weight of aminopeptidase by SDS-PAGE

1, Crude enzyme; 2, Purified enzyme; 3, Molecular weight marker

The enzyme showed the highest activity between 60 and 70 °C and was stable up to 60 °C for 30 min at pH 7.3 (Supplementary Figs. 4 and 5). It was active and stable in a broad pH range of 7 to 9 (Supplementary Figs. 6 and 7). Some metal ions (Cu²⁺, Mn²⁺, Ni²⁺, Pb²⁺, Fe³⁺, and Al³⁺) showed an inhibitory effect on the enzyme activity. On the other hand, the addition of Fe²⁺, Zn²⁺, Li⁺, Ca²⁺, and Mg²⁺ had a marginal effect on the activity (Supplementary Table 2). The addition of EDTA·2Na reduced the enzyme activity; however, even the presence of 10 mM EDTA·2Na did not completely inhibit the enzyme activity (Supplementary Table 3).

Supplementary Fig. 4.

Optimum reaction temperature

Supplementary Fig. 5.

Temperature stability

Supplementary Fig. 6.

Optimal reaction pH

Supplementary Fig. 7.

pH stability

Supplementary Table 2. Effect of metal ion on aminopeptidase activity

Metal ion	Relative Activities (%)
Non-Ion	100
Mg2+	110
Ca2+	105
Zn2+	102
Fe2+	101
Fe3+	45.5
Ba2+	98.5
Sn2+	83.7
Co2+	76.4
Al3+	47.0
Pb2+	42.0
Ni2+	39.5
Mn2+	35.7
Cu2+	13.7
Li+	105

Supplementary Table 3. Effect of EDTA·2Na inhibitor on aminopeptidase activity

EDTA·2Na (mM)	Relative Activities (%)
0	100
0.10	70.0
0.30	66.5
1.0	59.2
3.0	39.4
10	4.62

Enzyme specificity constants were estimated by using Leu-pNA, Ala-pNA, and Glu-pNA as substrates. Km values were calculated to be 0.220 mM for Leu-pNA, 0.294 mM for Ala-pNA, and 1.73 mM for Glu-pNA. Vmax values were calculated to be 36.9 units/(min·µg) for Leu-pNA, 0.0178 units/(min·µg) for Ala-pNA, and 0.007 94 units/(min·µg) for Glu-pNA. However, enzyme activity toward Glu-pNA was about 4 650 times lower than that of Leu-pNA, and it liberated alanine from Ala-pNA at a rate about 2 080 times slower than from Leu-pNA. These results suggested that the enzyme preferentially acted on N-terminal hydrophobic amino acids of peptides.

LapA as the purified aminopeptidase    An aliquot of the purified aminopeptidase was digested by trypsin and the obtained peptides were analyzed by LC-MS/MS, followed by a database search by means of Mascot Ver2.6. MS/MS data that share an identity of 81% amino acid sequence (not shown) concluded that the enzyme is leucine aminopeptidase A (LapA) from A. oryzae (gi|2055400650|pdb|7OEZ|A, EC 3.4.11.1). The probable N-terminal amino acid sequence of LapA purified from Proteax was AVTYPDSVQHNETVQNLIK. Matsushita-Morita et al. (2010) reported self-cloning of lapA in A. oryzae RIB40. The enzyme purified from Proteax showed a molecular mass of about 37 kDa, which resembled the size of the glycopeptidase F-untreated protein from strain RIB40; however, the N-terminal amino acid of LapA recombinant protein was Tyr, which corresponded to the 4th residue of that from Proteax.

Debittering of protein hydrolysates by the purified aminopeptidase    Time courses of bitterness scores and total amounts of free amino acids of casein and cod protein hydrolysates during LapA treatment are shown in Figs. 1 and 2. In the casein hydrolysate, the bitterness decreased significantly (p < 0.01) as the treatment proceeded (bitterness score decreased from 3.5 to 1.7 after 4 h). The free amino acids increased by 97.1% during further reaction of 17 h; however, the bitterness score hardly changed (Fig. 1). On the other hand, LapA reduced the bitterness of the cod protein hydrolysate even more effectively (p < 0.01), as its bitterness score declined from 2.9 to 0.54 (most of the panelists did not report bitterness at this level) after 21 h of reaction (Fig. 2).

Fig. 1.

Relationship between bitterness and free amino acid content of casein hydrolysate by LapA treatment

●, bitterness score; ▲, free amino acid content

**, p < 0.01 vs. The bitterness of AMP 0 h treatment

Fig. 2.

Relationship between bitterness and free amino acid content of cod protein hydrolysate by LapA treatment

●, bitterness score; ▲, free amino acid content

*, p < 0.05 vs. The bitterness of AMP 0 h treatment

**, p < 0.01 vs. The bitterness of AMP 0 h treatment

The molecular weight distributions of the casein and cod protein hydrolysates, and their respective LapA-treated solutions (after 21 h treatment) were examined by gel filtration chromatography (Figs. 3 and 4). Polypeptides in the casein hydrolysate might interact with each other, as molecules with a larger size than milk casein were observed, and the ratio of free amino acids and smaller peptides increased after the LapA treatment (Fig. 3). The molecular weights of peptides in the cod protein hydrolysate were mainly distributed from about 12 000 to 1 000, and there were marginal changes in the peptide fractions when LapA was applied (Fig. 4). This result strongly suggests that LapA released amino acids at the N-terminus of peptides until encountering certain amino acids and reduced the bitter taste without digesting internal peptide bonds, which is beneficial for food applications.

Fig. 3.

Gel filtration chromatography of casein hydrolysate and LapA-treated solution solid line, casein hydrolysate; dotted line, 21 h LapA-treated solution.

Fig. 4.

Gel filtration chromatography of cod protein hydrolysate and LapA-treated solution solid line, Cod protein hydrolysate; dotted line, 21 h LapA-treated solution.

The compositions of amino acids released by LapA treatment of the casein and cod protein hydrolysates are shown in Fig. 5 and Supplementary Table 4, and Fig. 6 and Supplementary Table 5, respectively. The compositions of free amino acids of the two protein hydrolysates were similar, as hydrophobic and basic amino acids were mainly released and Thr was also released in large amounts. On the other hand, LapA might hardly act on peptides containing acidic amino acids, Glu and Asp, as well as Cys, Gly, and Pro at their N-terminus. These results suggest that LapA could liberate preferred amino acids at the N-terminus of peptides until encountering the two acidic amino acids, Cys, Gly, and Pro. Amino acids having Δf values (free energy change by transferring it from ethanol to water) greater than 1.5 kcal/mol (6.3 kJ/mol) (Tanford, 1962), such as Val, Leu, Ile, Phe, Tyr, and Lys, accounted for 74 and 64% of the total free amino acids in the LapA-treated casein and cod protein hydrolysates, respectively, resulting in a reduction of the bitter taste.

Fig. 5.

The compositions of amino acids released by LapA treatment of the casein hydrolysate

The bars in the graph show the cumulative amount of each amino acid free.

Supplementary Table 4. The compositions of amino acids released by LapA treatment of the casein hydrolysate

Amino Acid	Amounts liberated (nmol/20µL)
	0–1h	1–2h	2–4h	4–8h	8–21h	Total
Asp	0.000	0.000	0.000	0.000	0.000	0.000
Glu	0.000	0.000	0.000	0.176	0.000	0.176
Pro	0.000	0.000	0.000	0.000	0.000	0.000
Thr	0.085	0.337	2.15	4.43	7.61	14.6
Ser	0.326	0.410	1.73	3.67	5.36	11.5
His	0.000	0.000	2.12	3.60	4.17	9.89
Lys	0.095	1.44	1.97	3.63	15.5	22.7
Gly	0.118	0.026	0.150	0.415	0.873	1.58
Ala	0.264	0.359	1.54	2.96	3.99	9.11
Cys	0.000	0.000	0.000	0.000	0.000	0.000
Arg	3.73	5.66	7.39	4.29	2.59	23.7
Met	1.33	1.68	3.05	0.823	1.66	8.54
Tyr	1.43	2.84	8.49	10.0	10.1	32.8
Val	5.44	6.78	6.09	3.53	6.06	27.9
Leu	12.7	13.3	20.2	24.0	23.6	93.8
Phe	9.19	4.70	9.39	9.23	8.05	40.6
Ile	0.436	0.690	2.27	2.25	2.98	8.62
Total	35.1	38.2	66.6	73.2	92.5	305

Fig. 6.

The compositions of amino acids released by LapA treatment of the cod protein hydrolysate

The bars in the graph show the cumulative amount of each amino acid free.

Supplementary Table 5. The compositions of amino acids released by LapA treatment of the cod protein hydrolysate

Amino Acid	Amounts liberated (nmol/20µL)
	0–1h	1–2h	2–4h	4–8h	8–21h	Total
Asp	0.000	0.000	0.000	0.000	0.000	0.000
Glu	0.000	0.000	0.000	0.000	0.000	0.000
Pro	0.074	0.017	0.121	0.124	0.000	0.336
Thr	0.261	0.495	1.34	2.41	4.91	9.42
Ser	0.018	0.018	0.070	0.253	1.07	1.43
His	0.011	0.025	0.129	0.476	1.13	1.77
Lys	0.055	0.211	0.439	0.948	2.56	4.22
Gly	0.000	0.000	0.000	0.000	0.151	0.151
Ala	0.095	0.108	0.314	0.983	2.70	4.20
Cys	0.000	0.000	0.000	0.000	0.000	0.000
Arg	1.57	1.84	2.46	2.42	2.85	11.1
Met	0.378	0.450	1.04	1.52	2.35	5.74
Tyr	0.293	0.369	0.673	1.10	2.32	4.76
Val	0.581	0.982	2.50	3.70	5.37	13.1
Leu	1.75	2.06	3.49	4.31	6.14	17.7
Phe	0.676	0.603	1.05	2.56	2.44	7.33
Ile	0.973	1.23	2.38	3.58	4.86	13.0
Total	6.74	8.41	16.0	24.4	38.9	94.4

When compared to the case of the cod protein hydrolysate, prolonged LapA treatments were not so effective on the casein hydrolysate, as its bitterness persisted (Fig. 1). According to the Food Composition Databaseⁱⁱ, milk casein is rich in the above mentioned six amino acids (38.3%, w/w) and has a quite high level of Pro (10.4%, w/w). In contrast, those for cod proteins are 32 and 4.01%, respectively. Ney's hypothesis (Ney, 1971) proposed that the bitterness and average hydrophobicity (Q value, calculated from Δf value) of peptides are highly correlated. A peptide can be felt to be bitter if its Q value exceeds 1.4 kcal/mol. Consistent with Ney's hypothesis, peptides containing more branched chain amino acids, Phe, and Tyr residues generate more bitterness (Ishibashi et al., 1988a; Ishibashi et al., 1987a; Ishibashi et al., 1987b). Two sites are postulated to be involved in the binding of bitter peptides to bitter taste receptors: one is the hydrophobic residue of the peptide, which acts as the primary binding site, and the other is called the stimulatory site, which involves hydrophobic or basic amino acid residues (Ishibashi et al., 1988c). Internal Pro residues may contribute to form a conformation suitable for bitter taste receptors by changing the conformation of the peptide molecule through folding of the peptide backbone by the imino-ring (Ishibashi et al., 1988b). As hydrophobicity of peptides is one of the determinants and the number of Pro is the other, the bitterness of peptides in casein hydrolysates could not be reduced beyond a certain level. Ishibashi et al. also reported that peptides with Leu (Ishibashi et al., 1987a), and Phe and Tyr (Ishibashi et al., 1987b) at the C-terminus tasted strongly bitter. To further reduce the bitterness of bitter-tasting peptides, removal of hydrophobic amino acids from both the N- and C-termini of the peptide needs to be considered.

Conclusion

Enzymatic treatment of proteins improves their solubility, digestibility, and thermal stability. In recent years, peptides with bioactivity such as antioxidant and hypotensive effects have been discovered from protein hydrolysates and applied to functional foods (Hajirostamloo, 2010). However, enzymatic treatment of proteins produces some low-molecular-weight peptides composed of hydrophobic amino acids, which produce a bitter taste, limiting their use in foods. In this study, an aminopeptidase purified from a food-grade commercial enzyme derived from A. oryzae was used to investigate the reduction of bitterness in protein hydrolysates. This enzyme was identified as Lap A (gi|2055400650|pdb|7OEZ|A, EC 3.4.11.1) by shotgun analysis using a mass spectrometer. The three-dimensional structure of LapA in the Protein Data Bank (PDB) showed that LapA possesses Zn, and His, Asp, and Glu residues are present around Zn. The characteristics of this enzyme were roughly consistent with the experimental results of previous studies (Matsushita-Morita et al., 2011), and the analysis of the free amino acid composition newly showed that the enzyme had slightly higher specificity for Thr and Tyr and lower specificity for Ser and Cys. The bitter taste was reduced when the enzyme was applied to casein and cod protein hydrolysates, which were respectively prepared by hydrolyzing casein and cod protein with pepsin. It was inferred that this enzyme specifically hydrolyzed hydrophobic and basic amino acids at the N-terminus of the peptides and released them, resulting in a reduction of bitterness. In addition, since this enzyme was purified from a commercial food enzyme reagent, it does not require safety testing, and it can be purified relatively easily by ion-exchange chromatography without any bacterial cultivation or extraction process. Therefore, it can be easily applied to the reduction of bitterness of protein hydrolysates.

Conflict of interest    There are no conflicts of interest to declare.

References

•  Adler-Nissen,  J. (1976). Enzymatic hydrolysis of proteins for increased solubility. J. Agric. Food Chem., 24, 1090-1093.
•  Hajirostamloo,  B. (2010). Bioactive Component in Milk and Dairy Product, World Acad. Sci. Eng. Technol., 72, 162-166.
•  Hellier,  M.D.,  Holdsworth,  C.D.,  McColl,  I., and  Perrett,  D. (1972). Dipeptide absorption in man. Gut, 13, 965-969.
•  Huang,  W.-Q.,  Zhong,  L.-F.,  Meng,  Z.-Z.,  You,  Z.-J.,  Li,  J.-Z., and  Luo,  X.-C. (2015). The structure and enzyme characteristics of a recombinant leucine aminopeptidase rLap1 from Aspergillus sojae and its application in debittering. Appl. Biochem. Biotech., 177, 190-206.
•  Ishibashi,  N.,  Arita,  Y.,  Kanehisa,  H.,  Kouge,  K.,  Okai,  H., and  Fukui,  S. (1987a). Bitterness of leucine-containing peptides. Agric. Biol. Chem., 51, 2389-2394.
•  Ishibashi,  N.,  Sadamori,  K.,  Yamamoto,  O.,  Kanehisa,  H.,  Kouge,  K.,  Kikuchi,  E.,  Okai,  H., and  Fukui,  S. (1987b). Bitterness of phenylalanine- and tyrosine-containing peptides. Agric. Biol. Chem., 51, 3309-3313.
•  Ishibashi,  N.,  Ono,  I.,  Kato,  K.,  Shigenaga,  T.,  Shinoda,  I.,  Okai,  H., and  Fukui,  S. (1988a). Role of the hydrophobic amino acid residue in the bitterness of peptides. Agric. Biol. Chem., 52, 91-94.
•  Ishibashi,  N.,  Kubo,  T.,  Chino,  M.,  Fukui,  H.,  Shinoda,  I.,  Kikuchi,  E.,  Okai,  H. and  Fukui,  S. (1988b). Taste of proline-containing peptides. Agric. Biol. Chem., 52, 95-98.
•  Ishibashi,  N.,  Kouge,  K.,  Shinoda,  I.,  Kanehisa,  H., and  Okai,  H. (1988c). A mechanism for bitter taste sensibility in peptides. Agric. Biol. Chem., 52, 819-827.
•  Izawa,  N.,  Tokuyasu,  K., and  Hayashi,  K. (1997). Debittering of protein hydrolysates using Aeromonas caviae aminopeptidase. J. Agric. Food Chem., 45, 543-545.
•  Komai,  T.,  Kawabata,  C.,  Tojo,  H.,  Gocho,  S., and  Ichishima,  E. (2007). Purification of serine carboxypeptidase from the hepatopancreas of Japanese common squid Todarodes pacificus and its application for elimination of bitterness from bitter peptides. Fish. Sci., 73, 404-411.
•  Matsushita-Morita,  M.,  Tada,  S.,  Suzuki,  S.,  Hattori,  R.,  Marui,  J.,  Furukawa,  I.,  Yamagata,  Y.,  Amano,  H.,  Ishida,  H.,  Takeuchi,  M.,  Kashiwagi,  Y., and  Kusumoto,  K. (2011). Overexpression and characterization of an extracellular leucine aminopeptidase from Aspergillus oryzae. Curr. Microbiol., 62, 557-564.
•  Mizushige,  T.,  Kawabata,  F.,  Uozumi,  K.,  Tsuji,  T.,  Kishida,  T., and  Ebihara,  K. (2010). Fast-twitch muscle hypertrophy partly induces lipid accumulation inhibition with Alaska pollack protein intake in rats. Biomed. Res., 31, 347-352.
•  Ney,  K.H. (1971). Voraussage der bitterkeit von peptiden aus derenaminosaurezus- ammensetzung. Eur. Food Res. Technol., 147, 64-68. (in German)
•  Nishiwaki,  T.,  Yoshimizu,  S.,  Furuta,  M., and  Hayashi,  K. (2002). Debittering of enzymatic hydrolysates using an aminopeptidase from the edible basidiomycete Grifola frondosa. J. Biosci. Bioeng., 93, 60-63.
•  Raksakulthai,  R. and  Haard,  N.F. (2003). Exopeptidases and their application to reduce bitterness in food: a review. Crit. Rev. Food Sci. Nutr., 43, 401-445.
•  Saha,  B.C. and  Hayashi,  K. (2001). Debittering of protein hydrolysates. Biotech. Adv., 19, 355-370.
•  Song,  P.,  Cheng,  L.,  Tian,  K.,  Zhang,  M.,  Singh,  S.,  Niu,  D.,  Prior,  B.,  Mchunu,  N. P., and  Wang,  Z.-X. (2020). A novel aminopeptidase with potential debittering properties in casein and soybean protein hydrolysates. Food Sci. Biotechnol., 29, 1491-1499.
•  Tanford,  C. (1962). Contribution of hydrophobic interactions to the stability of the globular conformation of proteins. J. Am. Chem. Soc., 84, 4240-4247.
• i) https://www.ncbi.nlm.nih.gov/ (Feb. 3, 2022)
• ii) https://fooddb.mext.go.jp/ (Feb. 3, 2022) (in Japanese)