Effect of Maillard reaction with xylose, yeast extract and methionine on volatile components and potent odorants of tuna viscera hydrolysate

Examples of tuna viscera include the spleen, stomach, intestines, gallbladder, liver, and pancreas, which are a by-product of Global Pet Care Business Unit, Thai Union Manufacturing, Samut Sakhon. Samples were prepared by thawing frozen tuna visceral samples at 4°C for 24 h. Fresh tuna viscera were chopped into small pieces using a silent cutter. Sampling was performed for chemical composition analysis, including moisture, protein, lipid, and ash by Horwitz (2000). Amino acid contents and histamine contents were analyzed with high performance liquid chromatography (HPLC) (Latimer, 2019).

Optimal conditions for protein hydrolysate preparation were performed using Factorial (3×3) experimental planning for three factors: temperature (50°C, 55°C, 60°C), enzyme (0.5%, 1.0%, 1.5%) and time (4, 6, 8 hours). The samples and water were mixed with the ratio 1:0.5 (w/w). Add SVC 223B endo/exo protease enzyme prepared from Aspergillus oryzae (activity > 450,000 HUT u/g, Siam Victory Chemicals, Bangkok, Thailand) and hydrolyzed using a SHKE480HP/SHKE481HP incubator shaker, Thermo Scientific (Waltham, MA, USA) product, at 160 rpm, subjected to the experimental conditions, after which the reaction was stopped by heating at 98°C–100°C for 10 min, filtered and centrifuged (SUPREMA 21, Tommy Seiko, Tokyo, Japan) at 12,000×g at 4°C for 30 min. The supernatants were collected and the sample volumes equalized and stored at –20°C for further analysis. Degree of hydrolysis (%DH) was measured following Sonawane & Arya (2017) and calculated according to Equation (1) to determine optimal conditions for protein hydrolysate preparation. The tuna viscera protein hydrolysate (TVPH) sample under optimal digestion conditions was measured for the flavor components, amino acids, and histamine.

A sample (0.5 g) was hydrolyzed with 5 mL of 6 M HCl using an oil bath controlled at 110°C for 24 h. The hydrolyzed sample was diluted with Milli-Q water and then filtered through a 0.22-µm membrane filter (CA, United States). The amino acid contents of the filtrate sample were analyzed with HPLC (1200 infinity series, Agilent, Santa Clara, CA, USA) equipped with poroshell-120 HPH-C18 column (4.6 mm × 100 mm, 2.7 µm, Agilent) and quantified with o-phthaldialdehyde method. The HPLC analytical condition for amino acid analysis was modified from Latimer (2016).

The TVPH samples were flavored by Maillard reaction supplemented with xylose (Sigma Chemical, Bangkok, Thailand), methionine (Sigma Chemical) and YE (Jim’s Group, Bangkok, Thailand) as shown in Table 1. TVFE treatments were heated at 120°C for 60 min to catalyze the Maillard reaction (Savolainen et al., 2019). All TVFE treatments were analyzed for flavor components.

Volatile components were analyzed using Head Space-Solid Phase Micro-extraction gas chromatography/time-of-flight mass spectrometry (GC–TOFMS; LECO, St. Joseph, MI, USA) and the analytical condition of GC–TOFMS adapted from Pupan et al. (2018). The fiber was divinylbenzene/carboxen on polydimethylsiloxane (50/30 µm of thickness; SUPELCO, Bellefonte, PA, USA) for entrapping volatiles. The samples (2 g), salt (3 g), distilled water (8 g), and internal standard solution (2-methyl-3-heptanone) at a concentration of 10 ppm (10 µL) in a screw top vial with PTFE silicon septa were mixed with vortex mixer and then pre-incubated at 55°C for 15 min. The fiber was exposed to sample for 20 min at 55°C and then was desorbed for 5 min in GC injector. The volatiles were separated with the RTX-5 column (10 m in length, 0.18 mm in diameter, and 0.20 µm in film thickness). The temperature was set at 40°C as an initial temperature. The temperature increased to 200°C with a rate of 6°C/min, after that the final temperature went up to 240°C (10°C/min). The volatile compounds were ionized by time-of-flight mass spectrometer using energy at 70 eV, scanning range of 33–400 m/z and acquisition rate at 20 spectra/s.

Volatile compounds were identified by comparing the spectrum data and retention indices of each compound on the database (Adams, 2007; Raza et al., 2020; Wang et al., 2017). The concentration of volatiles was calculated by finding the area under the peak of each substance and compared with the area under the peak of the internal standard with known concentration and reported in ppm (Pupan et al., 2018).

OAV was calculated as volatile compound concentration divided by odor threshold (Equation [2]). This accurately reflects the potent aroma components to the overall aroma. Compounds showing an OAV ratio > 1 contributed to the odor, and their OAVs were corresponding to the odor contribution (Jiang et al., 2022).

OIS represent the flavor profile of the most common flavoring agents, calculated according to Tamura et al. (2001), by approximate Stevens’ law (Stevens, 1958). OIS is shown as follows in Equation (3).

where U_o equal to OAV; ${\text{U}}_{\text{o}}^{\text{0}\text{.5}}$ : perceived odor intensity of individual compound, ${\text{U}}_{\text{omax}}^{\text{0}\text{.5}}$ : maximum value of ${\text{U}}_{\text{o}}^{\text{0}\text{.5}}$.

Factorial experimental plan and statistical analysis was using analysis of variance and comparing mean difference by Duncan test method at 95% confidence level by the packaged program (IBM SPSS Statistics Standard version 27, IBM, Armonk, NY, USA). TVPH and TVFE volatile metabolomics data were normalized and evaluated with multivariate analysis following Settachaimongkon et al. (2014). Heat-map visualization combined with Pearson’s correlation, principal component analysis (PCA) was conducted using multivariate analysis in MetaboAnalyst 4.0 software (https://www.metaboanalyst.ca/).

The tuna viscera samples were analyzed for their proximate composition. The main component was moisture (74.65%) while the protein content was 19.12%. The high amount of protein content in this sample implied to be a suitable recourse to process the flavor enhancer with Maillard reaction. Protein hydrolysate has been confirmed as a good resource for flavor enhancers. In consideration of safety, an analysis of histamine content found that the skipjack tuna viscera contained 17.15 mg/Kg of histamine. Histamine contents are limited lower than 50 ppm in fishery raw material for human consumption (FDA, 2011); therefore, the tuna viscera were safely used as raw material for the further process.

Results of the study of optimum conditions for protein hydrolysate preparation from skipjack tuna viscera (Table 2) showed that the degree of hydrolysis (DH) increased significantly with increasing temperature (p < 0.05). The 0.5% enzyme content incubated at 60°C had a significantly higher %DH (47.8%) (p < 0.05) compared to samples incubated at 50°C and 55°C at 0.5% enzyme content. After increasing the enzyme content to 1% and 1.5%, it was found that the %DH of the sample incubated at 60°C was significantly higher. It could be concluded that incubation at 60°C was the optimum temperature.

%DH tended to decrease with increasing digestion time (4–8 hours). The optimum condition for digestion was 0.5% enzyme concentration (w/w) at digestion time four hours, temperature 60°C, obtaining %DH equal to 47.80%. While in increasing the digestion time to six and eight hours, %DH decreased to 44.28% and 44.86%, respectively. The decrease in %DH observed with the long incubation period may be due to the restriction of enzyme activity by the enzyme-mediated substrate protein degradation until the peptide formed. The decrease in %DH may be caused from the reduced digestion rate due to lower substrate concentrations, enzyme inactivation and inactivation of enzymes (Guerard et al., 2002; Zheng et al., 2013).

Enzyme contents of 0.5%, 1.0% and 1.5% had a significant effect on the %DH of TVPH (p < 0.05), while optimum conditions for protein hydrolysate production significantly affected the quality of the tuna hydrolysate production (p < 0.05). Preparation of raw materials prior to decomposition must, therefore, make the shape of raw materials suitable for enzymatic degradation, with reaction temperature and reaction time all important. Therefore, the selection of the temperature conditions for each type of enzyme degradation should consider the rate of degradation and the stability of enzymes to temperature (Chotikachinda et al., 2018).

Based on the results, regarding cost and production time, the optimum condition for producing TVPH is 0.5% enzyme concentration (w/w) at a digestion time of four hours with temperature of 60°C, and %DH 47.80%. TVPH produced under the optimum condition contained histamine as 12.80 mg/kg which did not exceed the limitation (FDA, 2011).

The amino acid profile of skipjack tuna viscera, TVPH and YE are presented in Table 3. Cysteine and methionine were the main amino acid content in TVPH and YE and these two amino acids were crucial for meat flavor formation in Maillard reaction (Raza et al., 2020; Tomé, 2021). Glutamic acid and aspartic acid were the main amino acids found in both the offal of skipjack tuna and the YE and it is also an amino acid that gives taste. Supplementing with YE is an alternative flavor enhancer such as Asp and Glu which are clearly related to umami and meat flavor. As for the amino acids such as cysteine and methionine, they help to enhance the flavor formation. It has been reported that the Maillard reaction process using cysteine or methionine as precursors when heated will react with carbonyl compounds to form a flavoring agent in the sample (Gao et al., 2020).

TVFE were TVPH heated under Maillard reaction to create new volatile components and increase some volatile compounds, which changed the flavor characteristics from the original sample. The chemical classes of volatile contents in TVPH and TVEF are shown in Fig. 1. The major volatile groups in TVPH and TVEF were aldehydes and alcohols. After Maillard reaction, aldehyde and alcohol volatile contents of TVEF (Trt A-Trt G) notably increased compared with TVPH. The next volatile class in TVPH is sulfur groups and it was found that this group also notably increased after heat treatment, especially in Trt C, D, E, F and F. Moreover, pyrazine groups were newly formed and notably found after Maillard reaction. The positively identified volatiles of TVPH and TVEF are listed in Table 4. Sixteen compounds such as 2-methylpropanal, methylpyrazine, 2,5-dimethylpyrazine, trimethylpyrazine, dimethyl disulfide and 2-acetylthaizone were created during the Maillard reaction process. Temperature causes thermal degradation of cysteine and carbonyl compounds to form volatiles containing sulfur and nitrogen classes in the Maillard products. The formation of pyrazine group compounds after heating was observed in the Maillard reaction. According to Tan & Yu (2012), pyrazine formation in the Maillard reaction is depended on temperature and reaction time. The pyrazine will not form until the reaction temperature reaches 120°C. This is consistent with these experimental results that pyrazine compounds were found when TVPH was heated at 120°C. Trt A without xylose added, while Trt B and Trt C were 1 and 2 g of xylose added, respectively. It was found that the addition of xylose significantly (p < 0.05) increased the quantity of methylpyrazine and 2,5-dimethylpyrazine in Trt B and Trt C higher than Trt A. Methylpyrazine was found in Trt A, Trt B, and Trt C as 0.007, 0.015, and 0.016 ppm respectively. Trt A, Trt B, and Trt C composed of 2,5-dimethylpyrazine contained 0.014, 0.054, and 0.061 ppm as 0.014, 0.054, and 0.061 ppm, respectively. The addition of xylose accelerated the pyrazine volatile formation via Maillard reaction in this study.

Fig. 1. Relative content of volatile compound in TVPH and tuna viscera flavor enhancer (mg/kg). TVPH, tuna viscera protein hydrolysate; Trt A, Treatment A; Trt B, Treatment B; Trt C, Treatment C; Trt D, Treatment D; Trt E, Treatment E; Trt F, Treatment F; Trt G, Treatment G.

Download Original Figure

The TVFEs in this research were classified into three groups; first group (Trt B and C) were added only xylose, second group (Trt D and E) was fortified with xylose and YE and third group (Trt F and G) was mixed with xylose, YE and methionine. The sulfur and nitrogen containing volatiles in TVFE such as methylpyrazine, 2,5-dimethylpyrazine, methional, dimethyl disulfide, dimethyl trisulfide and 2-acetylthiazole were not significantly different (p > 0.05). Li & Liu (2022), Raza et al. (2020) and Tomé (2021) reported YE and methionine could enhance the formation of sulfur and nitrogen containing volatiles via Maillard reaction. The amounts of sulfur and nitrogen volatiles of Trt D, E, F, and G in this research were not significantly different (p > 0.05) from Trt B and C which may be the limitation of the Maillard reaction due to the excessive amount of amino acid and reducing sugar in Trt D, E, F, and G. The aldehyde volatile (2-methylpropanal) not detected in Trt G may occur from the carbonyl compound of this aldehyde reacting with amino acids or sugar products to transform other compounds during Maillard reaction (Gao et al., 2020).

From the Pearson’s correlation integrated with heat-map visualization to assess the similarity of the protein hydrolysate metabolite profiles (Fig. 2), there were marked differences in the metabolite profiles among sulfurs, aldehydes, and alcohols. This observation may be relevant to the huge number of volatiles related to the analysis. The colors on the heat map relate to volatile amounts. The red color means a higher amount, and the blue color means a lower amount of volatiles among samples (Phuenpong et al., 2021). The quantity of compounds in the heat-treated TVFE in the Maillard reaction compared between Trt A (without supplements) and Trt B, C, D, E, F, and G (with supplements) showed that the sulfur and nitrogen containing compounds (dimethyl disulfide, dimethyl trisulfide, methylpyrazine and 2,5-dimethylpyrazine) in Trt A were less than those of Trt B to G. Aldehyde containing compounds increased notably when heated in Maillard reaction to form TVFE, such as 3-methylbutanal, 2-methylbutanal, and benzeneacetaldehyde. There was also a tendency to increase the formation of alcohol containing compounds (2-ethyl-1-hexanol and 1-octen-3-ol) after Maillard reaction.

Fig. 2. Heat-map visualization and hierarchical clustering of volatile compound derived metabolite profiles of protein hydrolysate. 1-TVPH, tuna viscera protein hydrolysate; 2-Trt A, Treatment A; 3-Trt B, Treatment B; 4-Trt C, Treatment C; 5-Trt D, Treatment D; 6-Trt E, Treatment E; 7-Trt F, Treatment F; 8-Trt G, Treatment G.

Download Original Figure

PCA was performed to classify the volatile profiles obtained from protein hydrolysates and shown in Fig. 3. From the interpretation of principal components (PC) in a quantitative manner, the overall PCA score plot was generated with a total variance of 60.4%. PCA showed clear separation of TVPH and Trt A (TVFE-without Maillard supplements), while Trt B, C, D, E, and G (TVFE-with Maillard supplement) samples overlapped along PC1 (45.2% variance).

Fig. 3. PCA score plots for an overall comparison of volatile compound profiles of protein hydrolysate. PCA, principle component analysis; 1-TVPH, tuna viscera protein hydrolysate; 2-Trt A, Treatment A; 3-Trt B, Treatment B; 2-Trt C, Treatment C; 4-Trt D, Treatment D; 5-Trt E, Treatment E; 6-Trt F, Treatment F; 7-Trt G, Treatment G.

Download Original Figure

To determine potent odorant in samples, OAV were applied to evaluate individual volatile components. The volatile has an OAV (Table 5) greater than 1, indicating that the volatiles contributed to its odor in the sample. From the odor threshold in Table 5, the concentrations of 1-Octen-3-ol and dimethyl trisulfide were very low (less than 0.1 ppb). Odor threshold of ٣-methylbutanal, (Z)-٤-heptanal, methional, dimethyl disulfide, and ٢-methylpropanal were proximately ٠.١–٠.٩ ppb, while ٢-methylbutanal, hexanal, benzeneacetaldehyde, and trimethylpyrazine had relatievely low thresholds between 1–3٠ ppb. These volatile compounds give the odor in tuna organ protein hydrolysates (TVPH and TVFE). Volatile compounds showing OAV > 1,000 such as 1-octen-3-ol, 3-methylbutanal and dimethyl trisulfide, contributed to meaty, chicken and chocolate-like aroma. Volatile components having OVA between 100–500 (methional, Z-4-heptenal and 2-methylbutanal) gave nutty and creamy aroma. Compounds containing OAV < 100 (2-methylpropanal, dimethyl disulfide and hexanal) gave fatty, onion, toasted aroma.

To visualize aroma profiles of sample, OIS was calculated and is shown in Fig. 4. Among the 25 volatile compounds in tuna visceral protein hydrolysate (TVPH and TVFE), 10 compounds showed odor activity above 5%. The potent volatile compounds were classified from OIS activity into three groups. First was the non-Maillard sample (TVPH) which lacked OIS in the sulfide volatiles, such as dimethyl disulfide and dimethyl trisulfide. 1-Octen-3-ol was the strongest odorant having OIS as 100% while 3-methylbutanal, (z)-4-hepteanl and methional showed OIS proximately 20% supported to TVPH aroma. Second was Maillard sample without supplements (Trt A) composed of OIS from alcohol, sulfide and aldehyde volatiles such as dimethyl trisulfide, 1-octen- 3-ol, benzeneacetaldehyde. 3-Methylbutanal was the strongest odorant containing OIS as 100% while the next potent odorant was 1-octen-3-ol (75%). Methional and dimethyl trisulfide were the next potent volatile odorants (40%). 2-Methylpropanal, 3-methylbutanal, dimethyl disulfide (z)-4-hepteanl and benzeneacetaldehyde contained OIS from 10-20% supported to Trt A aroma. Third were Maillard samples with supplements (Trt B, C, D, E, F, and G). 1-Octen-3-ol was the strongest odorant having OIS as 100%. The next potent volatile odorants were 3-methylbutanal, methional and dimethyl trisulfide having OIS from 37%–88%. 2-Methylpropanal, 3-methylbutanal, dimethyl disulfide (z)-4-hepteanl and benzeneacetaldehyde showing OIS from 10%–25% supported to the aroma profiles of these Maillard samples with supplements. OIS of volatile compounds provides a more convenient and faster procedure for the characterization of odor qualities and the determination of potential odors in various samples.

Fig. 4. OIS of volatile compound in TVPH. ¹ 2-Methylpropanal. ² 3-Methylbutanal. ³ 2-Methylbutanal. ⁴ Dimethyl disulfide. ⁵ Hexanal. ⁶ (Z)-4-heptenal. ⁷ Methional. ⁸ Dimethyl trisulfide. ⁹ 1-Octen-3-ol. ¹⁰ Benzeneacetaldehyde. OIS, odor impact spectrum; TVPH, tuna viscera protein hydrolysate; Trt A, Treatment A; Trt B, Treatment B; Trt C, Treatment C; Trt D, Treatment D; Trt E, Treatment E; Trt F, Treatment F; Trt G, Treatment G.

Download Original Figure