The experiment was conducted at the grape research base of the Horticulture Institute (87.28’E, 45.56’N), Xinjiang Academy of Agriculture Sciences, under the fruit quality control post of the national grape industry technology system, from July to December 2018. The experimental base covers an area of four hectares and is equipped for irrigation, fertilization, weeding, and spraying. The base is well constructed and developed. In the current study, 18 varieties of fresh grapes were used as test materials over the course of 6 years. These varieties covered the main types with varying maturities, shapes, and colors, as outlined in Table 1 . All varieties were planted on the same location in 2012 and began bearing fruit in 2014. Grapes plants were spaced in 1×3.5 m, and standard cultivation practices were followed. For each variety, 15 healthy trees with uniform growth and the same flowering period were selected. Each group consisted of five trees with three biological repeats and berries were harvested at their optimum technological maturity as per the OIV resolution VITI 1/2008 (OIV, 2008).

After the fruit sample was cleaned well, the exocarp and mesocarp were quickly separated with a scalpel, cut into small pieces, and put in liquid nitrogen. Some extra samples were transferred to the laboratory and stored in a freezer at -80°C. The grape berry samples were freeze-dried in a vacuum, and the freeze-dried berry was ground (30 Hz, 1.5 minutes) to powder by using a grinder with zirconia beads. Weigh 20 mg grape powder and add 500 μL methanol: isopropanol: water (3:3:2 V/V/V) extract, vortex for 3 min, and ultrasonic in ice water for 30 min. Centrifuge, add an internal standard, and then freeze dry. The chemical derivation method is as follows: a small molecule carbohydrate sample and 100 μL methoxy ammonium salt pyridine (15 mg/mL) solution were mixed, and the mixture was incubated at 37°C for 2 h. Then, 100 mL of BSTFA solution was added to the mixture and incubated at 37°C for 30 min to obtain the derivatized solution. The solution was diluted with n-hexane to an appropriate concentration and stored in a brown sampling bottle for analysis.

The vertical and horizontal diameters of fruit were measured with an electronic digital vernier caliper, and the fruit shape index was calculated (Barbagallo et al., 2020). The weight of a single fruit was measured with a 1/10000 electronic analytical balance. The content of total soluble solids concentration (°Brix) was determined by a PAL-1 digital refractometer, and the total acid was determined by a PAL-BX/ACID-2 Brix acid-meter.

Sugar determination was carried out using the GC-MS method (Medeiros and Simoneit, 2007; Ruiz-Matute et al., 2011; Milkovska-Stamenova et al., 2015; Wang et al., 2022). Briefly, sugar substances in grape berries were analyzed by gas chromatography (Agilent 7890B), mass spectrometry (7000 d), and a DB-5MS column. With helium as the carrier gas, the flow rate is 1 mL/min. The injector and source temperatures were maintained as per standard procedures. The oven temperature ramp progress was maintained at 170°C, 250°C, 280°Agilent 7890B), mass spectrometry (7000 d), and a DB-5MS column. With helium as the carrier gas, the flow rate is 1 mL/min. The injector and source temperatures were maintained as per standard procedures. The oven temperature ramp progress was maintained at 170°C, 250°C, 280°C, and 310°C. The details of the GC-MS analysis for specific parameters are presented in Table 2 .

Based on the GC-MS platform, MetWare software (Wuhan, China, http://www.metware.cn/) was used to do a qualitative and quantitative analysis of the sugar components (Wang et al., 2022). For each group of samples, three biological replications were maintained. Sugar standards were procured from Olchemim, Aladdin (Shanghai), and Sigma (America). The detected sugar components include 9 monosaccharides and 4 disaccharides ( Table 3 ).

The Agilent MassHunter qualitative and quantitative software was used for data processing. The total ion current diagram is shown in Figure S1 . It can be seen from this study that the repeatability and reliability of QC sample data are well, as shown in Figure S2 . Attached Table S1 is the linear equation for different carbohydrate species.

Calculation:

B: Sugar concentration value (μg/m)

C: Volume of solution for constant volume (μL)

D: Volume of supernatant (μL)

E: Volume of extract (μL)

F: Weighed sample mass (g)

The mean of three replicates was used to express all berry quality parameters and different sugar data. IBM SPSS v25.0 software (IBM SPSS Inc., Chicago, IL, USA) was employed to analyze Duncan’s test at a different level of significance. Pearson’s correlation coefficient was used to determine the correlation between variables. R software was utilized to conduct principal component analysis (PCA) and plot the grapes.

The phenotypes, such as appearance and morphology, of the 18 grape varieties are known for their own characteristics at fruit maturity ( Figure 1 ). The basic characteristics of each grape variety at fruit maturity are measured, including weight, size, firmness, and total acid, and presented in Table 4 . Based on the data in the table, it is evident that all of the varieties used in this study differ significantly. For instance, results showed that in terms of individual bunch weight, ‘XinYu’ has the largest bunch weight of 954.97g, followed by ‘Red Globe’, and so on. The ‘Wuhecuibao’ variety had the smallest panicle weight of 201.46g, while the panicle weight of ‘XinYu’ was nearly five times higher. A large variation was observed in the longitudinal diameter of individual grape bunches among different varieties. The results showed that bunch longitudinal diameter ranged from 17.58 to 30.78 cm. The highest bunch longitudinal diameter was observed in ‘Centennial seedless’ and the smallest was observed in ‘Wuhecuibao’. The bunch longitudinal diameter of ‘Centennial seedless’ was 75.08% larger than ‘Wuhecuibao’. In terms of bunch diameter, ‘Yatomi Rosa’ had the largest diameter of 22.21cm, while ‘Wuhecuibao’ had the smallest bunch diameter of 8.49 cm. The bunch diameter of ‘Yatomi Rosa’ was 161.60% larger than that of ‘Wuhecuibao’.

Grape berries of different varieties with different colors and shapes are shown in Figure 1 . The parameters related to the berry index were also measured to understand the structural variation among the varieties. The longitudinal diameter of barriers ranged from 14.28 to 35.43mm. The longitudinal diameter of ‘Rizamat’ was the largest, and ‘Thompson seedless’ had the smallest longitudinal diameter. The fruit longitudinal diameter of ‘Rizamat’ was 148.11% larger than that of ‘Thompson seedless’. In terms of berry transverse diameter, ‘XinYu’ was in top rank with 27.74 mm, and ‘Thompson seedless’ is the smallest, reaching 10.9 mm. The berry transverse diameter of ‘XinYu’ was 53.34% larger than that of the smallest variety. As for single berry weight, ‘XinYu’ was the largest, and ‘Thompson seedless’ was the smallest.

Total soluble solid (TSS) is among the most essential attributes that determine the quality of fruits. TSS is dominated by total sugar content, with a minor contribution from soluble proteins, amino acids, and other organic materials (Xu et al., 2022). In this study, the TSS of ‘Summer Black’ and ‘Flame Seedless’ was found to be 25.12 and 24.71° Brix, respectively, which was higher than other grape varieties, while ‘Victoria’ had the lowest TSS of 14.93° Brix. ‘Summer Black’ and ‘Flame Seedless’ had 68.25% and 65.50% higher TSS than ‘Victoria,’ respectively. Furthermore, ‘Black Monukka’ and ‘Centennial Seedless’ had the highest total acid content of 0.67%, while ‘Victoria’ had the lowest total acid content of 0.27% among all grape varieties studied.

The study conducted by Huang et al. (2021) used the soluble solids to acid ratio as an indicator to evaluate grape maturity and flavor. The TSS/TA ratio was calculated for all grape varieties, and the results showed that the ratio ranged from 86.24 (Wuhecuibao) to 28.06 (Black Monukka), indicating significant differences among the varieties. Specifically, ‘Wuhecuibao’ had a TSS/TA ratio that was 207.34% higher than that of ‘Black Monukka’. Berry maturity is a crucial stage for harvesting, and fruit quality often depends on maturity indices (Niimi et al., 2017; Niimi et al., 2018). The grape varieties used in this study were classified as early and late based on their harvest dates. ‘Summer Black’, ‘Victoria’, and ‘Flame seedless’ matured early, while ‘Crimson seedless’ and ‘Munake’ matured relatively late.

The GC-MS analysis revealed significant differences in the amounts of sugar components present in the berries of the 18 grape varieties tested ( Figure 2 ). The heatmap showed that, with the exception of maltose, all identified sugar components were detected in all tested varieties, whereas maltose was only detected in five varieties. Notably, the concentration of the other nine sugar components was found to be low, while D-fructose and glucose were present in high concentrations, followed by sucrose. The heatmap depicts the average concentration of different sugar components among the varieties.

Based on the analysis of the components and contents of various sugars in different grape varieties using GC-MS, principal component analysis (PCA) was conducted to investigate the differences both within and among replicates. The results, depicted in Figure 7 , reveal that the three replicates of each of the 18 grape varieties were closely clustered together, suggesting good repeatability and ensuring the accuracy of the data. Moreover, the samples of different grape varieties were clearly distinguishable, indicating that the sugar composition and content among different grape varieties differed significantly, reaching a significant or extremely significant level. The findings highlighted the substantial variation in sugar composition and content among grape varieties.

According to Figure 8A , the total sugar content in these grape varieties is determined by the sum of monosaccharides and disaccharides detected by GC-MS, including their absolute values. ‘Melissa’ has the highest total sugar content, followed by ‘Yatomi Rosa’ and ‘Huozhouhongyu’, and then ‘Flame seedless’, ‘Wuhecuibao’, ‘Thompson seedless’, and ‘Xin Yu’, among others. Conversely, ‘Victoria’, ‘Centennial Seedless’, and ‘Rizamat’ were found to have lower sugar content. In 18 varieties, D-fructose, glucose, and sucrose contribute to over 99% of the total sugar content, with D-fructose slightly higher than glucose in 12 of them ( Figure 8B ). Additionally, inositol was identified as a higher contributor to the minor sugar components in all grape varieties ( Figure 8C ).

The correlations between various sugar components are presented in Figure 9A using Spearman correlation coefficients. The correlation matrix revealed that some sugar components showed a positive correlation with others. For instance, Xylitol content exhibited a positive correlation with L-fucose (p<0.001) and L-rhamnose (p<0.001). Similarly, lactose content was positively correlated with D-sorbitol (p<0.01), L-rhamnose (p<0.05), and trehalose (p<0.05) content. A strong correlation between sucrose content and glucose content was observed. Furthermore, glucose content in varieties was found to have a strong correlation with D-fructose content (p<0.001). Additional positive and negative correlations were identified, and further details are presented in Figure 9A .

Results shown in Figure 9B revealed that there is no significant positive correlation between total sugar and soluble solid content. For example, although ‘Yatomi Rosa’ has the second-highest total sugar content, its proportion of soluble solids is lower. Similarly, ‘Summer Black’ has the highest soluble solids, but its total sugar content is not particularly high. In contrast, ‘Flame Seedless’ has relatively high levels of both total sugar and soluble solids, while ‘Victoria’ has lower levels of both. There are significant differences in sugar and total soluble content among some varieties. For instance, the total soluble solid content in ‘Summer Black’, ‘Flame Seedless’, and ‘Thompson Seedless’ is similar, but the total sugar content varies greatly. Therefore, the plot showing the relationship between total sugar and total soluble solids indicates that there is no positive correlation between these parameters.

The morphology and characteristics of berries vary among the varieties and different species of grapes ( Table 4 ). Quantifying the phenotypic parameters of grape berries and bunches is important for precision agriculture (Liu et al., 2022a). The cultivated grapes are known to have high variation compared with wild resources, which largely resemble round berries in shape (Rodríguez et al., 2011; Zhang et al., 2021). In many fruit crop markets, demand is closely associated with the shape and quality of fruit crops. Varietal development programs are also associated with fruit shape in many fruit crops (van der Knaap et al., 2014; Zhang et al., 2021). The morphological characteristics observed in our study showed that the popular grape varieties grown in the Xinjiang region are round to oval in shape. Many studies have shown that berry appearance is highly associated with berry index and appears to influence the acceptability and preferences of consumers when it comes to fresh fruit consumption (Lecourieux et al., 2014; Zhang et al., 2022b). The studies by Zhang et al. (2022a) and Zhang et al. (2022b) revealed that in some regions, spine grapes are eaten because of their large shape and flavor. The berries are characterized by high variability in grapes and these traits are also used to describe different grape varieties and domestication processes (Barbagallo et al., 2020). We measured various berry traits to highlight the phenotypic diversity of berries in different varieties grown in Xinjiang. The varieties used in the current investigation showed higher variability in various berry characteristics. The significant differences were reported in bunch parameters, berry parameters, berry index, TSS, total acid and harvesting time. The technological maturity parameters of Italian table grapes were measured in some popular varieties. The shorting of varieties on a commercial level showed significant variation in different chemical parameters, including TSS, TA, TSS/TA, pH, and different major sugar components (Segade et al., 2013). Barbagallo et al. (2020) used a number of grape varieties from the Sicilian genetic pool and precisely observed the variability in seed and berry traits. Similar research was carried out by Bouby et al. (2013) to identify the difference in elongation of the pip body between primitive cultivars and highly domesticated cultivars. The results revealed that traits of domestication are related to the strength of selection pressure (Bouby et al., 2013).

The quality of grape berries is mainly determined by the type and amount of sugar, soluble solids, and organic acids content. Among these factors, sugar content is the most important. Xinjiang is home to a wide variety of grape germplasm, but different grape varieties have varying levels of sugar. In a study of 18 grape varieties, we found that fruit size, shape, and color characteristics differed at maturity, as did sugar content. Sugar accumulation during fruit development has been extensively studied in various species, and the amount of total soluble sugars typically increases with growth, reaching a peak at maturity or ripening (Basson et al., 2010; Li et al., 2012; Cao et al., 2013). However, the patterns and concentrations of sugar accumulation can differ between species. Glucose and fructose are typically the major proportion of soluble sugars in most fruits, while sucrose is predominant in some species like peaches, citrus, and litchi (Yakushiji et al., 1996; Desnoues et al., 2014; Li et al., 2020).

In our studies, we observed a similar pattern of sugar distribution, where glucose and fructose contributed the majority of sugars. The sugar-to-acid ratio reflects fruit taste, and the flavor of grape berries is closely related to sugar-acid content (Guo et al., 2017; Zhang et al., 2022b). Fructose and glucose are the most common sugars in most fruits, and they also contribute to the flavor of grape berries. Other metabolic changes occur during grape berry ripening, such as the accumulation of sugars in the form of glucose and fructose in the berry vacuoles (flesh and skin) following sucrose translocation from the leaves (Durán-Soria et al., 2020). Recent research by Li et al. (2020) compared sugar profiles in major fruit crops and found that glucose and fructose were the most abundant sugars, which is consistent with our findings. However, sucrose was found to be the predominant sugar in some fruits, including peaches.

The recent studies conducted on grapes have shown that the sugar content of wild grapes is primarily divided into fructose and glucose, as reported by Liu et al. (2006); Jiang et al. (2017); Zhang et al. (2022b) and Segade et al. (2013). Our current study supports these findings and indicates that the sugar composition and level vary significantly across grape varieties, as observed in previous studies on grapes (Zhang et al., 2022b) and citrus (Zhou et al., 2018). Specifically, our analysis of individual sugar components revealed that fructose, glucose, and sucrose are the three primary sugar components in grapes, accounting for over 99% of the total sugar content. Among the 18 grape varieties studied, fructose accounted for 42.68%-50.95% of the total sugar, while glucose accounted for 42.13%-46.80%, both accumulating at significantly higher levels than sucrose and other sugar substances, consistent with the findings of Li et al. (2020) and Liu et al. (2006). Similar results were observed in the sugar concentration of other fruits, where D-fructose, sucrose, and glucose were the major sugar contributors (Ma et al., 2017). The proportion of sucrose content in total sugar ranged from 6.17% to 12.69%. Orak (2009) reported a variation in glucose content of 5.98% to 12.21% and fructose content of 5.93% to 12.66% in 24 significant grape varieties, which is similar to our findings. The carbohydrate composition on grape varieties identified by thin layer chromatography, and spectrophotometric Dubois method also revealed that glucose and fructose as major constituents in black grape varieties (Iosub et al., 2013). However, the variation in fructose and glucose ranges may be due to different analytical methods used for sugar content estimation. Interestingly, our current research found that the majority of grape varieties had trace amounts of monosaccharides and disaccharides, unlike many other studies. The concentration of lactose and maltose was measured in some Korean fruits and vegetables, but they were not present in most of them. Shanmugavelan et al. (2013) found that the concentration of trace sugar was not enough to be detected by HPLC in Campbell early and green varieties of grapes. In contrast, GC-MS-based sugar quantification in our study showed that many trace sugars were present in most of the grape varieties, with only five varieties showing the presence of maltose. The differences in the identification of various sugars can be attributed to analytical methods or variety differences. In our study, the sugar composition determined by GC-MS was combined with the data of soluble solids, and it was found that the sugar content, an important quality factor, differed significantly among the 18 grape varieties. GC-MS has been found to be effective in previous studies for the quantitation of carbohydrate intermediates (Milkovska-Stamenova et al., 2015). In short, current study allowed to characterize economically important grape varieties grown in Xinjiang grapes according to physiochemical and sugar composition.