The TP contents varied from 1,129.8 to 2,710.4 mg GAE/L, for the Cabernet Sauvignon wines and from 860.2 to 1,656.5 mg GAE/L, for the Merlot wines. The Cabernet Sauvignon and Merlot wines with the highest amount of TP were both from the NXYQY region, whereas both single varietal wine samples tested in the SXXN region had the lowest values. The content of TP decreased in the order: NXYQY > HBSC > HBCL > SXXN regions for Cabernet Sauvignon and Merlot wines. For Cabernet Sauvignon samples, the TP content in NXYQY region was nearly 2.0~2.4 times those in the SXXN and HBCL regions, being approximately 1.2 times that in the HBSC region. For Merlot samples, the content of TP in the NXYQY region was 1.7~1.9 times those in the SXXN and HBCL regions, whereas the TP content in the HBCL region was non-significantly higher than that in the SXXN region.

The content of TFO varied from 859.9 to 2,290.3 mg CTE/L, for the Cabernet Sauvignon wines and from 660.6 to 1,375.4 mg GAE/L, for the Merlot wines. The TFO contents of Cabernet Sauvignon and Merlot wines in the NXYQY region were significantly higher than those of the other three regions. In addition, the Merlot wines from the HBCL and SXXN regions contained non-significantly lower content of TFO. The content of TFO decreased in the order: NXYQY > HBSC > HBCL > SXXN for Cabernet Sauvignon wines and NXYQY > HBSC > SXXN > HBCL for Merlot wines, respectively.

The content of TFA varied from 277.7 to 666.4 mg CTE/L, for the Cabernet Sauvignon wines and from 272.8 to 497.4 mg GAE/L, for the Merlot wines. The TFA content of Cabernet Sauvignon wines in the NXYQY and HBSC regions were more than double that in the SXXN region, and its content in Merlot samples in the NXYQY region was more than 70% those in the HBCL and SXXN regions. The content of TFA decreased in the order: HBSC > NXYQY > HBCL > SXXN (Cabernet Sauvignon wines) and NXYQY > HBSC > SXXN >HBCL (Merlot wines), respectively.

The grape pigments or anthocyanins are present in red grapes only and are largely responsible for the color of red wines. The content of TA varied from 261.5 to 400.3 ME mg/L, for the Cabernet Sauvignon wines and from 157.5 to 350.3 ME mg/L, for the Merlot wines. The TA contents of two single varietal wines in the NXYQY region were the highest, and the contents in the HBCL region were the lowest. In order to identify the phenolic contents of Cabernet Sauvignon and Merlot wines from four wine grape-growing regions, we chose all the raw materials (grape berries) and wines had the same situations, including the same vintage and cultivation management, the same winemaking techniques and ageing. The results confirm a variation in phenolic content among wine samples tested. As is well known, the amounts of phenolic materials vary considerably in different wine grape-growing regions, depending on the grape variety, environmental factors of vine growth [8]. On the whole, in the present study, either for Cabernet Sauvignon or for Merlot, the NXYQY wine samples contained significantly more contents of phenolics than did the other regional wines, while on the contrary, its contents in SXXN and HBCL regions were the lowest. The high phenolic contents indicated that the NXYQY wines had potential antioxidant activity. In the study of Li et al. [9], the TP contents of Cabernet Sauvignon and Merlot wines were reported. The TP content varied from 1,580 to 2,927 mg CTE/L for the Cabernet Sauvignon wines, its content in the Merlot wines varied from 1,977 to 2,246 mg GAE/L. Considering the fact that there exist differences of environmental conditions, winemaking techniques and aging, these values are analogous to the values found in this study for Cabernet Sauvignon and Merlot wines.

Analysis of DPPH• scavenging activity is one method for detecting antioxidant activity and is commonly used to evaluate the ability of free radicals to react with antioxidants [10,11]. In the DPPH• scavenging assay, antioxidants reacting with DPPH• produce yellow α,α-diphenyl-β-picrylhydrazine. The degree of discoloration indicates the radical scavenging activity of the antioxidant [12]. The free radical-scavenging activity of all the samples was determined by the DPPH• methods and the results were shown in Figure 1(a). For DPPH•, the values of wines varied from 4,670.4 to 6,179.4 μM TE/L for the Cabernet Sauvignon wines, and 3,857.0 to 5,304.1 μM TE/L for the Merlot wines. The values of DPPH• decreased in the order: XNYQY > HBSC > SXXN > HBCL (Cabernet Sauvignon) wines and XNYQY > HBSC > HBCL > SXXN (Merlot wines), respectively. The results of investigation show that the higher the concentration of antioxidant, the lower is the amount of remaining DPPH• and the higher is the free radical-scavenging activity.

The reducing power property indicates that the antioxidant compounds are electron donors and can reduce the oxidized intermediates of the lipid peroxidation process. In this research, we use the CUPRAC assay which is based on the reduction of Cu(II) to Cu(I) by antioxidants. This assay is superior to the widely used ferric ion reducing antioxidant power (FRAP) assay. Firstly, the reaction conditions are more adaptable to real body systems; secondly, it can detect the reducing ability of thiol compounds [9]. All analyzed wine samples demonstrated significant antioxidant capacity in the CUPRAC test [Figure 1(b)]. For CUPRAC, the values of wines varied from 10,227.2 to 19,632.8 μM TE/L for the Cabernet Sauvignon wines, and 9,194.1 to 17,693.0 μM TE/L for the Merlot wines. The CUPRAC of the NXYQY Cabernet Sauvignon and Merlot wines with the highest values was nearly 1.9 times higher than those of the SXXN Cabernet Sauvignon and the HBCL Merlot wine, respectively. Not significant differences were also observed in the SXXN and HBCL Merlot wines. The values of CUPRAC decreased in the order: NXYQY > HBSC > HBCL > SXXN Cabernet Sauvignon wines and XNYQY > HBSC > SXXN > HBCL Merlot wines.

The superoxide radical is one extremely reactive free radical formed in biological systems and has been implicated as a highly damaging species in free radical pathology, capable of damaging almost every molecule found in living cells [13]. The superoxide radical-scavenging activities of all the wine samples were shown in Figure 1(c). In this research, the Cabernet Sauvignon and Merlot samples exhibited from 59.9% to 68.5% and 62.6% to 66.2% superoxide radical-scavenging activity, respectively. The inhibition of superoxide radical decreased in the order: XNYQY > HBSC > HBCL > SXXN (Cabernet Sauvignon) wines and XNYQY and HBSC > SXXN > HBCL (Merlot wines).

Five flavan-3-ols were observed in four regional wines by HPLC-ESI-MS/MS (Table 2), including three monomers (catechin, epicatechin and gallocatechin) and two dimers. With the exception of gallocatechin, which was only detected in the NXYQY Cabernet Sauvignon wines and the HBSC Merlot wines, the wine samples from four regions contained the same composition of four flavan-3-ols. For the Cabernet Sauvignon wines, the wine sample with the highest content in five flavan-3-ols was that from the NXYQY region, except for procyanidin dimer B2, and the wine from SXXN region showed the lowest levels of five flavan-3-ols. The contents of five flavan-3-ols in HBSC and HBCL regional wines varied among these regions, being comparable. In more detail, the contents of catechin, epicatechin and procyanidin dimer B2 in the NXYQY “Cabernet Sauvignon” wines were almost 8, 14 and 34 times those in the wines from the SXXN region.

For the Merlot wines, the HBSC regional wine had the highest level of catechin and epicatechin, which were almost two times that in the wines from the SXXN and HBCL region, and nearly three times that in the SXXN wine sample. The interregional differences were also present in the contents of the procyanidin dimers.

Flavonols are yellow pigments that mainly exist as the 3-O-glycoside of four main aglycones: myricein, quercetin, kaempherol and isorhamnetin [14]. They generally act as UV protectors [15,16] and co-pigments of anthocyanins in flowers and fruits [17,18]. As shown in Table 2, there were 13 flavonols identified in all wines, and the composition of flavonols varied widely with geographical origin. Compared to the other three regions, the number of flavonols detected in HBSC Cabernet Sauvignon and HBCL Merlot wines was the least, respectively. Only quercetin-glucuronide, dihydroquercetin-3-O-rhamnoside, laricitrin-3-O-glucoside and myricetin-3-O-galactoside were common in the Cabernet Sauvignon wines of four regions, besides including the former three flavonols, myricetin-3-O-glucoside was common in all the Merlot wines. The greatest range of flavonol compounds was found in the NXYQY wines. Except for myricetin-3-O-galactoside and common in the Cabernet Sauvignon wines of four regions, besides including former three flavonols, myricetin-3-O-glucoside was common in all the Merlot wines. The greatest range of flavonol compounds was found in the NXYQY wines. Except for myricetin-3-O-galactoside and myricetin-3-O-glucoside, the contents of other flavonols were the highest in the NXYQY region, as shown in Table 2.

The phenolic acids could be simply classified into two groups: hydroxybenzoic acid and hydroxycinnamic acid. There were four hydroxybenzoic acids and three hydroxycinnamic acid identified in all wines (Table 2). As shown by this data, the contents and composition of phenolic acids varied widely in wine samples of different regions. Only dimer(epi)gallocatechin-(epi)catechin and syringic acid were common in the Cabernet Sauvignon and Merlot wines of all four regions. Syringic acid is one of the most important hydroxybenzoic acids in wines. Its contents showed significant differences in the wines from different regions. The NXYQY wines had the highest level of syringic acid while the wines from SXXN presented the lowest level. Among various hydroxycinnamic acids, the content of trans-fertaric acid showed significant differences in the different regions. The NXYQY Cabernet Sauvignon wines had nearly six times more trans-fertaric acid than the HBCL wines.

Since stilbenes are phytoalexins and their existence in grapes is directly related to environmental stress, such as botrytis infections and UV-irradiation [19], healthy plants contain small amounts. The trans-piceid was only found in all Merlot wine samples, amounts were comparable with the concentration range found in the literature [8]. Wine samples from the HBSC region have the highest amount of trans-piceid.

The phenol profile in a wine comes primarily from the grape berries, but the polyphenolic composition of wines is more complex as compared to their corresponding grape berries, because of the numerous reactions involving phenolic compounds that occur during the wine making and maturation processes [14]. In the present study, we chose all the raw materials (cultivation management) and wines (winemaking techniques and ageing) to have the same situations, yet the four regional wines displayed different qualitative and quantitative profiles, attributed to the unique soil and climatic conditions of the four regions and are indicative of the “terroir” influence on a wine’s polyphenolic profile. It must be noted that these individual phenols contents with different composition were all under 10.0 mg/L. The results were partially inconsistent with the results in some previous studies [8,20] showing that a similar composition in phenolic compounds was observed in different regional wines due to the biosyntheses of phenolic compounds depending largely on the genotypes of the grape cultivar rather than the environmental factors. Downey et al. [21] also suggested that flavonoid biosynthesis in plants is affected by many factors, including light, temperature, altitude, soil type, water, microbial interactions and nutritional status. Our finding was in accordance with these studies on polyphenolic composition involving wines of different geographical origins [22]. The NXYQY wines had the highest level of individual phenols, especially for four flavan-3-ols and quercetin-3-O-glucuronide, dihydroquercetin-3-O-rhamnoside, laricitrin-3-O-glucoside and isorhamnetin-3-O-glucoside. Price et al. [23] and Spayd et al. [24] have suggested that there was a positive correlation between sunlight exposure and an increase in flavonols. Previous researches showed that light, water deficits and higher temperature differences between daytime and nighttime could significantly increase the contents of flavonoids [25,26]. In addition, both the HBSC and HBCL regions that are geographically close to each other have similar individual phenol contents in Cabernet Sauvignon wines.

The individual phenolics (including flavan-3-ols, flavonols, hydroxybenzoic acids, hydroxycinnamic acids, and stilbenes), a wine sample with 100 mL was diluted with the same volume of distilled water, and then extracted thrice with ethyl acetate (80 mL). The combined ethyl acetate phase was removed by a rotary evaporation at 28 °C and the remainder was resolved in methanol (chromatography grade) up to a final volume of 5 mL. The final sample was filtered by 0.22 μm organic membranes prior to analysis by HPLC-MS/MS.

Qualitative and quantitative analyses of individual phenolic compounds were performed in Center for Viticulture and Enology, College of Food Science & Engineering, China Agricultural University. Based on the optimum HPLC-ESI-MS/MS conditions established for the individual phenolic compounds using standard solutions we further combined retention time with (molecular and fragment) ions information to qualitatively analyze these individual phenolic compounds detected [42]. For the individual phenolic compounds, an Agilent 1200 series instrument, equipped with a G1322A Degasser, a G1312B Binary pump, a G1367C HiP-ALS, a G1316B TCC, a G1314C VWD, and a ZORBAX SB-C18 column (3 × 50 mm, 1.8 μm) were used. The mobile phase comprised of (A) 10% (v/v) acetic acid, and (B) 90% (v/v) acetonitrile containing 10% (v/v) acetic acid. The elution gradient was from 5% to 8% B for 5 min, from 8% to 12% B for 2 min, from 12% to 18% for 5 min, from 18% to 22% for 5 min, from 22% to 35% for 2 min, from 35% to 100%B for 2 min, 100% B for 4 min and from 100% to 5% B for 2 min, with a flow rate at 1.0 mL/min. The injection volume was 2 μL, and the detection wavelength was 280 nm. The column temperature was 25 °C. The MS conditions were as follows: ESI, negative ion model; nebulizer, 35 psi; dry gas flow, 10 L/min; dry gas temperature, 325 °C; scan, 100–1,000 m/z.

Flavanols, flavonol, hydroxybenzoic acid, hydroxycinnamic acids, and stilbenes were quantified using catechin, quercetin, gallic acid, caffeic acid, and resveratrol as standards, respectively. The gradience with eight concentrations of mixture standard (catechin, quercetin, gallic acid, caffeic acid, and resveratrol) were set with three replications, five groups of standard curves of the concentration were made based on the average area of individual compounds, and five polyphenol compounds were qualified by the external standard method.

The ability to scavenge DPPH• free radicals was determined. Scavenging activity was based on the slightly modified method [43]. Briefly, 0.1 mL of sample solution (with appropriate dilution if necessary) was added to 3.9 mL of a 60 μM solution of DPPH• in methanol. A control sample, containing the same volume of solvent in place of extract, was used to measure the maximum DPPH• absorbance. After the reaction was allowed to take place in the dark for 30 min, the absorbance at 515 nm was recorded to determine the concentration of remaining DPPH•. Results were expressed as Trolox equivalent antioxidant capacity.

The cupric reducing antioxidant capacity was determined according to the method of Apak et al. [44]. Results were expressed as Trolox equivalent antioxidant capacity.

The method used by Robak and Gryglewski [45] for determination of superoxide anion scavenging activity was followed after some modification. The superoxide radical was generated in 3 mL of Tris-HCl buffer (20 mM, pH 8.3) containing 1 mL of NBT (150 μM) solution, 1 mL of NADH (468 μM) solution and 1 mL of sample solution (with appropriate dilution if necessary). The reaction was started by adding 1 mL of PMS solution (60 μM) to the mixture. The reaction mixture was incubated at 25 °C for 5 min, and the absorbance was measured at 560 nm against the corresponding blank solution. The result was expressed as inhibition in relation to a control test.