Yanzhihong apricot fruits were harvested from an orchard in Wanjiagou, Hohhot. To ensure the representativeness of the samples, a preliminary experiment was conducted with ten randomly selected mature apricot fruits. An a priori power analysis was performed using G*Power software (effect size d = 1.2, significance level α = 0.05, and sample size per group = 10), yielding a power value of 0.85. This result confirmed that the sample size provided sufficient statistical power (>0.8) to detect metabolic differences between groups. After harvest, a portion of the fruits was immediately dissected using a scalpel to obtain the target tissue regions for spatial metabolomics analysis, which were then placed on dry ice. The remaining fruits were cut into red and yellow sections, stored in centrifuge tubes on dry ice, and transported to the laboratory for physiological measurements.

Frozen sections of apricot tissue were prepared using a Leica CM1950 cryostat at a thickness of 50 μm. The fruit tissues samples, retrieved from a −80°C freezer, were equilibrated at -20°C in the cryostat for 1 hour before sectioning. The samples were mounted onto sample holders, and their orientation was carefully adjusted prior to securing them onto the cryostat’s positioning stage. The sections were transferred using a pre-cooled brush onto pre-chilled indium tin oxide (ITO) glass slides. To ensure adhesion, the back of the slides was pressed against the hand until the sections became transparent. Gentle finger rubbing on the back of the slide facilitated the evaporation of residual moisture, turning the sections from transparent to white. The prepared ITO slides were then dried in a vacuum desiccator for 30 minutes.

A 15 mg/mL solution of 2,5-dihydroxybenzoic acid (DHB) was prepared using a 90:10 acetonitrile:water solvent system. The DHB matrix solution was evenly sprayed onto the tissue sections using a TM-Sprayer matrix deposition system. The spraying parameters were set as follows: temperature, 60°C; flow rate, 0.1 mL/min; pressure, 7 psi; 25 spraying cycles, with a drying time of 5 seconds between cycles.

The matrix-coated ITO slides were placed on the mass spectrometer target plate(Mass Spectrometer Model: tims TOF flex, Mass Resolution: 40,000, Laser Scan Count per Pixel: 400, Laser Energy: 80%, Germany). The DataImaging software (Bruker) was used to define the tissue regions for analysis, with an imaging resolution of 50 μm (i.e., the smallest unit of the 2D matrix was 50 μm × 50 μm). The imaging area was divided into a two-dimensional array of sampling points according to the sample size. The detection range was set to 50–1300 Da. Under the same laser energy, the tissue sample is detected by focusing the laser beam onto the sample through an optical path and continuously scanning the sample. The released molecular ions were then detected by the mass spectrometer, providing mass-to-charge ratio (m/z) information and raw signal intensity data for each pixel. The raw data were imported into SCiLS Lab software, where they were normalized using the Root Mean Square (RMS) method. This process generated relative intensity values for different m/z signals at each spatial point, which were subsequently visualized as heatmap images (Deininger et al., 2011).

Standardized metabolic data were analyzed using multivariate statistical methods, including Principal Component Analysis (PCA) and Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA), to evaluate differences between fruit regions. One-way analysis of variance (ANOVA) was used to determine significant differences in metabolite concentrations between groups. Pearson correlation coefficients were calculated to assess linear relationships among metabolites, particularly between pigment accumulation and sugar–acid ratios. To control for multiple comparisons, false discovery rate (FDR) correction was applied to all p-values. Additionally, metabolic pathway enrichment analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database to interpret the functional significance of metabolite variations.

The Yanzhihong apricot is distinguished by its unique appearance, characterized by two distinct color regions—red and yellow. These regions not only exhibit significant visual differences but also display distinct physicochemical properties. To systematically investigate the biochemical composition of these fruit regions, this study measured the contents of glucose, fructose, sucrose, ascorbic acid, malic acid, citric acid, chlorophyll a, chlorophyll b, carotenoids, and total anthocyanins.

The results revealed differences in the relative abundance of various sugars between the two color regions ( Figure 1 ). Glucose levels showed no significant difference between the yellow (5.71 mg/g) and red (5.39 mg/g) regions. However, fructose content was significantly higher in the yellow region (31.54 mg/100g) than in the red region (16.88 mg/g), whereas sucrose content was markedly higher in the red region (21.39 mg/g) compared to the yellow region (17.79 mg/g). This difference suggests that the yellow region primarily derives its sweetness from fructose, while the red region’s sweetness is more dependent on sucrose. The differential sugar distribution may be influenced by localized regulation of sugar metabolism pathways and enzyme activity. Significant differences were also observed in the distribution of organic acids. The yellow region exhibited higher levels of ascorbic acid (11.03 mg/100g), malic acid (3.61 mg/g), and citric acid (6.41 mg/g) compared to the red region (8.29 mg/100g, 2.34 mg/g, and 4.27 mg/g, respectively). These results indicate that the yellow region possesses a more pronounced acidic flavor profile.

Regarding pigment accumulation, chlorophyll a content was higher in the yellow region (8.38 mg/g) than in the red region (6.06 mg/g), whereas chlorophyll b content was substantially higher in the red region (9.56 mg/g) than in the yellow region (1.89 mg/g). Additionally, carotenoid content was significantly elevated in the red region (29.25 mg/g) compared to the yellow region (19.27 mg/g), contributing to the enhanced color saturation of the red region. The accumulation of anthocyanins further intensified the red hue of the fruit, albeit with a relatively low overall concentration. The red region exhibited a slightly higher anthocyanin content (2.95 mg/100g) than the yellow region (2.35 mg/100g), reinforcing its distinct color.

This study systematically investigated the interrelationships among key metabolites, including sugars, organic acids, anthocyanins, and carotenoids, to elucidate their metabolic interactions ( Supplementary Figure S1 ). The results demonstrated significant correlations between these metabolites, highlighting their complex regulatory relationships within metabolic pathways.

A particularly strong positive correlation was observed between ascorbic acid and citric acid (r = 0.9971). Conversely, ascorbic acid showed a significant negative correlation with Cyanidin-3-O-(6’’-O-malonyl) glucoside (r = -0.9795), suggesting that an increase in ascorbic acid may suppress the synthesis of certain anthocyanins, thereby influencing fruit coloration and pigmentation intensity. Furthermore, citric acid exhibited a strong negative correlation with carotenoid accumulation (r = -0.9953), indicating a potential inhibitory role of citric acid in carotenoid biosynthesis. Similarly, malic acid was also negatively correlated with carotenoid levels (r = -0.9916), further supporting the regulatory role of organic acids in carotenoid metabolism.

The metabolic relationship between anthocyanins and carotenoids was also noteworthy. Pelargonidin 3-O-beta-D-glucoside 5-O-(6-coumaroyl-beta-D-glucoside) exhibited a negative correlation with carotenoid accumulation (r = -0.9858), and a similar trend was observed for Cyanidin-3-O-(6’’-O-malonyl) glucoside (r = -0.9780). These findings suggest that the balance between anthocyanins and carotenoids may influence the final fruit coloration and the overall nutritional composition.

Figure 2A presents the results of Principal Component Analysis (PCA) of the differently colored regions (yellow and red) of apricot fruit, revealing the overall metabolic differences between the two groups. The x-axis represents the first principal component (PC1), which explains 80.1% of the variance, while the y-axis represents the second principal component (PC2), accounting for 7.2% of the variance. The PCA score plot clearly demonstrates a distinct separation between yellow and red fruit samples in the PC1–PC2 plane, indicating significant differences in metabolic characteristics between these color regions. The yellow fruit samples (yellow_1, yellow_2, yellow_3) clustered in the lower PC1 region, whereas the red fruit samples (red_1, red_2, red_3) were concentrated in the higher PC1 region. Furthermore, the high intra-group clustering suggests that the metabolic features associated with fruit color are highly stable and reproducible. Figure 2B displays the results of Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA), which further distinguishes the metabolic profiles of the yellow and red fruit samples. The x-axis represents the predictive principal component [T score (Zhou et al., 2020)], explaining 80.9% of the variance, while the y-axis represents the orthogonal principal component [Orthogonal T score (Zhou et al., 2020)], explaining 6.18% of the variance. The analysis revealed a clear separation between the yellow and red fruit samples along the T score (Zhou et al., 2020) axis, with yellow fruit samples (yellow_1, yellow_2, yellow_3) clustering on the right and red fruit samples (red_1, red_2, red_3) on the left. This distinct separation pattern further confirms systematic differences in the chemical composition and physicochemical properties of the two groups. Compared to PCA, OPLS-DA provided a more targeted classification by enhancing the visualization of inter-group differences while minimizing the influence of irrelevant variables, resulting in a clearer distinction between the sample groups. Figure 2C presents the correlation matrix of the different colored regions, providing insights into the overall consistency and variability in metabolite composition. The yellow fruit group (yellow_1, yellow_2, yellow_3) exhibited high internal correlation, with correlation coefficients approaching 1, indicating extremely stable metabolic composition with minimal inter-individual variation. In contrast, while the red fruit group (red_1, red_2, red_3) also showed high correlation, the values were slightly lower than those of the yellow group, suggesting a higher degree of metabolic variability within the red fruit samples. Figure 2D displays the heatmap of metabolite abundance in different fruit regions, categorizing metabolites into different classes, such as organic acids, amino acids and derivatives, and flavonoids. Standardized processing was applied to facilitate visualization of metabolite expression levels across samples. The relative abundance of metabolites in different samples further highlighted systematic metabolic differences between the yellow and red fruit regions. The yellow fruit exhibited higher expression levels in multiple metabolite categories, particularly in organic acids, flavonoids, and amino acids. In contrast, the red fruit showed lower or more dispersed metabolite expression in these categories, particularly for flavonoids and organic acids, which exhibited greater variability.

Overall, the heatmap analysis clearly illustrated the metabolic divergence between different fruit color regions. The yellow fruit exhibited a more stable metabolic profile with consistently higher expression levels across various metabolite categories, whereas the red fruit displayed greater variability, suggesting that its color formation, flavor profile, and nutritional value may be influenced by more complex regulatory mechanisms.

Figure 3A presents the differential metabolite analysis between the red and yellow regions of apricot fruit. The volcano plot displays Log₂(Fold Change) on the x-axis, representing the magnitude of metabolite expression changes, while the y-axis shows Log₁₀(P-value), reflecting statistical significance. Each point represents a metabolite, with its color, size, and position indicating statistical significance, expression fold change, and Variable Importance in Projection (VIP) values, respectively. Overall, most metabolites (gray points) exhibited significant expression differences between the yellow and red fruit regions. However, distinct clusters of upregulated (red points) and downregulated (green points) metabolites highlighted the pronounced metabolic differences between the two color regions. Specifically, 101 metabolites were significantly upregulated in the red fruit, while 626 metabolites were significantly downregulated. Additionally, 311 metabolites showed no significant changes, suggesting stable expression levels across both fruit regions.

Figure 3B illustrates the enrichment patterns of various metabolite classes based on KEGG metabolic pathway analysis. The metabolites analyzed include lipids, amino acids and derivatives, organic acids, alkaloids, flavonoids, terpenoids, phenolic acids, nucleotides and derivatives, steroids, lignans and coumarins, and quinones, which display significant variations in abundance across different metabolic pathways. Heatmap analysis reveals marked differences in metabolite abundance, with red and yellow indicating high levels, while green signifies lower levels. These findings provide compelling evidence for the functional roles and regulation of metabolites across metabolic pathways, offering new insights into the dynamic changes within metabolic networks and their biological underpinnings.

Figure 3C further illustrates the results of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, revealing metabolic differences between the red and yellow regions of apricot fruit. The results indicate that primary metabolic pathways accounted for the highest proportion (50.79%), suggesting that most differential metabolites are involved in essential physiological processes such as carbohydrate metabolism and amino acid metabolism. These differences highlight the variations in fundamental metabolic activities between the two fruit regions.

The biosynthesis of secondary metabolites comprised 38.1% of the differential metabolic pathways. These compounds play a crucial role in plant defense mechanisms, antioxidant capacity, and aroma formation, potentially influencing fruit flavor and coloration. Amino acid metabolism accounted for 5.94%, suggesting that differences in amino acid synthesis and metabolism may contribute to variations in taste and nutritional composition.

Carbohydrate and glycogen metabolism pathways represented 3.17% of the enriched pathways, indicating that differences in sugar metabolism may regulate apricot sweetness and energy storage. Lipid metabolism accounted for 1.59%, suggesting potential effects on fruit ripening, antioxidant properties, and oil content. Nucleotide metabolism made up 2.13%, implying a role in cell division, gene expression, and fruit maturation regulation.

Furthermore, variations in the biosynthesis pathways of isoflavones and flavonoids highlighted their significant roles in pigment accumulation, flavor formation, and antioxidant properties in apricot fruit. These findings provide valuable insights into the metabolic mechanisms underlying color differentiation in apricot fruit and their potential implications for fruit quality and sensory attributes.

By integrating physiological measurements with spatial metabolomics data, we identified significant spatial variations in sugar distribution between different colored fruit tissues. As shown in Figure 4A , the signal intensity of N-Acetyl-D-mannosamine was relatively low in the red fruit flesh. This metabolite exhibited a localized distribution in the red region but showed a stronger signal near the peel, indicating that sugar biosynthesis is more active in the peel region. In contrast, the yellow fruit exhibited a significantly stronger N-Acetyl-D-mannosamine signal, which was more evenly distributed across the peel, central fruit flesh, and areas near the seed, suggesting a more uniform sugar distribution ( Figure 4B ). Additionally, Alpha-D-Galactosamine 1-phosphate Inhibitor showed a strong signal in the red fruit flesh, primarily concentrated in the region between the peel and inner flesh. This finding suggests that the red fruit region exhibits higher glucosidase activity, leading to more active sugar hydrolysis and conversion. In contrast, the yellow fruit showed weaker and more uniform distribution of this metabolite, with particularly low concentrations near the seed. Another metabolite, 2-(Acetyl)-3-(3-isobutyl)-1’-(methyl 2-hydroxypentanoate)-3’-lauroyl-sucrose, exhibited high concentrations in both red and yellow fruit regions ( Figure 4C ). However, its abundance was relatively lower in areas adjacent to the peel and seed, indicating spatially distinct metabolic activity in different fruit regions. These findings highlight the complex spatial heterogeneity of sugar metabolism in apricot fruit, with differences in metabolite distribution reflecting localized metabolic regulation, sugar biosynthesis activity, and enzymatic conversion processes in different tissue regions.

In the analysis of organic acids in apricot fruit ( Figure 5 ), we found that malic acid abundances were higher in the red fruit region and were predominantly distributed in both the peel and flesh. In contrast, in the yellow fruit, malic acid was less abundant and showed a limited distribution in the central flesh. Phosphinomethyl malate was also more abundant in the red fruit region, whereas its concentration was significantly lower in the yellow fruit, where it was primarily localized in the peel and near the seed. In the yellow fruit, L-Ascorbic acid 2-phosphate was more abundant and widely distributed, with approximately six times the concentration found in the red fruit. However, its spatial distribution pattern remained similar in both fruit regions.

Ecgonine methyl ester was found at higher concentrations in the yellow fruit. IMS relative quantitation analysis revealed that this metabolite exhibited a relatively uniform distribution across both fruit color regions. These findings suggest that the spatial distribution of organic acids in apricot fruit is highly variable, with distinct accumulation patterns between red and yellow regions, potentially influencing fruit acidity, antioxidant capacity, and metabolic interactions in different tissue compartments.

Our study revealed significant differences in pigment accumulation ( Figure 6 ), antioxidant capacity, and flavor formation between the red and yellow fruit regions. The red fruit flesh was rich in various anthocyanins, including Cyanidin-3-O-(6’’-O-malonyl) glucoside, Petunidin 3-(6’’-p-coumarylglucoside), and Pelargonidin 3-O-beta-D-glucoside 5-O-(6-coumaroyl-beta-D-glucoside). These compounds not only enhanced pigment deposition, resulting in a more vibrant red coloration, but also conferred stronger antioxidant properties. IMS relative quantitation analysis indicated that these anthocyanins were predominantly distributed in the peel and adjacent fruit flesh of the red region, suggesting a crucial role in peel protection and pigment biosynthesis. In contrast, the yellow fruit flesh exhibited lower anthocyanin abundances, leading to a lighter coloration. Petunidin 3-(6’’-p-coumarylglucoside) was more abundant in the red fruit and was evenly distributed throughout the entire tissue section.

However, in the yellow fruit ( Figure 6 ), its abundances was lower and primarily localized in the fruit flesh and near the seed. Similarly, Pelargonidin 3-O-beta-D-glucoside 5-O-(6-coumaroyl-beta-D-glucoside) accumulated at higher levels in the red fruit, with IMS relative quantitation showing its presence in both the flesh and seed regions. In contrast, its abundances in the yellow fruit was lower and exhibited a more uniform distribution. Additionally, 6-Hydroxycyanidin was found at higher abundances levels in the red fruit, with heatmap analysis revealing its predominant localization in the peel and flesh regions. In the yellow fruit, this compound was more evenly distributed but with an overall lower abundances. Notably, 13-Apo-beta-carotenone, a key carotenoid derivative, was significantly more abundant in the yellow fruit compared to the red fruit. Heatmap visualization indicated that its signal was concentrated in the peel and flesh regions, suggesting that the yellow fruit had a more active carotenoid accumulation pathway. This pattern of distribution implies that carotenoid biosynthesis in the yellow fruit may contribute not only to its coloration but also to its enhanced sweetness.