2.1 Materials

Wheat alkylresorcinols were prepared according to the method reported in our previous study (Liu et al., 2018). Wheat flour (10.5% protein, 1.7% fat, and 75.5% carbohydrates) were purchased from Beijng Er Shang Group (Beijing, China). Baker's yeast was obtained from Angel Yeast (Hubei, China). 2-Methyl-3-heptanone were obtained from Sigma Aldrich (St. Louis, Mo, USA). Nε-(carboxymethyl)lysine was purchased from J&K Scientific Ltd. (Beijing, China).

2.2 Bread preparation

Wheat breads were prepared using 100 g of wheat flour, 75 g of water, 1.5 g of dry yeast, 1.6 g of salt, 6.3 g of sugar, and 2 g of corn oil. ARs were ground into powder and mixed with wheat flour at different levels (0%, 0.03%, 0.1%, and 0.3% w/w). After being completely mixed, kneaded, and rested for 20 min, the dough was fermented for 1h and baked for 15 min at 170℃ using a bread maker (NB-H3202, Panasonic, Japan). Subsequently, the baked bread was cooled to room temperature and divided into evenly sized pieces without separating the crust and crumb for later analysis.

2.3 LC–MS analysis

The method used to determine CML was according to a previous study with minor modification (Tareke et al., 2013). Briefly, 0.1 g of bread sample was hydrolyzed by 1 ml of 6 M HCl at 110°C for 24 hr (Jiao et al., 2017) and then dried by nitrogen purge. Subsequently, dried hydrolysate was dissolved in 1 ml of water and filtered through a 0.22 μm filter before analysis using Agilent Technology 6420 Triple Quad LC-MS (Milford, MA, USA) with a Shimadzu Inertsil ODS-SP C18 column (4.6 mm × 250 mm id, 5 μm). The injection volume was 10 μL with a flow rate of 0.2 ml/min, and the oven temperature was set at 35°C. The mobile phase consisted of 0.1% FA -water as solvent A and acetonitrile as solvent B. The gradient elution program was started with 10% of solvent B at the beginning, after 1 min, increasing with a linear gradient to 40% B in 3.5 min, then held 40% B until 21mins. After that, the gradient program was returned to the initial conditions. The MRM-MS was conducted at a positive ionization mode and the optimized parameters were as following: spray voltage of 4 kV, source temperature at 350°C. For CML detection, the transitions m/z 205.0 → 84.0 (135 V, 30eV) were used for quantification.

2.4 Volatile compounds analysis

The volatile compounds in bread were extracted by headspace solid phase microextraction (HS-SPME) and analyzed using GC-MS (Trace-1300 ISQ-Mass, Thermo Fisher Scientific, San Jose, CA, USA) according to a previous report with some modification (Pu et al., 2019). Briefly, 2 g bread sample was mixed with 0.5 μL 2-methyl-3-heptanone (0.816 μg/μl in n-hexane) loaded in a 20 ml glass vial. After equilibration, the mixture was incubated in a water bath for 30 min at 80°C. Volatile compounds were adsorbed using SPME needle with a 65 μm PDMS/DVB fiber for 20 min in the headspace, followed by desorption at 250°C for 5 min in GC-MS with a TG-WAXMS Column (30 m x 0.25 mm x 0.25 µm; Thermo Fisher Scientific Inc. Waltham, MA, USA). The program of column temperature was set as follows: 40°C, held for 1 min; increased by 2°C/min to 100°C, held for 1 min; increased by 4.5°C/min to 170°C, held for 1 min; increased by 3°C/min to 220°C, held for 1 min. The carrier gas was Helium with a flow rate of 1.2 ml/min. The temperature of ion source and injection port was 250°C. The split ratio was set to splitless. The ionization mode was EI, and the electron energy was 70 eV. The mass scan range was 40–400 m/z.

2.5 Identification and quantitative analysis of volatile compounds

Identification of volatile compounds was carried out by comparing their MS data and calculation of the retention indices (RIs) with the database according to NIST (MS Search 2.0). Semi-quantification of the volatile compounds was conducted using 2-methyl-3-heptanone as the internal standard (IS). The concentration of the target compounds based on their peak area was calculated by comparing the concentration of IS and the peak area of IS. All of the compounds were quantified as 2-methyl-3-heptanone equivalents.

2.6 Odor activity value (OAV)

The contribution of the volatile compound to the overall aroma of the bread was evaluated by their odor activity value (OAV). OAV was calculated as the ratio of the individual compound concentration to their odor thresholds (Caliari et al., 2015).

2.7 Sensory evaluation

The sensory analysis was carried out as described by Souza-Borges and Conti-Silva (2018). The bread samples were cooled for three hours to room temperature before serving to an untrained sensory panel (7 males, 8 females) composing of 15 consumers. The consumers received a sample of each bread on a dish encoded with random 3-digits. Everyone should score for five sensorial parameters, including appearance, flavor, hardness, springiness, and overall acceptability. Sensory evaluation of breads was carried out using ten-point hedonic scale (1–2: dislike extremely; 3–4: dislike; 5–6: neither like nor dislike; 7–8: like; 9–10: like extremely).

A quantitative descriptive analysis was applied for evaluating bread samples. Before the quantitative descriptive analysis, a common bread was divided into evenly sized pieces and served to nine trained panelists to discuss aroma compositions of the sample, until everyone agreed to use them as the attributes. Panelists scored the intensities of the buttery, flowery, malty, sweet, and fatty odor of breads based on a 0 to 10 scale (0: not detectable; 1–2: extremely weak intensity; 3–4: weak intensity; 5–6: medium intensity; 7–8: high intensity; 9–10: extremely high intensity). The sensory scores were calculated as the average score of each attribute plus the standard deviation.

2.8 Texture profile analysis

CT3 texture analyzer (Brookfield, Massachusetts, USA) with TA4/1000 cylindrical probe was used for texture profile analysis (TPA). The instrument parameters were set according to American Association of Cereal Chemists (AACC) standard (74–09). The bread samples were cooled for 30 min at room temperature and sliced into pieces of 25 mm thickness before TPA analysis. The determinations were calculated in triplicate for each sample.

2.9 Statistical analysis

All results were represented as mean ± standard deviation (SD). Statistical significance was determined using Student's t test in instances of two experimental groups. For more than two groups, statistical evaluation of the results was performed by one-way ANOVA with Duncan's test. P-values of less than 0.05 were considered statistically significant.

3.1 Effects of ARs on CML formation in wheat bread

ARs-supplemented bread (ARs-B) at the dose of 0.03%, 0.1%, and 0.3% (w/w) could significantly decrease the formation of CML by 21.70%, 35.11%, and 42.18%, respectively, compared to the bread without addition (Figure 1). The inhibitory efficiency of ARs on the formation of CML in wheat bread sample might be ascribed to its chemical structure, the molecule backbone of which is resorcinol (1,3-dihydroxybenzene). It has been reported that flavonoids owning phenolic hydroxyl groups could bound reactive dicarbonyl species and hinder the generation of CML (Mildner-Szkudlarz et al., 2017; Yoon & Shim, 2015). Another report indicated that flavonoids and polyphenols could inhibit the formation of CML in food systems by inhibiting the oxidation of glucose and scavenging free radicals (Rice-Evans et al., 1996). Our previous study also indicated that ARs could significantly inhibit oxidative damage and reactive oxygen species (ROS) generation by improving antioxidant enzymes activities, such as catalase, NAD(P)H: quinone oxidoreductase 1 (NQO1) and heme-oxygenase-1 (Wang et al., 2019). Therefore, the inhibitory ability of ARs on CML formation may due to the elimination of free radicals, which were important intermediates of CML formation.

3.2 Sensory characterization and acceptability of ARs-supplemented bread

Bread with or without ARs supplementation was subjected to aroma profile analysis, which was performed by the evaluation of the organoleptic qualities of bread using five aromas descriptors including “buttery”, “flowery”, “malty”, “sweet”, and “fatty” (Majcher et al., 2019; Markus et al., 2017; Zhu et al., 2015). Based on the sensory evaluation scores, ARs-B exhibited predominantly buttery-like aroma characteristics compared with the control (Figure 2). Moreover, the results showed there were equal degrees of acceptability in relation to appearance, hardness, and springiness. However, ARs-B had a higher score in flavor and overall acceptability compared with control bread, suggesting ARs addition had a positive effect on the quality and flavor of wheat bread in some extent (Figure 3). Especially in Table 1, the results of correlation analysis obtained from control and ARs-B which illustrated that ARs supplement resulted in a significant positive correlation between flavor and overall acceptability. The results indicated that ARs may improve the overall acceptability of wheat bread by promoting the flavor characteristics of bread.

3.3 Effect of ARs on volatile profile of wheat bread

Total of 25 volatile compounds were detected from wheat ARs extraction (Table 2). And a total of 33 volatile compounds were detected from ARs-supplemented bread, including 9 alcohols, 3 esters, 6 acids, 8 aldehydes, 2 ketones, and 5 heterocycles (Table 3). A Student's t test was applied to investigate the effect of ARs addition on the volatile profile of wheat bread. Significant differences were found between control and ARs-B in different volatile families including alcohols, represented by 3-methyl-1-butanol, 1-hexanol, and 2-phenylethanol; ketones, represented by acetoin and 2,3-butanedione and heterocycles, represented by maltol (Table 3).

3.4 Quantitation of volatile compounds and calculation of odor activity values

Among the 33 identified and quantitated compounds, alcohols accounted for the highest proportion of volatile components both in control (3.24 mg/kg) and ARs-B (4.06 mg/kg). ARs addition significantly improved the total content of alcohols, ketones and heterocycles by 25.6%, 250%, and 317%, respectively.

The concentrations of 3-methyl-1-butanol (3, malty-like), 1-hexanol (5, flowery-like), 2-phenylethanol (9, flowery-like), 2-methylbutanal (14, malty-like), 2-(E)-octenal (17, fatty-like), 2-methylbutanoic acid (24, sweaty-like), hexanoic acid (25, fatty-like), octanoic acid (26, fatty-like), acetoin (28, buttery-like), 2,3-butanedione (29, buttery-like) and maltol (33, warmy-fruity-like) were significantly increased in ARs-supplemented bread compared with control (Table 3). Among these volatile compounds, the buttery-like smelling acetoin had the highest concentration (1,100 μg/kg), followed by 2-phenylethanol (1,020 μg/kg) and 3-methyl-1-butanol (620 μg/kg). Analysis of odor activity values (OAVs) is an effective method for the verification of the volatile-active compounds, because its evaluation criterion has taken into account the amounts of the compound and the odor threshold value (Birch et al., 2014). Therefore, in our study, OAVs of all compounds were calculated as the ratio of the volatile component concentration to odor threshold value determined in water. When OAV > 1, the volatile component was considered to contribute aroma to the sample. A total of 13 volatile components with OAV > 1 were listed in Table 4. Based on the volatile compounds with obvious difference in concentration between control and ARs-B and the results of OAV, acetoin, 2,3-butanedione, 3-methyl-1-butanol, 2-phenylethanol and 2-methylbutanal were the main volatile compounds to the aroma of bread (Table 4). The OAVs of other aroma compounds such as 1-hexanol, 2-(E)-octenal, 2-methylbutanoic acid, hexanoic acid, octanoic acid and maltol were <1, suggesting these odorants may have a minor contribution to the overall aroma.

Methyl-1-butanol and 2-phenylethanol formed by decarboxylation reactions through Ehrlich pathway during dough fermentation might contribute to the malty and flowery characteristic aromas of wheat bread (Pico et al., 2015; Sahin & Schieberle, 2019). In our study, the concentration of 3-methyl-1-butanol and 2-phenylethanol in ARs-B group was 620 μg/kg and 1,020 μg/kg, which was nearly 2.3 and 3.1 times higher than that in control bread. The result may explain the effects from quantitative descriptive analysis that ARs addition can improve the malty and flowery aroma of bread in some extent (Figure 2). Moreover, the addition of ARs significantly increased the concentration of acetoin and 2,3-butanedione, which were the metabolites derived from the fermentation process of yeast and directly linked to the yeast activity, hinting that ARs might promote yeast fermentation activity and increase fermentation productions (Birch et al., 2014; Pico et al., 2015). Acetoin and 2,3-butanedione were positively correlated with the characteristic aroma (buttery) in wheat bread and could be used as food flavoring and fragrance in bakery products due to their pleasant odor characterization (Birch et al., 2013; Di Renzo et al., 2018; Pico et al., 2015). In our present study, ARs addition significantly increased the content of acetoin and 2,3-butanedione by 3.37 and 2.17 times compared with control bread, which could partially explain the result of sensory characterization that ARs-supplemented bread pronounced predominantly buttery aroma compared with control bread. In addition, the concentration of 2-methylbutanal in ARs-B (7 μg/kg) was 1.75 times higher than that in control bread. 2-Methylbutanal was also produced by the Ehrlich pathway (Martínez-Anaya, 1996), which contributed a pleasant malt aroma in bread and could be responsible for the bread crust aroma because of its high OAV (Zehentbauer & Grosch, 1998). Taken together, the increased concentration of acetoin, 2,3-butanedione, 3-methyl-1-butanol, 2-phenylethanol and 2-methylbutanal, induced by ARs addition might be responsible for the flavor promotion of bread.

3.5 Effect of ARs on the texture of wheat bread

As shown in Figure 4, there was no significant difference on hardness, springiness, and chewiness between control bread and ARs-B, suggesting that ARs addition (3 g/kg) would not cause negative effect on bread texture. The results were consistent with the conclusion of sensory evaluation (Figure 3). However, a study from Annica et al found that ARs with a higher amount (5–10 g/kg) had a negative effect on bread texture, causing lower bread volume and baker's yeast activity (Annica Am et al., 2011). Therefore, ARs addition in this study not only enhanced the flavor profile of wheat bread, but also improved their sensory property and quality, possibly due to its lower supplementation or different homologues composition.