Antioxidant synergistic effects of Osmanthus fragrans flowers with green tea and their major contributed antioxidant compounds

The antioxidant synergistic effects of Osmanthus fragrans flowers with green tea were evaluated, and their major antioxidant compounds contributed to the total amount of synergy were determined. The antioxidant compounds in O. fragrans flowers with green tea were identified by LC-MS and quantified by UPLC-PDA. The synergistic antioxidant interactions between O. fragrans flowers with green tea and their antioxidant compounds were tested using the Prieto’s model after the simulated digestion. The main antioxidant compounds in O. fragrans flowers were acteoside and salideroside, whereas the main antioxidant compounds in green tea were caffeine, gallic acid, and L-epicatechin. The significant synergistic effect between O. fragrans flowers and green tea was observed and among nearly all of the combinations of their antioxidant compounds. Among the combinations, acteoside and gallic acid contributed most to the antioxidant synergy between O. fragrans flowers and green tea. However, the simulated digestion decreased this antioxidant synergy because it reduced the contents and the antioxidant capacities of their compounds, as well as the antioxidant synergy among the compounds.

However, the type of interactive effects is often determined through extremely simple relationships and through basic procedures instead of classical approaches 14 . Recently, a procedure has been described to identify and quantify the interactive effects between two antioxidants and applied to investigate interactive mechanisms in complex mixtures of antioxidants 15 . Furthermore, digestion plays an essential role in the effects of natural antioxidants in the human body. The total phenylethanoid glycoside contents and antioxidant activities in O. fragrans flowers were reported to decrease significantly after digestion 16 . Moreover, other studies showed that the total antioxidant capacity in tea extracts decreased after digestion in vitro 17 .
Thus, in this study, we aimed to measure the antioxidant synergy between O. fragrans flowers and green tea as well as among the antioxidant compounds in O. fragrans flowers with green tea using a simulated digestion model. Consequently, the major compounds which have the most contribution to antioxidant synergy were identified. We also aimed to analyse the influence of simulated digestion and examine the factors involved in this influence.

Results and Discussion
Antioxidant synergy of O. fragrans flowers and tea. The antioxidant interactions between O. fragrans flowers and each of the four kinds of tea including Longjing Tea, Pu' er Tea, Black Tea and Tieguanyin Tea were investigated using the Prieto's model. The results showed that O. fragrans flowers had synergistic antioxidant effects with all the teas (Fig. 1). Among all the combinations, the combination of O. fragrans flowers with Longjing Tea (0.23%) showed the most potent antioxidant synergy, whereas the combinations of O. fragrans with Black Tea (0.16%), with Tieguanyin Tea (0.04%) and with Pu' er Tea (0.12%) had less potent antioxidant synergies. The antioxidant compounds and their quantities in different types of tea considerably varied because of the distinct manufacturing process of each tea type 18 (Table S1). Thus, the major antioxidant compounds in O. fragrans flowers and green tea were selected for further investigation.

Main antioxidant compounds in O. fragrans flowers and green tea.
According to the mass spectra of the antioxidant compounds presented in Fig. 2A, acteoside, salidroside, isoacteoide, chlorogenic acid and caffeic acid were identified and quantified in O. fragrans flowers, with contents of 10.77, 1.51, 0.39, 0.27 and 0.08 mg/g dry weight (DW), respectively. The same compounds were determined in various O. fragrans flowers using UPLC/PDA/MS after the simulated digestion 16 . Acteoside had the highest content among the antioxidant compounds in O. fragrans flowers, and this finding was consistent with that of Lu et al. 4 .
However, EGCG, one of the most common compounds in green tea, was not detected. The reason was that EGCG underwent complex reactions during digestion and formed EC and an EGCG derivative with high molecular weight 20 . Thus, the EGCG content was below the quantification limit, as indicated in our results.  Table S2. The value of parameter m was equivalent to the EC 50 in Table 1. Among all the antioxidant compounds, gallic acid showed the highest antioxidant capacity with an EC 50 value of 3.52 μ g/mL, whereas caffeic acid had the lowest with an EC 50 value of 11.4 μ g/mL.

Individual antioxidant capacity of
Most compounds exhibited lower antioxidant capacity compared with the compounds in previous studies 21 . Meanwhile, acteoside and isoacteoside had higher DPPH radical scavenging activities compared with those in the results of Kim et al. 22 . The differences in the results were due to the simulated digestion, which resulted in the transformations and the losses of different antioxidant compounds to varying degrees.
In addition, salidroside and caffeine had poor DPPH free radical scavenging capacity, which could not fit the dose-response model well. Their antioxidant capacities were not determined by DPPH assay 23 . Therefore, they were excluded during the investigation of antioxidant interactions.

Antioxidant interactions among antioxidant compounds of O. fragrans and green tea.
The surface responses in each pair of the 45 combinations were fitted to models, as shown in Fig. 3. Their RUV values were then quantified ( Table 2). The global confidence interval of an RUV is [− 0.17%, 0.33%]; thus, any RUV value outside this confidence interval is considered as statistically consistent. On the basis of this confidence limit, 16 combinations had significant synergistic interactions.
Among the 32 combinations of the 4 antioxidant compounds in O. fragrans flowers with the 8 antioxidant compounds in green tea, most showed synergistic antioxidant interactions. Particularly, 13 combinations had significant synergistic interactions, whereas 3 combinations had significant antagonistic interactions. The combination of caffeic acid and ECG showed the most effective antioxidant synergy with RUV of 0.85%, followed by the combination of chlorogenic acid and EC (0.77%). Among the 4 antioxidant compounds of O. fragrans flowers, chlorogenic acid showed the best synergistic effect with the antioxidant compounds in green tea as it had significant synergies with 6 green tea compounds, such as EC, GC, ECG, GCG, EGCG and EGC. By contrast, caffeic acid had the most antagonistic effects with the antioxidant compounds in green tea. It showed significant antagonistic effects with EGCG, GA and GC. Among the 8 antioxidant compounds in green tea, ECG had the best synergistic interactions with the antioxidant compounds in O. fragrans flowers. It exhibited significant synergy with caffeic acid, chlorogenic acid and isoacteoside. Meanwhile, the synergies among polyphenols and flavonoids were observed in previous studies. Palafox-Carlos et al. 24 and Hajimehdipoor et al. 8 revealed the antioxidant synergistic effects among chlorogenic acid, caffeic acid and gallic acid. In addition, de Kok et al. 25    These results suggested that the antioxidant synergistic effect between O. fragrans flowers and green tea can be attained through the accumulation of the synergistic effects of all the antioxidant compounds. Figure 4 showed the major antioxidant pairs (relative contribution higher than 1%) that contributed to the antioxidant synergy between O. fragrans flowers and green tea. The combination of acteoside and gallic acid had the most contribution (35.05%) to the antioxidant synergy, followed by other pairs, namely acteoside and EC (19.17%), acteoside and EGC (11.69%) and acteoside and catechin (9.97%). The combination of acteoside and gallic acid had the most contribution after simulated digestion because gallic acid was the major product of the decomposition of tea polyphenols, such as ECG and EGCG, which were extremely unstable in neutral or alkaline conditions 26 . Furthermore, acteoside had the highest content among the compounds in O. fragrans flowers, and gallic acid had the highest content among the compounds in green tea. Their combination had an RUV value of 0.54%, which mostly contributed to the antioxidant synergy between O. fragrans and green tea.

Relative contributions of the antioxidant combinations to the antioxidant synergy in O. fragrans green tea.
Acteoside was present in all the antioxidant pairs. It had high relative contributions because it had the highest content in the O. fragrans flowers. Acteoside isolated from several plants, such as Premna serratifolia 27 and Euphrasia rostkoviana 28 , had also been shown to account for the major antioxidant effects in the in vitro assay models.

Effect of simulated digestion on the antioxidant synergy between O. fragrans and green tea and their antioxidant compounds. A significant decrease in the antioxidant synergy between O. fragrans
flowers and green tea after the simulated digestion was observed. Particularly, 32 RUV values of the 45 antioxidant combinations decreased after the simulated digestion (Table 2). Of these combinations, 15 decreased by more than 0.5%. Among the 15 combinations, 7 decreased by more than 1%. By contrast, the combinations where the RUV value increased slightly after simulated digestion did not increase by more than 0.5%.
The decrease of synergy resulted from the loss of the antioxidant compounds; moreover, their antioxidant capacities decreased (Table 1). Most phenylethanoid glycosides, such as acteoside, were typically hydrolysed or enzymolysed during simulated digestion 29 . Acteoside, which was present in all antioxidant pairs with high relative contributions, was reported to be unstable at pH 7, wherein it transformed into isoacteoside and other oxidative products; thus, its antioxidant synergy with other antioxidant compounds decreased. Similarly, some ester-type catechins, such as EGCG and ECG, undergo complex reactions during digestion to form EC and a EGCG derivative with high molecular weight, which has no antioxidant capacity 20 . Therefore, the relative contribution of the combination of acteoside and EGCG and that of acteoside and ECG to the total antioxidant synergy significantly decreased after the simulated digestion.
In conclusion, the general antioxidant synergistic effects among the antioxidant compounds accumulated and thus resulted in the antioxidant synergy between O. fragrans flowers and green tea. Among all the combinations among the antioxidant compounds, the combination of acteoside and gallic acid was a major antioxidant pair that had the most contribution to the antioxidant synergy of O. fragrans and green tea. However, the simulated digestion decreased the antioxidant synergy between O. fragrans flowers and green tea because of the decrease in the contents, antioxidant capacities and antioxidant synergies among the antioxidant compounds. Our research    Sample preparation. The O. fragrans flowers and teas were extracted using a modified method described by Xiong et al. 30 . Approximately, 1 g of the dried sample was added to 45 mL boiling water. The mixture was then stirred and extracted at 55 °C for 0.5 h. The extract was then filtered and diluted to 50.0 mL in a brown volumetric flask and stored in a refrigerator at 4 °C for the evaluation of the antioxidant activities within 30 days.
The simulated digestion was performed using a modified method described by Versantvoort et al. 31 . Saliva with α -amylase was added to a buffer system with sodium and potassium at pH of 6.8 ± 0.2. Gastric juice containing pepsin was added to another buffer system with sodium, potassium and calcium at a final pH of 2.00 ± 0.02. Lastly, duodenal juice containing parenzyme and lipase was added to a buffer system at pH 8.1 ± 0.2.
The prepared extracts (1 g each) were incubated with 3 mL of saliva for 5 min and then were mixed with 6 mL of gastric juice for 120 min. The resulting mixture was then mixed with 6 mL of duodenal juice for 120 min. All the incubations were performed at 37 °C on a rotating wheel. After the digestion, ethanol was added to ensure the   Table 2. Antioxidant interactive effects of 45 combinations of standard antioxidant compounds and extracts before and after the simulated digestion. The concentration ranges used for each case were showed in μ g for the final reaction volume (250 μ L). The degree of antioxidant interactive effects (RUV) was computed as described in Eq. (5). *Value was significantly different from the control at P < 0.05. Each sample was combined with itself, and the results were used simply as a control.  Table S6.
inactivation of the enzymes. The mixtures were then filtered by a vacuum pump. The filtrates were concentrated at 45 °C and then diluted to 10 mL with methanol. A controlled trial was operated without the prepared samples to improve the accuracy. The digesta of the extracts and antioxidant compounds were stored at 4 °C for the evaluation of the antioxidant activities within 30 days.

Identification of the antioxidant compounds by LC-MS. The antioxidant compounds present in the
O. fragrans tea were identified using a modified method described by Gruz et al. 32 . Analysis was performed on LC (ACQUITY, Waters, Milford, MA, USA) equipped with a mass spectrometre (Xevo TQ, Waters). The Zorbax XDB C18 column (2.1 mm × 150 mm, 3.5 μ m) was maintained at 35 °C. Mobile phase A consisted of purified water with acetonitrile, whereas mobile phase B consisted of 0.1% formic acid (flow rate = 0.2 mL/min). The gradient elution profiles were as follows: 8% A for 0− 3 min, from 8% to 12% A for 3-7 min, from 12% to 15% A for 7-11 min, from 15% to 20% A for 11-13 min, from 20% to 90% A for 17-19 min, from 90% to 8% A for 19-20 min and 8% A for 20-23 min. The injection volume was 1 μ L. An ESI source was operated at negativation mode. The full-scan mass spectra ranged from 100 m/z to 1000 m/z. The parameters of the mass spectrometer were set at a capillary voltage of 3 (+ ) and 3.5 kV (− ), capillary temperature of 325 °C, cone voltage of 135 V and sheath gas flow rate of 660 L/h 33 .

Quantification of antioxidant compounds by UPLC-PDA. The antioxidant compounds present in the
O. fragrans green tea were quantified using a modified method described by Trautvetter et al. 34  where A 516 is the absorbance at 516 nm. In each antioxidant, 15 increasing concentrations were used. In each concentration, 20 μ L of sample was tested using the DPPH antioxidant assay 36 . The variation of the radical scavenging activity (R) as function of the dose of the antioxidant can be described satisfactorily using the Weibull cumulative distribution function 37 . Thus, the effect of increasing concentrations of an antioxidant (A) can be described in general terms, as follows: where K is the specific antioxidant asymptotic value of the response, m is the concentration producing the half-maximal response and α is a shape parameter related to the slope. The parameter m of Eq. (2) directly provides the classical EC 50 (μ g of antioxidant), which effectively summarises all the effects (time and dose) of the response, providing the key information needed to achieve a very specific response (50%). In this instance, the lower the m value is, the higher the potency of antioxidant is.
Determination of the synergistic and antagonistic effects. The antioxidant interaction was determined using a modified procedure developed by Prieto et al. 15 . The microplate assays were performed according to a complete design for 8 × 8 arrays of the two antioxidant mixtures at equally increasing concentrations (64 independent dose combinations), which were freshly prepared. Thus, 10 μ L of each antioxidant solution was added to each well and was tested with the DPPH antioxidant assay. All the other conditions were the same as those previously described.
The dose response surface was fitted to the null interaction (NI) model  An index that summarises the possible complex effects above was described to compute the percentage RUV between the surface volume (SV) produced by the null interaction and the surface volume with interactions, as follows: Relative contribution to synergism. The relative contribution of the antioxidants to the antioxidant synergy between O. fragrans flowers and green tea was calculated using the methods modified by Palafox-Carlos et al. 38 and Lee et al. 39 . Firstly, the relative contribution of the antioxidant to the total antioxidant capacity of O. fragrans green tea (RCA) was calculated, as follows: = = % of RCA EEC /TAC, being EEC C/EC , where EEC 50 is the equivalent EC 50 which is the content (C) of the antioxidant compound in O. fragrans tea divided by its EC 50 , which is the parameter m in Eq. (2); TAC is the total antioxidant capacity of the O. fragrans tea and is the sum of the EEC 50 values of all the antioxidants present in O. fragrans tea. Thus, for a certain pair of antioxidants, its relative RUV is as follows: where RCA 1 and RCA 2 are the relative contributions of the antioxidant compounds to the total antioxidant capacity of O. fragrans tea. In the end, the relative contribution of the antioxidant pair to the antioxidant synergy (RCS) between the O. fragrans flowers and green tea was calculated as follows: where RR is the relative RUV of the antioxidant pair; TRR is the total relative RUV of all the antioxidant pairs.

Statistical analysis.
All the experiments were performed in triplicate, and the results were expressed as mean ± SD. The simulated and experimental results were adjusted to the proposed models by nonlinear least squares methods (quasi-Newton) using Matlab 2013. Student's t and Fisher's F tests were used for model consistency. The statistically significant differences were expressed as *(P < 0.05) and **(P < 0.01).