3-Nonanone (98%, Heowns, Tianjin, China), organic acid standards (oxalic acid, tartaric acid, malic acid, lactic acid, acetic acid, citric acid and succinic acid) were from Shanghai yuanye Bio-Technology Co., Ltd. Catechin standards (C, catechin; EC, epicatechin; GC, gallocatechin; CG, catechin gallate; EGC, epigallocatechin; CAF, caffeine; ECG, epicatechin gallate; EGCG, epigallocatechin gallate.) were from Sigma-Aldrich (Shanghai, China). Saccharomyces cerevisiae (lager yeast, s-33) was from Fermentis (Marquette-lez-Lille, France). 2,2-Diphenyl-1-picrylhydrazyl (DPPH), 2,2-azino-bis (3-ethylbenthiazoline-6-sulfonic acid (ABTS), and tripyridyl triazine (TPTZ) were from Sigma-Aldrich (Shanghai, China). All other chemicals and solvents used were of analytical grade, from local suppliers.

HOPB and JTB were produced in a 300 L pilot-scale brewing plant (Zhejiang Gongshang University). Pilsner-type (75 kg) and wheat (25 kg) malt was crushed using a two-roll mill and then transferred to a stainless-steel mashing vessel and mixed with water (45°C, 300 L). The process was initiated with the mash-in at 45°C (20 min), and the temperature was then gradually increased and maintained at 50°C (30 min), 65°C (40 min), 72°C (20 min) for maltose saccharification and 78°C (10 min) for enzyme inactivation, respectively. After complete mash conversion, the sweet wort was separated from the spent grains by lautering, then transferred to the kettle for boiling. The wort was boiled for 110 min with the addition of hops pellet (100 g, alfa-acid proportion: 13%) after 30 min and jasmine tea extract (360 g) after 100 min. The coarse break was separated through settling and the wort was cooled to about 40°C and transferred to the fermentation tank where the brewing yeast (150 g/L, activated for 30 min in sterile water) was added. After 7 days of primary fermentation at 20°C, the yeast slurry is drained from the bottom of the fermenter, followed by a temperature adjustment to 4°C and a closed venting valve for post-fermentation and maturation for 60 days, promptly sampling, storage (−80°C) and analysis. The control samples were brewed in the same way except for the addition of different ingredients at the late hopping stage. For the control sample, 150 g (alfa-acid proportion: 13%) of hops were added after 100 min of boiling.

Beer analysis was performed after maturation, following procedures in Chinese standard GB/T4928-2008. Color was determined by the spectrophotometric method (5.6.2) (Unico-2000, Unico, Shanghai, China); foam stability with a foam measuring cup (7.2); total acid content by titration with 0.1 mol/L NaOH; ethanol concentration was determined by quantitative distillation according to Dietz et al. (12). The alcohol concentration was determined with an SBA-40E biosensor (Shandong Biosensor Institute, China). Turbidity was determined using a HACH-TL2300 Turbidity Meter (HACH, Shanghai, China, detection limit = 0.001 NTU). A pH meter (FivrGo-2, Mettler Toledo, Shanghai, China) was used for pH measurements.

HPLC was used to analyze some targeted components, such as GA, GC, EGC, C, EC, EGCG, GCG, ECG, and caffeine, as described previously (22, 23). The HPLC (Agilent Technologies, Santa Clara, CA) was equipped with an Infinity binary pump, an autosampler, a column thermostat (set at 30°C), a diode array detector and an Agilent Zorbax SB-Aq C18 column (250 × 4.6 mm i.d., 5 μm). The mobile phases were 0.2% v/v aqueous formic acid (A) and methanol (B). The initial solvent was 5% B, which was ramped linearly to 20% B at 5 min, 25% B at 18 min, 42% B at 25 min, held for 7 min, then increased to 100% B at 40 min. The total run time was 40 min, the flow rate 1.0 ml/min, the injection volume 5 μl and the detection wavelength 278 nm.

The analysis was performed as described previously (24). Bottles of beer were maintained at 4°C to minimize loss of volatiles. Beer sample (4 ml), water (4 ml), internal standard (3-Nonanone, 20 μl, 9.8 mg/L) and NaCl (1 g) were added to 20 ml SPME headspace vials and sealed with a polytetrafluoroethylene (PTFE)-silicon septum. The septum covering the vial headspace was pierced with the needle containing the SPME fiber and retracted, and the fiber was exposed to the headspace for 30 min at 50°C, then inserted directly into the GC-MS injection port. The carrier gas was helium at a flow rate of 1 ml/min. Samples were analyzed on a DB-5MS UI column (30 m by 0.250-mm inside diameter by 0.25-μm film thickness, Agilent). The oven temperature was programmed as follows: initial temperature 35°C, held for 5 min, increased at 4°C/min to 130°C, held for 3 min, then at 5°C/min to 230°C, held for 5 min. Electron impact (EI) ionization was used at 70 eV, scanning from m/z 10 to 250. Background subtraction was performed on the raw GC-MS data using data processing software. The National Institute of Standards and Technology (NIST 14) database was used for qualitative analysis of the MS peaks corresponding to the chromatographic peak signals at different retention times (>70%). The peaks were quantified by comparison with the internal standard (3-Nonanone) to calculate the relative content of each substance and the data were imported into Simca-P software (Version 14.1, MKS Umetrics AB, Umeå, Sweden). The outputs were subjected to Orthogonal partial least squares-discriminant analysis (OPLS-DA). The variable importance in projection (VIP) value was used to evaluate data generated by OPLS-DA; only data with VIP values >1 were selected for further analysis.

Non-targeted metabolomics analysis was carried out using UPLC-HRMS (Q-Exactive system, Thermo Fisher Scientific, Rockford, IL), as described previously (25) with some modifications. The column was an Acquity UPLC BEH C18 (100 nm × 2.1 mm; 1.7 μm, Waters, Manchester, UK). The mobile phases were aqueous, 0.1% v/v formic acid (A) and acetonitrile (B), and the linear elution gradient program was 0–1.0 min, 5% B; 2.0 min, 10% B; 6.0 min, 35% B; 8.5–9.5 min, 100% B; and 10.0–12.0 min, 5% B. The total analysis time was 12 min and the flow rate was 0.3 ml/min. The column and autosampler were set at 40 and 10°C, respectively.

Mass spectrometric analysis was performed with the Q-Exactive Orbitrap mass analyzer in negative ionization mode at a spray voltage of 3.0 kV. The capillary temperature and auxiliary gas heater temperature were both 300°C. The flow rates of sheath gas and auxiliary gas were set to 45 and 10 arbitrary units, respectively. The full scan MS/data-dependent MS/MS (ddMS2) mode was used, in which the resolution was 70,000 and 35,000 for full MS and ddMS2, respectively. The mass scan range was from m/z 66.7 to 1000.

All the samples were filtered through a 0.45 μm Millipore filter. The raw data acquired were processed on Compound Discoverer software (Version 3.0, Thermo Fisher) to obtain all the ion fragment information through peak picking and alignment. This information was used for partial least-squares discriminant analysis (PLS-DA) to screen for differential metabolites with VIP >1.2 and p < 0.05, which was performed on Simca-P v14.1. Thereafter, the Human Metabolome Database (http://www.hmdb.ca/), our laboratory's standards library and previous metabolomics studies (26, 27) were used for identification of the differential metabolites.

The total phenolic content of beer was determined spectrophotometrically with Folin–Ciocalteu reagent (28). A calibration curve was plotted using gallic acid as standard. The beer samples were diluted with deionized water to adjust the concentration of phenolic compounds to the linear calibration range of gallic acid. Results were expressed as mg of gallic acid equivalent (GAE) per liter.

α-Amylase inhibition was determined as described previously (33), with minor modifications. Test samples (200 μl) were added to sodium phosphate buffer (300 μl, 0.02 M pH 6.9) containing 30 unit/ml porcine pancreatic α-amylase and preincubated at 37°C for 10 min. After preincubation, starch solution (300 μl, 0.5%) was added, then the reaction mixtures incubated at 37°C for 15 min. The reaction was stopped by adding 3,5-dinitrosalicylic acid reagent (600 μl), heating in a boiling water bath for 10 min, then cooling. After adding 900 μl of distilled water, the absorbance was measured at 540 nm. The α-amylase inhibitory activity was calculated as follows:

The α-glucosidase inhibition was determined as described previously (34) with minor modifications. α-Glucosidase (100 μl, 70 U/ml, in 0.5 ml sodium acetate buffer, pH 5.0) was premixed with beer sample (0.5 ml) and incubated at 37°C for 15 min. p-Nitrophenyl-α-D-glucopyranoside (0.5 ml, 2.5 mmol/L) was added, then the mixture was incubated at 37°C for 15 min and stopped by adding sodium carbonate solution (1 ml 0.2 mol/L). The α-glucosidase activity was determined by measuring the release of p-nitrophenol at 405 nm. α-Glucosidase inhibition was calculated as follows:

The beers were evaluated by 29 untrained tasters from the Tea Research Institute, Chinese Academy of Agricultural Sciences, aged between 20 and 59 years. For each taster, a 40 ml sample of the beer was served in a disposable, clear, acrylic glass. The tasters evaluated the aroma, taste, foam, appearance and overall score using a nine point hedonic scale form, where 1 = dislike strongly; 5 = neither like, nor dislike; and 9 = like strongly (35). A second four level scale (not sensed, faintly sensed, mildly sensed, and strongly sensed) was used to grade bitter and astringent tastes, and malty aroma, fruital aroma, floral aroma, and fresh-scent aroma. Training of tasters in grade evaluation prior to the experiment. All the participants (healthy and non-smokers from TRICAAS) were conducted considering the principle outlined in the Declaration of Helsinki, and informed written consent was obtained.

The data are presented as the mean ± standard error of the mean. All experiments were carried out in triplicate. The results were analyzed with SPSS 18.0, using one-way analysis of variance to determine differences between sample groups, with p < 0.05 considered statistically significant.

The effects of JTE on the physicochemical properties of the beers were compared (Table 1). JTE addition did not affect (p > 0.05) the color, turbidity and diacetyl content, however, it affected (p < 0.05) the contents of alcohol, total acids, total polyphenols, total catechins, and the foam-stability. The foam stability, alcohol concentration, total polyphenol concentration, and total catechin concentration of JTB were 528.0 ± 22.7 s, 6.6 ± 0.14% (v/v), 921.95 ± 1.6 mg/L, and 302.4 ± 1.3 mg/L, respectively, 1.52-, 1.04-, 1.6-, and 34.72-fold those of HOPB, respectively.

The catechin concentration of JTB was markedly higher than that of HOPB (Table 2), because of the high catechin concentration of JTE (17), and since catechins are the major JTE polyphenols, the polyphenol content of JTB was also significantly higher than that of HOPB. The foam stability of JTB was significantly higher, which may be related to its greater polyphenol content (18) and the foam-stabilizing effect of some polyphenols (36). Studies have shown that the type and content of polyphenols and catechins in beer are important factors affecting the antioxidant capacity and flavor stability of beer (37). In this study, it was found that beers with added tea extracts had stronger antioxidant capacity. Taken together, these results suggest that there is an association between JTE addition and beer quality.

Organic acids are important indicators of product quality and contribute to the organoleptic properties of beer, as well as being indicators of fermentation performance, so the organic acids (oxalic, tartaric, malic, lactic, acetic, citric, and succinic acid) in JTB and HOPB were determined (Figure 1). The total organic acid content of JTB (1445.38 mg/L), was 1.52-fold that of HOPB (p < 0.05). The main organic acids in both beers were lactic, acetic and citric acid and they differed significantly. The lactic, acetic and citric acid contents of JTB were 288.88 ± 1.02, 288.75 ± 7.34, and 356.44 ± 11.03 mg/L, respectively, 1.48-, 1.58- and 2.43-fold those of HOPB, respectively. The malic and succinic acid contents of JTB were 173.03 and 240.56 mg/L, 1.78- and 1.21-fold than those of HOPB, respectively. The tartaric acid content of JTB was significantly lower (0.44-fold) that of HOPB. The oxalic acid contents were similar, mainly because oxalic acid is primarily derived from the wort (38).

After LC-MS data preprocessing, a total of 1,113 compound ion features were obtained from univariate and multivariate analysis. The QC samples were closely grouped in the PCA scores plot, indicating that the metabolomic analysis was reliable (Figure 4A). The JTB and HOPB samples were both closely grouped, but the two beer type groups were well separated, i.e., LC-MS analysis could clearly distinguish the two beer types. The criterion of PLS-DA VIP value ≥1.2 (Figure 4B) was used to screen for compounds that significantly differed between the two beers, and identified 123 differential compounds (Figure 4B, purple).

The 123 compounds were initially identified on the basis of their exact molecular masses and fragmentation spectra. Thirty-nine differential compounds were identified by comparing with the HMDB database (http://www.hmdb.ca/), laboratory standard libraries and previous reports (Table 4), namely, six bitter acids and derivatives, three amino acids, five phenolic acids, five organic acids, seven dimeric catechins, nine flavone/flavonol glycosides, and four others. Overall, JTE addition resulted in significant changes in metabolic profile.

A heatmap was plotted to visualize the differential metabolites resulting from JTE addition (Figure 5). The content of phenolic acids, amino acids, dimeric catechins and flavone/flavonol glycosides in JTB was significantly higher than that in HOPB, whereas bitter acid and some organic acids were less abundant in JTB. These differences are consistent with the substitution of late hopping with the addition of JTE, as amino acids, dimeric catechins, flavone/flavonol glycosides and phenolic acids are abundant in tea, whereas bitter acids are only found in hops (47–50). Dimeric catechins, amino acids and flavone/flavonol glycosides from the JTE would give the beer a tea-like flavor, especially the theanine, which has an umami taste (51). The high content of iso-alpha-acids is the main reason for the bitterness of beer (52), which is consistent with the sensory evaluation results (see below). HOPB had a higher bitterness intensity.

Taste, flavor and other sensory attributes are the main determinants of beer quality (53). The appearance, foam, aroma, taste, and overall acceptability of the two beers were compared by sensory evaluation (Figure 6). There were differences in flavor intensity between the beers, with HOPB having a higher bitterness intensity and JTB having a higher floral and light fresh-scent intensity (Figure 6A); the aroma score for JTB was 8.8, 31.0% higher than that of HOPB. The aroma differences are consistent with the relative volatile profiles of the beers (section 3.2), i.e., JTB contained higher concentrations of linalool, α-terpineol and citronellol. The score for taste of JTB was 7.8, 32.5% higher than that of HOPB, whereas HOPB scored higher for bitterness. JTB had a richer, whiter and finer foam, with a score of 7.7, compared with 7.0 for HOPB. However, JTB scored lower for appearance than HOPB, because JTB was a little more turbid than HOPB. Overall, JTB had a higher organoleptic rating than HOPB. The taste of JTB was mellower, softer and with a pleasing tea flavor (Figure 6B). JTB had a higher aroma score than HOPB (p < 0.05), apparently because of the abundant floral and fresh fragrances released from JTE (54, 55). HOPB had a lower taste score and higher bitterness score than JTB, probably resulting from the late hopping of HOPB; late hopping enhances the bitterness of beer (56). The study of Oladokun et al. (57) showed a significant effect of polyphenol content on the perceived intensity and characteristics of bitterness, with higher polyphenol content resulting in stronger bitterness and poorer bitterness characteristics expression in beer. It is noteworthy that although the beer with tea extract added in this study had higher polyphenol content, the bitterness intensity of JTB did not become stronger.

The antioxidant capacity influences the functional properties and the oxidation resistance of the beer during storage (30). The DPPH and ABTS+ radical scavenging capacities, and ferric reducing antioxidant power (FRAP) were determined to compare the antioxidant capacity of the beers (Figure 7). The DPPH, ABTS+, and FRAP capacities of JTB were 80.9, 42.3, and 50.3 mg/L, respectively, 52.6, 28.8, and 47.7% higher (p < 0.05) than those of HOPB, respectively.

The higher antioxidant capacities of JTB are consistent with its 1.6-fold higher phenolic content than HOPB; polyphenols, flavonoids, and flavonols account for the antioxidant capacity of beer (58, 59). Similarly, addition of fresh fruits during beer fermentation significantly enriched the content of phenolic compounds and increased the antioxidant capacity of the beer (60). Increased antioxidant capacity improves the storage stability of beer (36).

α-Glucosidase and α-amylase are the key enzymes in the digestive system that hydrolyze dietary carbohydrates. Inhibition the two amylase can reduce and control postprandial blood glucose spikes, delaying hydrolysis of carbohydrates and suppressing postprandial hyperglycemia in prediabetes, diabetes, and obesity patients (61). The inhibitory effect on these digestive enzymes of the beers was determined (Figure 8); both beers inhibited α-amylase activity by 42.5 ± 1.5% (HOPB) and 72.1 ± 1.0% (JTB). α-Amylase inhibition by JTB was 30.5% higher than that of HOPB, probably because of its higher phenolic content; tea catechins strongly inhibit α-amylase activity (62, 63) and gallocatechin gallate is the strongest inhibitor among the catechins (61). HOPB inhibited α-glucosidase by 50.3%, whereas JTB had no significant effect. The characteristic compounds in hops inhibit both α-amylase (64) and α-glucosidase activity (65–67), which is consistent with the inhibition of both enzymes by HOPB.