Peach (Prunuspersica (L.) Batsch, cv. Feicheng) fruits were harvested at commercial maturity in Feicheng, Shandong, China, and were transferred to the laboratory within 6 h. Uniform fruits free of defects and mechanical damage were selected and randomly divided into three groups, the 0 d samples, stored at 0 ± 0.5°C, CEO-treated and stored at 0 ± 0.5°C. We previously assessed the dose–response effects of CEO on IB index of peaches, and the concentration of 50 μL/L for subsequent investigation have been selected (). To perform the CEO treatment, peach fruits were placed in a polypropylene container with 6.65 L air space. Subsequently, two pieces of sterilized filter paper were attached on the two sides of the inner surface with 332.5 μL of CEO added, then the lid was quickly covered and the final concentration of 50 μL/L of air (v/v) was obtained. The containers were sealed with PVC cling film and then stored at 0 ± 0.5°C for 28 d. Sampling points for IB measurement were set at 0, 7, 14, 21, 28 d. According to the IB results, we selected day 21 of cold storage to perform the examination of aroma content in peaches. Briefly, after cold storage for 21 d, the fruit were transferred for 2 d of shelf-life at 20°C for subsequent ripening. For clarity, these 2 d were indicated with “21 + 2”. Control fruits were subjected to the same condition without any CEO treatment. For each sampling point, 12 fruits were sampled with 4 fruits in each replicate. After determination of IB, all samples were immediately frozen in liquid nitrogen and stored at-80°C refrigerator for molecular analysis.

Pure-grade CEO (barks steam distillation; origin: China) was purchased from Guangzhou Hengxin Spice Co., Ltd., Guangzhou, China, and stored in the dark, at room temperature. The main composition of CEO was given in Table 1.

Internal browning (IB) index was used to assess the degree of flesh browning by calculating the brown extent of each fruit after cutting peach along the axial diameter. IB index was measured based on the following formula of previous method (32): IB index = 100% × Σ [(internal browning scale) × (number of fruit at that internal browning scale)]/ [4 × total number of fruit in each treatment]. The internal browning scale was 0 = no browning; 1 = less than 1/4 browning; 2 = 1/4–1/2 browning; 3 = 1/2–3/4 browning; 4 = more than 3/4 browning.

CIELAB color space could provide parameters that are highly correlated with human visual perception of fruit color. Each color parameter is represented in a color diagram, where L* value ranges between 0 and 100, indicating the gradient from white to black; the positive a* value indicates a reddish-purple color, and negative a* value indicates a bluish-green color; the positive b* value indicates a yellow color, and negative b* value indicates a blue color (33). In this study, a portable NR10QC colorimeter (3nh Corp., China) was used to measure peach flesh browning. L* and a* values were measured on the flesh adhering firmly to the core (three points in each fruit). A total of nine peaches were randomly selected from the three replicates with three fruits of each group.

The volatile compounds analysis of peach fruits samples was implemented using a HS-GC-IMS instrument (Flavor Spec®, G. A. S., Department of Shandong HaiNeng Science Instrument Co., Ltd., Shandong, China). Two grams of the minced peach pulp (close to the skin, ca. 0.5 cm thickness) was transferred to a 20 mL headspace bottle and incubated at 40°C for 15 min. Then, 500 μL of headspace gas was sampled and automatically injected into a heated syringe at 45°C. GC was performed with a 15 m capillary column (FS-SE-54-CB-1, inner diameter: 0.35 mm) to separate the volatile compounds and coupled to IMS at 45°C. Nitrogen (99.999% purity) was used as the carrier gas at the following programmed flow rates: 2 mL/min for 2 min, 100 mL/min for 18 min, and then the flow stopped. The analytes were eluted, separated at 60°C in the column, and then ionized in an IMS ionization chamber at 45°C. A flow rate of 150 mL/min was used as the drift gas for IMS. All analyses were performed in triplicate.

GC-IMS data were analyzed using Laboratory Analytical Viewer (LAV), GC-IMS Library Search, the Reporter, Gallery plot and dynamic PCA plug-ins. LAV was used to view the analysis spectrum where each point represents a volatile compound. The GC-IMS Library Search used the IMS database to qualitatively analyze the substances. The Reporter plug-in could directly compare the difference between samples using 3D topographic plots. The Gallery plot plug-in could intuitively and quantitatively compare each volatile compound between different samples by fingerprint comparison. The dynamic PCA plug-in could cluster samples and determine the unknown substances using dynamic principal component analysis.

The potential effect of CEO on peach fruit aroma was analyzed in a panel test following the method of Duan et al. (34), with some modification. Peach flesh was cut into slices and placed in labeled glass dishes with lids. The interval between cutting fruit and sensory test was less than 30 min. It was presented the 21 + 2 d-CEO treated sample and 21 + 2 d-control fruit with an alternated order. After smelling, panelists were asked to indicate which one had more peach odor, and to evaluate their preference and acceptance.

Trizol reagent (Invitrogen) was used for RNA isolation. RNA was converted to cDNA using PrimeScript™ RT reagent Kit with gDNA Eraser (Takara, Dalian, China) following the manufacturer’s instructions. Real-time PCR was conducted with CFX96™ Real-Time System (Bio-Rad, Hercules, United States) using TB Green Premix Ex Taq (TliRNaseH Plus) (Takara, Dalian, China). The fold change of the expression of the mRNA was calculated by the 2^-ΔΔCt method. Primers were listed in Table 2. The temperature program for qPCR was as follows: 94°C for 3 min, 45 cycles of 94°C for 5 s and 60°C for 30 s, followed by a final melting curve step from 65°C to 95°C.

All data were expressed as mean value ± standard deviation (SD). Data were analyzed by one-way ANOVA and Duncan’s multiple range tests using SPSS 16.0 software. p value <0.05 was considered statistically significant.

As shown in Figure 1, browning from the core toward the flesh was evident in both control and CEO treatment from 7 d onwards. With the extension of cold storage, the browning color became deeper while the browning scale became larger. Especially during the period 14–21 d, the browning extent of the peach fruit flesh increased sharply. However, fruits treated with CEO exhibited less internal browning than the control, especially at 7, 14, and 21 d, and the difference between them was visually significant.

Furthermore, lightness (L* value) and chrome a* value, combined IB index were used to quantitatively assess the degree of browning of the internal tissues of “Feicheng” peach fruit during cold storage. Previous studies have reported a high correlation between the L* parameter and peach flesh browning, with decreasing L* values being a reliable indicator of browning (35, 36). a* parameter was frequently used to assess fresh-cut products, as reported in fresh apple or pear slices, and it was considered that a* value of the pulp was the best indicator of browning (37, 38). In our study, the changes in L* and a* values of the flesh around the core may indicate browning (dark and red flesh) because before storage flesh was white. The L* value tended to decrease with increasing refrigeration time, while a* value increased significantly, indicating that the low temperature stress darkened the internal browning tissue to a more intense red color (Figures 2A,B). At the same time, higher IB index was observed throughout experimental period (Figure 2C). Compared with the control, the internal flesh color of the fruit treated with CEO was much lighter (L* value increased), less chromatic (a* value decreased), and the internal browning scale showed slight decline. The differences of L* value at 7, 14, 21 d, a* value at 7 d, and IB index at 7, 14, 21 d, between the CEO-treated and the control fruit were significant (p < 0.05), of which the parameters L* value and IB index were highly correlated with the above visual evaluation. These data revealed that CEO treatment was effective in alleviating the severity and delaying the development of internal browning of “Feicheng” peach fruit in early stage of cold storage.

The data for volatile compounds in different treated groups analyzed by HS-GC-IMS were presented by 3D topography in Figure 3A, where the X-axis represented the ion migration time for identification, the Y-axis represented the retention time of the gas chromatograph, and the Z-axis represented the peak height used for quantification. As shown in Figure 3A, we found that the volatiles composition in each treatment group were very similar, but the signal intensities were slightly different.

By normalizing the time of ion migration and the position of the reactive ion peak (RIP), we obtained the top view of GC-IMS topographic plot of volatile compounds in peach fruits with different treatments (Figure 3B). The whole spectrum represents the headspace composition of the sample. It could be seen that the majority of the signals was located in the retention time of 100–400 s and the drift time of 1.0–1.8. Signal intensity was represented by the color. White indicated lower intensity and red indicated higher intensity. The darker the color was, the greater the intensity was. Compared with the 0 d samples, the red spot areas framed by black boxes at the retention time range of 110–130 s, 130–170 s, 180–210 s, 240–270 s, and 340–380 s were distinctly smaller or even disappeared in the control samples stored after 21 d of cold storage plus 2 d of shelf life, showing that the signals of volatile substance were much lower than those of the day 0 samples. However, almost all of the above weakened spots or areas were strengthened in CEO-treated samples under the same storage conditions. Moreover, the signals framed by white boxes for the two storage groups located in the retention time range of 160–230 s, 240–260 s, and 330–380 s were stronger than those observed in the 0 d samples. Among them, it was noteworthy that only the CEO-treated samples exhibited higher signal intensity during the retention time of 240–260 s. After long-term cold storage, the signals of characteristic aroma-related volatiles disappeared or their intensity decreased. In contrast, CEO treatment effectively prevented the loss of these volatiles, leading to a certain increase in these signal intensities.

The fingerprinting technique was used to quantitatively reveal the dynamic changes of each volatile compound. The appreciated visual signals were listed, respectively, by Gallery plot for intuitive comparison. Therefore, variance between different treatments were evident. Due to the formation of adducts between the analyzed ions and neutral molecules when passing through the drift tube, some single compounds were observed more than one signal (e.g., dimers or trimers) in the drift time. The results demonstrate that the formation of dimers or high clusters associated with compounds having high proton affinity and/or higher concentration in the analyte.

A total of 24 typical target compounds from topographic plots were identified by the GC-IMS Library, including ester, alcohol, aldehyde, and ketone, based on retention and drift times (Figure 4; Table 3). As presented, the signal intensities of propanal, benzaldehyde, (E)-2-hexenol, 2-methylpropanol, acetophenone, hexyl acetate, amyl acetate, isoamyl acetate, isobutyl acetate, propyl acetate, and ethyl acetate in the 0 d samples were much higher than those in the control samples after 21 d of storage plus 2 d of shelf life (p < 0.05), whereas ethyl hexanoate, ethyl 3-methylbutanoate, and ethyl formate were not detected in the latter samples. Among those weaker or lost signals, improved signal intensities of propanal, 2-methylpropanol, ethyl hexanoate, hexyl acetate, amyl acetate, isoamyl acetate, isobutyl acetate, propyl acetate, ethyl acetate and ethyl formate were seen in CEO-treated samples (p < 0.05), suggesting that the declined concentration level of these compounds were partly recovered after CEO fumigation.

On the other hand, some volatile compounds including nonanal, linalool, and butyl acetate were distinctive and had higher concentrations in the two storage groups than that on harvest day (p < 0.05). There was only one compound, propyl acetate, detected a remarkable promotion of signal intensity in CEO-treated samples, whereas almost undetectable in both samples at day 0 and the control at day 21 + 2. Propyl acetate, frequently described as having “fruity” and “sweet” odor attributes, was identified as a major compound that contributed to aroma formation in strawberry and apple fruits (39, 40), but it was rarely detected in peach fruits. Furthermore, CEO-treated samples had higher concentrations of amyl acetate, isoamyl acetate, and butyl acetate than those in the two other groups (p < 0.05). The above 4 esters, both newly formed or quantitatively increased, resulted in a richer and more diverse peach fruit flavor perception. Thus, the CEO treatment effectively improved the refrigerated aroma quality of “Feicheng” peach.

Principal component analysis (PCA) was performed to highlight the differences in volatile profiles of peach fruits based on signal intensities. PCA of volatile compounds in different treated-peach fruit was shown in Figure 5. It presented the first two principal components that described 56 and 30% of the accumulative variance contribution rate. The three treatment groups were remarkably distinguished in the distribution map. Fruits samples for 0 d were mainly located in the positive areas of PC1 and PC2. Peach fruits of the control at day 21 + 2 were absolutely located in the negative areas of PC1 and PC2, while peach fruits of CEO-treated samples were mainly located on the negative area of PC1 and the positive area of PC2. Except ethanol, all of the content of volatile compounds listed in Table 3 were significantly different among treatment groups (p < 0.05). These exhibited major changes of aroma volatiles in control after 21 d of cold storage plus 2 d of shelf life compared with 0 d peach and obvious recovery in CEO-treated fruits.

The aroma profiles of peach fruits with different treatment were perceived by a sensory panel consisting of 10 trained individuals. The panelists recognized that CEO treated fruits had more peach aroma intensity than the control after 21 d of cold storage plus 2 d of shelf life, with more preference and better acceptance (Figure 6), which further suggested that CEO treatment could effectively alleviate the occurrence of aroma-related volatiles loss and promoted their recovery in peach fruit during long-term cold storage.

For further understanding the molecular mechanism of the effect of CEO on peach fruit volatiles, changes in transcript level of genes involved in LOX pathway were determined. The transcript level of PpLOX1 tended to successive decline over cold storage period, followed by an obvious elevation when comparing the level at 21 d with that during 2 d of shelf life in both of two groups. PpLOX1 in CEO-treated group displayed higher expression at 21 d and 21 + 2 d (p < 0.05; Figure 7A). However, PpLOX2 showed different response patterns. PpLOX2 in both two groups showed sharp increase at the very beginning of cold storage, subsequently decreased till the end of shelf life, and CEO treatment significantly promoted its expression from 7 d to 21 + 2 d (p < 0.05; Figure 7B). During cold storage, the transcript level of PpHPL1 was generally decreased, with a burst after transferring to 20°C. Significant higher level was found in control at 7 d, while CEO induced its expression after 2 d of shelf life (p < 0.05; Figure 7C). Similar expression trend was also observed for PpADH1 throughout the experimental period, but no significant difference was observed between CEO treatment and the control, while CEO induced its expression since 21 d (p < 0.05; Figure 7D).