Next Article in Journal
Fruit-Based Fermented Beverages: Contamination Sources and Emerging Technologies Applied to Assure Their Safety
Previous Article in Journal
Functional Bread Produced in a Circular Economy Perspective: The Use of Brewers’ Spent Grain
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Foods
Volume 12
Issue 4
10.3390/foods12040837
foods-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Flavoromics Approach in Critical Aroma Compounds Exploration of Peach: Correlation to Origin Based on OAV Combined with Chemometrics

by
Qianqian Li
1,
Bei Li
2,
Rong Zhang
1,
Shuyan Liu
1,
Shupeng Yang
1,
Yi Li
1 and
Jianxun Li
1,*
1
Key Laboratory of Agro-Products Quality and Safety Control in Storage and Transport Process, Ministry of Agriculture and Rural Affairs, Institute of Food Science and Technology, Chinese Academy of Agricultural Sciences, Beijing 100093, China
2
Key Laboraory of Tropical Fruits and Vegetables Quality and Safety for State Market Regulation, Hainan Institute for Food Control, Haikou 570314, China
*
Author to whom correspondence should be addressed.
Foods 2023, 12(4), 837; https://doi.org/10.3390/foods12040837
Submission received: 27 November 2022 / Revised: 19 January 2023 / Accepted: 31 January 2023 / Published: 16 February 2023
(This article belongs to the Section Foodomics)
Browse Figures
Versions Notes

Abstract

:
It is essential to seek the critical aroma compounds to identify the origins of peach as well as provide a guidance for quality evaluation. In this study, the peach was characterized by HS-SPME/GC-MS. Subsequently, the odor activity value (OAV) was calculated to specify the primary aroma-active compounds. Afterwards, the chemometrics methods were employed to explore the potentially critical aroma on the basis of p value, fold change (FC), S-plot, jack-knifing confidence interval, variable importance for projection (VIP), and the Shared and Unique Structures (SUS) plots. As a result, five compounds (methyl acetate, (E)-hex-2-enal, benzaldehyde, [(Z)-hex-3-enyl] acetate, and 5-ethyloxolan-2-one) were considered as critical aromas. Moreover, the multi-classification model was developed with an outstanding performance (accuracy of 100%) using the five critical aroma. Moreover, the potential chemical basis of odors was sought through sensory evaluation. In addition, this study provides the theoretical and practical foundation for geographical origin traceability and quality evaluation.

Graphical Abstract

1. Introduction

The peach (Prunus persica L. Batsch) has been cultivated for more than 3000 years in China [1]. It is known that the germplasm resources of peaches are abundant throughout the world, which is attributed to its long history of cultivation and wide geographic distribution. The peach is ranked as one of the most nutritionally and economically valuable fruits due to its widespread consumption worldwide [2]. Moreover, the peach qualifies as a healthy food in view of high nutrients, low calories, as well as an ample amount of vitamins and minerals [3]. To a certain extent, the peach provides health benefits by improving digestion, reducing cholesterol, and boosting the immune system, which is particularly dependent upon the high content of dietary fiber, antioxidants (vitamins C, vitamins A, polyphenols), calcium, and potassium [4]. Aroma is one of the essential factors for evaluating the fruit quality [5]. It is reported that the unique aroma compounds in peach are C6 compounds, lactones, aldehydes, alcohols, esters, and so on [6,7]. Typically, the aroma is contributed by a multitude of aroma-active chemicals with the concentrations over the odor threshold [8,9]. Definitely, the OAV is essential to assess the contributions of volatile compounds to the aroma [10]. As the aroma could provide fingerprint information on its geographical origin, it is imperative to search the critical aroma compounds connected with each geographic region, which could not only identify the origin of peach but also provide guidance to reduce the denaturation or damage to the flavor-characterizing compounds and active ingredients of the fruit.
Generally speaking, gas chromatography-mass spectrometry (GC-MS) is the frequently used method for volatile compound separation and quantification in fruit [11]. As a rapid volatile extraction technique, headspace-solid phase microextraction (HS-SPME) has the advantages of simple, quick, solvent-free treatment and green environmental protection [12,13,14]. The chemometrics methods are powerful approaches to mine data thoroughly, develop efficient and robust models, and give a rational interpretation of complex data sets [15]. Principal components analysis (PCA) [16,17], partial least squares-discrimination analysis (PLS-DA) [18,19] and orthogonal partial least squares-discrimination analysis (OPLS-DA) [20] are powerful tools to explore the critical aroma compounds. Moreover, the Shared and Unique Structures (SUS) plots generated using OPLS-DA are firmly established as a forceful tool to investigate the shared and unique structures of multiple classification models [21].
In this study, the volatile profiles were characterized using HS-SPME/GC-MS. Subsequently, the volatile contribution of OAV was calculated to specify the primary aroma-active compounds. After that, the chemometrics methods were carried out to reveal the differences and explore the statistically and potentially significant aroma compounds among the three origins on the basis of S-plot analysis, significant difference, fold change (FC), and jack-knifing confidence intervals. In addition, the SUS plots were performed to give a visualization of the shared and unique biomarkers among multiple groups. Finally, the critical compounds through the two strategies of OAV and chemometrics were employed to develop the multi-classification models and to explore the potential chemical basis of odors using sensory evaluation.

2. Materials and Methods

2.1. The Peach Samples Collection

The peach (Prunus persica (L.) Batsch) samples were acquired from three characteristic provinces of Beijing (BJ), Shandong (SD), and Hebei (HB) in northern China in 2021. A total of 144 samples were collected from 18 plantation bases of BJ, SD, and HB, that is, 48 samples were collected from six bases for each province. In order to make the results more representative, four samples were mixed as one for determination. The mature peach samples were gathered according to the standard of GB/T 8855-2008 in China. The samples were stored at 4 °C and analyzed within 24 h after collection.

2.2. The Peach Samples Preparation for Volatile Compounds Analysis

2.2.1. Extraction of Volatile Compounds

The extraction of volatile compounds from the peach samples was performed using the HS-SPME technique which was equipped with carboxen/polydimethylsiloxane fused silica (CAR/PDMS) coated fiber (30 × 0.25 × 0.25) of 2 cm (Supelco, Inc., Bellefonte, PA, USA). A mixture of 3 g peach sample, 0.8 g NaCl, 2 mL distilled water, and 20 uL internal standard of 2-butanol, were placed into a 20 mL vial with a magnetic rotor, and sealed with silicone septa (Sigma Chemical Co., St. Louis, MO, USA) to incubate at 50 °C. After 20 min of equilibration, the CAR/PDMS coated fiber was exposed in the headspace, placed for 40 min before being inserted into the GC-MS injector port at 250 °C for 3 min.

2.2.2. Volatile Compounds Detection

The volatile compounds were identified using GC-MS (Agilent 7890A GC and Agilent 7000B mass selective detector) with the HP-5MS capillary column (30 m × 0.25 mm × 0.25 µm) for separation. The highly pure helium (≥99.99%) was utilized as the carrier gas and the flow rate was set as 1 mL/min. The temperature program of the GC oven was optimized as follows: 40 °C, kept for 3 min, rise to 180 °C at 5 °C/min, held for 3 min, increased to 300 °C at 20 °C/min, held for 3 min. The ionization method was electron ionization (EI). The MS was scanned in an EI mode of 70 eV over the mass range of 40 to 500 with source temperature set as 230 °C and transfer line temperature set as 250 °C. The 2-butanol (10 μL, with a final concentration of 2.70 mg L−1) was set as an internal standard for semi-quantitative analysis [22].

2.2.3. Volatile Compounds Identification

Prior to mass spectral matching, the peaks were analyzed using the deconvolution approach (TraceFinder 5.1) on AMDIS software (www.amdis.net accessed on 1 August 2022). Moreover, the volatile compounds were identified through the NIST library combined with the retention index (RI), as reported. As for the compound identification, the compound was selected according to the minimum ΔRI (the calculated RI minus the library RI). Strictly speaking, a sequence of n-alkanes (from C7 to C40) was adopted as external references to calculate the RI for each volatile. The RI was calculated according to Equation (1), where n was the number of carbon atoms of normal alkanes, ta was the retention time of the volatile compound, tn and tn+1 were the retention time of Cn and Cn+1, respectively.
R I = 100 × n + 100 × t a − t n t n + 1 − t n
The identified volatile compounds were semi-quantified using 2-butanol as an internal standard with Equation (2), where Ac and Ai were the peak areas of the volatile compound and internal standard, respectively. The concentration of internal standard was Ci.
C = A C A i × C i

2.3. OAV Calculation of the Aroma Compounds

The contribution of aroma compounds was estimated using the OAV. The OAV of the aroma compounds was calculated as the ratio of concentration to the odor threshold [23]. The aroma compounds with OAV > 1 were considered as contributing greatly to the aroma profile. It is worth noting that the OAVs presented in this study have some limitations since the odor threshold values were calculated in water. Although it was not the real situation for peach, it could reflect the OAVs of volatile compounds to some extent, as more than 80% of the peach pulp was water.

2.4. Sensory Evaluation

Aroma profiling was carried out using ten assessors (five females and five males, aged from 22 to 30) that had rich experience in sensory evaluations and quantitative description analysis. The assessors evaluated the aroma, gave comments, and scored the peach samples according to the standard of GB12313-1990 in China for traditional sensory evaluation. The traditional sensory evaluation was conducted three times in parallel. Subsequently, the aroma terms were sought to describe the aroma characteristics according to the standard of GB/T10221-2021 in China from the traditional sensory evaluation. On the basis of relevant literature [24], the inexact words were deleted. Ultimately, four sensory attributes and their definitions of peach samples were established, which were demonstrated as follows: methyl acetate for ‘sweet, fruity-like’ note, [(Z)-hex-3-enyl] acetate for ‘green and grassy’ note, hexyl acetate for ‘fruity’ note, and nonan-1-ol for ‘floral’ note.
Based on the established aroma terms, ten assessors evaluated the peach on a digital scale of 0 to 5. In particular, the intensity scales ranged from 0 to 5 for non-existent, just detectable, weak, medium, strong, and very strong, respectively. The samples were in randomized order for three replications.

2.5. Statistic Analysis

The OPLS-DA analysis of HS-SPME/GC-MS data was performed on SIMCA 14.1 (Umetrics, Umeå, Sweden); the PCA analysis was carried out on MatLab (Version 2021b, Natick, MA, USA). The HS-SPME/GC-MS data were preprocessed using pareto scaling (Par) before multivariate analysis. The Par is calculated as Equation (3), where x ¯ denotes the mean value of variables. For the OPLS-DA model, R2 and Q2 indicate the model developed and predicted ability, respectively. The closer the two parameters get to 1, the better is the performance of the model.
P ar = x − x ¯ S D
As an unsupervised multivariate method, PCA is a frequently-used chemometrics tool to reduce the dimension of complicated data to the subspace with fewer principal components (PCs). PCA is adept at revealing the underlying pattern of a dataset in order to seek the similarities and differences. In this work, PCA was utilized to reduce the dimensions and to observe a primary assessment of the variation between classes through the HS-SPME/GC-MS datasets, with Par pretreatment.
The PLS-DA approach is a supervised multivariate statistical method [15], which decomposes the X matrix into the categorical response of Y as described in Equation (4),
X = T P T + E
The OPLS-DA approach is a modification of the PLS-DA method [20], which decomposes the X matrix into three parts according to the categorical response of Y as described in Equation (5),
X = T P P P T + T O P O T + E
where TP and PP are the score and loading of X matrix for prediction, respectively. Likewise, To and Po are the score and loading that is orthogonal to Y. E is the residual matrix of X. Namely, the systematic variation of X is divided into three parts of Y-predictive, Y-orthogonal, and the residual matrix.
The volatile compounds were analyzed on SPSS 26.0 (SPSS Inc., Chicago, IL, USA) for t-test with the significant difference being p < 0.05. When measuring the content of volatile compounds, three replicates were employed, and the results were based on three replicates expressed as the mean ± standard deviation.

3. Results and Discussion

3.1. Characterization the Profile of Volatile Components by HS-SPME/GC-MS

The peach samples from three cultivation regions of BJ, SD, HB were analyzed using HS-SPME/GC-MS. The volatile compounds of peach from the cultivation regions are shown in Table 1. As demonstrated in Table 1, a total of 23 volatile compounds were identified using AMDIS through the deconvolution approach and ΔRI included seven esters, five alcohols, three hydrocarbons, two aldehydes, two benzenes, two ketones, one lactone, and one ether. In fact, the identified volatile compounds number was 17, 17, and 15 for BJ, SD, HB, respectively, which accounted for 70.83%, 73.91%, and 71.43% of the total volatile compounds with the exclusive volatile compounds observed in one or two cultivation regions. Strictly speaking, the volatile compounds of pentan-3-one, methyl hexanoate and phenylmethanol were the specific compounds of BJ, HB, and SD, respectively. In addition, esters, alcohols, and aldehydes were the main volatiles in peach according to the number of compounds. As shown in Table 1, the volatile compounds distribution for the peaches from different cultivation regions were different. Some compounds were in extremely high concentrations among the three planting regions, i.e., benzaldehyde, hexyl acetate, and (Z)-hex-2-enyl acetate (>1500 µg/kg). As for the volatile compounds, some exhibited notably high concentrations (>1500 µg/kg) in individual areas, such as [(Z)-hex-3-enyl] acetate was high only in BJ and (E)-hex-2-enal was high only in SD. Yet, some compounds were distinct among the three groups; for instance, the methyl acetate was significantly higher in SD, which was almost six times as much as BJ and HB, and the concentration of ethylbenzene in BJ was twice as much as SD and HB. In brief, the volatile compounds from BJ, SD, HB have been characterized, and they will be the inputs for the following OAV and chemometrics analysis to explore the critical compounds.

3.2. Critical Aroma Compounds Exploration from the Perspective of OAV

The contribution of individual volatile compounds was evaluated using OAV upon taking its concentration and the corresponding odor threshold into account. Namely, a higher OAV was positively associated with a stronger aroma. The volatile compounds with OAV > 1 were specified as primary aroma compounds. The OAV of volatile compounds which were accompany with an odor description and threshold values are shown in Table 2. A total of seven aroma compounds including three esters of hexyl acetate, [(Z)-hex-3-enyl] acetate and methyl acetate, two aldehydes of (E)-hex-2-enal and benzaldehyde, one alcohol of nonan-1-ol, one lactone of 5-ethyloxolan-2-one were considered as primary aroma-active compounds with OAV > 1 (Table 3), which is consistent with a previous study [23].
Generally, the esters of hexyl acetate and [(Z)-hex-3-enyl] acetate with relatively high OAV values, contributed to ‘fruity’ and ‘green and grassy’ odor notes [25]. In addition, the methyl acetate had ‘sweet and fruity’ odor properties. The three esters were regarded as primary contributors to the peach [26]. The substrates for esters were afforded via the lipoxygenase (LOX) biosynthetic pathway [27]. The alcohol acyl transferase (AAT) was a critical enzyme that catalyzed the conversion of alcohols into aliphatic esters. As for the (E)-hex-2-enal, it presented the notes of ‘green and grassy’ [28]. Admittedly, the unsaturated fatty acids were oxidized into various C6 aldehydes during the homogenization process. Specifically, the (E)-hex-2-enal and nonan-1-ol derived from linoleic acid and linolenic acid underwent dioxygenation of a lipoxygenase-catalyzed reaction by LOX [29]. The nonan-1-ol with the odor characteristics of ‘floral and rose’ was yielded using hydroperoxide lyases (HPL) acting on hydroperoxides, which was metabolized to alcohol using alcohol dehydrogenase (ADH) [30]. As for 5-ethyloxolan-2-one, it contributed to the ‘sweet and coconut’ odors for peach samples. The 5-ethyloxolan-2-one was derived from the fatty acid metabolism through the β-oxidation pathway [31], which was yielded by means of a series of procedures of dehydrogenating, epoxidizing, hydroxylating, shortening by β-oxidation, and esterifying by hydroxyacetyl-coenzyme A [32]. In addition, benzaldehyde, responsible for ‘sweet’ notes [33], gives the ‘sweet’ flavor for the peach. As a matter of fact, the benzaldehyde, originated from amino acids of phenylpropanoid compounds [34], was synthesized from phenylalanine using aromatic amino acid decarboxylases. Moreover, the phenylalanine was biosynthesised through shikimate or arogenate pathways, and deaminated via the enzyme of phenylalanine ammonialyase [31]. In the previous study, we speculated that the generation of benzaldehyde might be associated with the enzyme of chorismate synthase [35].
As could be seen from Table 3, the richness of the aroma characteristics were different among the three cultivation regions, which was mainly due to the different combinations of the primary aroma compounds and the distinct contributions of each compound endowed with unique odors. Consequently, seven aroma compounds of methyl acetate, (E)-hex-2-enal, benzaldehyde, [(Z)-hex-3-enyl] acetate, 5-ethyloxolan-2-one, nonan-1-ol, and hexyl acetate were screened and ascertained as the primary aroma-active compounds of the three cultivation regions of BJ, SD, HB, in accordance with the OAV evaluation.

3.3. Critical Aroma Compounds Exploration from the Perspective of Chemometrics Methods

3.3.1. Volatile Components Modeling with PCA and OPLS-DA Approaches

The PCA algorithm, one of the most frequently used chemometrics approaches, was used to project the HS-SPME/GC-MS data set to a lower dimensional space and to compare the differences among the three groups. To develop the PCA model, the first two PCs were selected with a total variance of 84.4%. Specifically, the first principal component (PC1) accounted for 50.9% of the total variance while the second principal component (PC2) contributed 33.5%. The overview of the two-dimensional score (Figure S1) plot was clearly segmented, which implied that there were indications of grouping among different cultivation regions. Moreover, the score plot also reflected the impact of cultivation regions on peach aroma composition. The results indicated that BJ and HB had tiny difference according to PC1, while they could be separated along the PC2 dimension. In contrast, there were striking differences between SD vs. BJ as well as SD vs. HB along the PC1 dimension. It could be concluded that the discrepancy between SD vs. BJ and SD vs. HB were significantly greater than that between BJ vs. HB. It might be deduced that HB vs. BJ were relatively near to each other compared with SD. This result suggests that the identified aroma using the HS-SPME/GC-MS-based method could be utilized in peach discrimination. To uncover the underlying difference among BJ, SD, and HB, the supervised method OPLS-DA was applied.
In order to discover the differential critical aroma among the three cultivation regions of the three pair-wise models of BJ vs. SD, BJ vs. HB, SD vs. HB, the supervised chemometrics methods of OPLS-DA was performed for pairwise comparison. In addition, the permutation test was carried out to estimate whether the performance of the OPLS-DA model was acceptable [36] by comparing the multiple random permutated y variables and the original y variables through random sampling. Seven-fold cross-validation was executed to verify the accuracy and reliability of the model [37] on the three pair-wise models with the evaluation parameters of the total explaining the X-variance R2(X), the model developing the ability of R2(Y), and the cross-validated predictive ability Q2(Y). As for the three models (Figure 1a–f), they were all established with one predictive and one orthogonal component. The R2(Y) and Q2(Y) for the three models were all satisfied with R2(Y) > 0.980 and Q2(Y) > 0.969. Actually, the permutation test was connected with the slope of regression plot and the intercept on the y axis. A greater slope together with a smaller intercept would yield a more reliable model with better quality and predictive ability. The results of the permutation test with 200 iterations for the three models are demonstrated in Figure 1 and Table 4. Moreover, the Q2 and R2 for the three OPLS-DA models were all higher than the permutation tests (Table 4), which indicate that the models were suitable for discriminating and seeking potentially critical aromas. The results of the three permutation tests revealed that the classification models were not over-fitting [38]. As postulated above, the three robust OPLS-DA classification models were developed with an acceptable separation ability for exploring the potential critical aroma compounds.

3.3.2. Critical Aroma Components Exploration by S-Plot, Jack-Knifing Confidence Interval, VIP, t-Test, and FC

To reveal the potential differential aroma compounds among the three cultivation regions, the following restrictions were taken into consideration. Firstly, the S-plot together with the jack-knifed confidence intervals were employed for potential critical aroma exploring [21]. Secondly, the VIP, expressing the contribution of each variable, was utilized to seek the potential aroma with the threshold greater than 1 [39]. Thirdly, the potential compounds were characterized on the basis of the p values (<0.05) and FC (|log2 (FC)| > 1). The S-plot manifested the covariance and the correlation between the X variables and the predictive score [21]. Both the contribution (covariance, Cov(t, X)) and the reliability (correlation, Corr(t, X)) are shown in the scatter plot. As a matter of fact, the two aspects of high magnitude and high reliability acted together to demonstrate a low risk of spurious correlations. The calculation of Corr(t, X) corresponded to Cov(t, X) and the standard deviation of the t and X variable ( C orr ( t , x ) = C o v ( t , X ) σ t σ x ). In this study, the criteria for the S-plot was set with |p(cov)| ≥ 0.1 and |p(corr)| ≥ 0.5. The visualization of candidate aroma compounds to distinguish the pair-wised comparison of BJ vs. SD, BJ vs. HB, SD vs. HB together with the jack-knifing confidence intervals are demonstrated in Figure 2. The jack-knifing confidence interval ( C I = S E C V t ( α , d f ) ) was determined using SECV, the standard error for each loading, t, the statistical value with α of 0.05, and df, the degrees of freedom.
As shown in Figure 2a of the S-plot, eight compounds (solid pink circle) were identified as the potential differential compounds that were responsible for the paired comparison of BJ and SD. In the loading plot (Figure 2b), the highlighted potential critical aroma compounds were matched with the S-plot one-by-one. As demonstrated in Figure 2a,b, the confidence interval for all potentially critical aroma compounds in the loading plot was relatively low, which indicated that all the eight potentially critical aroma compounds were located in credible regions. With regards to the VIP, four compounds were filtered as candidates to characterize the differences with VIP > 1 (Table S1). As a consequence, four compounds of methyl acetate, benzaldehyde, (E)-hex-2-enal, and [(Z)-hex-3-enyl] acetate were considered as the potential critical compounds for BJ and SD by taking a comprehensive consideration of the evaluation criterion, including p-values and FC. Analogously, as for the paired comparison of BJ and HB, eight compounds were screened as potentially critical compounds by means of the S-plot (Figure 2c). It can be deduced that the 1,2-xylene (with blue arrow) might give an ambiguous response for the two discriminated groups (Figure 2d). In addition, the raw data (Figure S2a) further confirmed that the two regions were slightly overlapped. As far as the VIP, four compounds were found as critical aroma compounds (Table S1). On the whole, four compounds of benzaldehyde, (E)-hex-2-enal, [(Z)-hex-3-enyl] acetate, and 5-ethyloxolan-2-one were regarded as the potentially critical aroma compounds to distinguish BJ and HB using S-plot, jack-knifing confidence, VIP, p values and FC. Likewise, for the pair-wise comparison of SD and HB, six compounds were screened using the S-plot (Figure 2e). It could be speculated that the benzaldehyde and phenylmethanol (blue arrow) with large confidence intervals might be situated in uncertain regions (Figure 2f). However, the raw data (Figure S2b,c) ascertained the benzaldehyde did not seem to be vital. Table S1 details the identified discriminant aroma with the relative influence of VIP values on the model. Accordingly, four aroma compounds were considered as candidate critical aroma compounds. After taking p values and FC into account, three aroma compounds of methyl acetate, (E)-hex-2-enal, and 5-ethyloxolan-2-one were ultimately determined as the potential critical compounds to distinguish SD and HB.
To sum up, the critical aroma components of the three groups of BJ vs. SD, BJ vs. HB, and SD vs. HB were ascertained with an S-plot together with the jack-knifed confidence intervals and VIP through the paired OPLS-DA approach in cooperation with the p values and FC. Overall, the five primary aroma compounds methyl acetate, (E)-hex-2-enal, benzaldehyde, [(Z)-hex-3-enyl] acetate, and 5-ethyloxolan-2-one were specified as differential biomarkers for the three groups.

3.3.3. Critical Compounds Characterization upon the Strategy of SUS-Plot

SUS-plot, the scatter plots for correlation from the predictive component of Corr(tp, X) derived from two individual OPLS models, was produced to give a visualization to determine the shared and unique structures of multiple groups [40]. It was of great interest to find both shared and unique compounds among the multi-model for an overall understanding of the difference as a comparison of the models [21]. Specifically, the compounds on the diagonal represented the shared compounds that had the same effect in both models. Alternatively, the compounds that were located far away from the diagonal symbolized the unique compounds that were specific to either of the two models.
It is worth mentioning that the five compounds acted as the critical potential aroma for three paired comparisons of BJ vs. SD and BJ vs. HB, SD vs. BJ and SD vs. HB, HB vs. BJ and HB vs. SD. To obtain a clear overview on the SUS plots, the five aroma compounds were visualized, see Figure 3. The SUS-plot of the two models of BJ vs. SD and BJ vs. HB are demonstrated on Figure 3a. As seen from Figure 3a, all the five potential crucial compounds were regarded as the shared critical aroma along the diagonal line of SD and HB. With respect to the two models of SD vs. BJ and SD vs. HB (Figure 3b), the shared aroma compounds of HB and BJ spanned four compounds (methyl acetate, (E)-hex-2-enal, benzaldehyde, 5-ethyloxolan-2-one), while one ([(Z)-hex-3-enyl] acetate) was ascribed as a unique feature (concentration significantly higher) to BJ with the p(corr) close to 0 for HB. As far as the models of HB vs. BJ and HB vs. SD (Figure 3c), four critical compounds of methyl acetate, (E)-hex-2-enal, benzaldehyde, 5-ethyloxolan-2-one were considered as the common characteristics of SD and BJ, whereas [(Z)-hex-3-enyl] acetate was determined as the particular compound for BJ with the p(corr) close to 0 for SD.
The box plots of the five critical aroma compounds are demonstrated in Figure S3. The concentration of two compounds of methyl acetate, (E)-hex-2-enal and benzaldehyde were significantly higher in SD with the concentration of 602.58, 2688.33, and 4858 µg/kg (Figure S3a–c). As seen in Figure S3d, the concentration of [(Z)-hex-3-enyl] acetate was remarkably higher in BJ, being 1567.83 µg/kg compared with HB and SD. Moreover, 5-ethyloxolan-2-one was significantly higher in HB compared to BJ and SD with a concentration of 384.83 µg/kg (Figure S3e). In conclusion, as shown in the SUS plots and box-plots, the five compounds of methyl acetate, (E)-hex-2-enal, benzaldehyde, [(Z)-hex-3-enyl] acetate, and 5-ethyloxolan-2-one were identified as the critical markers for peach geographical origin traceability in BJ, HB, and SD.

3.4. Critical Aroma Compounds for Qualitative Discrimination Model and Potential Chemical Basis of Odors via Sensory Evaluation

3.4.1. Critical Aroma Compound: Qualitative Discrimination Model Based on PLS-DA

The five critical aroma compounds of methyl acetate, (E)-hex-2-enal, benzaldehyde, [(Z)-hex-3-enyl] acetate, and 5-ethyloxolan-2-one were characterized as critical components by taking both the OAV and chemometrics methods into account. It is universally acknowledged that the PLS-DA approach is powerful in dealing with multi-classification problems; therefore, the qualitative analysis model for BJ, HB and SD was developed by the PLS-DA algorithm with the five critical aroma compounds to investigate whether the three cultivation regions could be distinguished using the determined primary compounds through the multi-classification model.
The first two latent variables (LVs) were selected to develop the multi-classification model according to the seven-fold cross-validation procedure with the total covariance of 0.972. As a result, the model developing ability of R2(Y) and predicted ability of Q2(Y) were 0.954 and 0.951, respectively. As obtained from the confusion matrix (Figure 4a), all the samples were located on the main diagonal, the classification accuracy for the three cultivation regions were all 100%. The PLS-DA model gained a favorable classification capacity according to accuracy, which was sufficient to suggest that the PLS model could feasibly distinguish the three cultivation regions using the five critical markers.
To sum up, the five critical markers acquired from the shared compounds from OAV and chemometrics methods were enough to develop the PLS-DA model for discriminating the three cultivation regions. Moreover, it can be concluded that the screened markers were overwhelmingly practicable to enable geographical origin traceability. It was also sufficient to indicate the marker explored methods of OAV, coupled with chemometrics approach, work effectively.

3.4.2. Critical Aroma Compound: Potential Chemical Basis of Odors by Sensory Evaluation

The peach samples were used for a sensory evaluation in order to produce an overview of the flavors prevalent among the three cultivation regions. The standard deviation (σ) was applied to assess the stability of the peach flavor. In general, a larger σ indicates a greater difference in flavor. As demonstrated in Figure 4b, the ‘sweet’, ‘fruity’, ‘green and grassy’ odors were considered as the dominating aroma in peach. Among the four typical aroma attributes, ‘sweet’ and ‘floral’ exhibited the minimum variation (both with σ = 0.58), whereas, the σ for ‘green’ (σ = 1.15) and ‘grassy’ (σ = 1.00) were relatively larger. The intensity of the ‘sweet’ and ‘fruity’ odor in SD was stronger when compared with BJ and HB, which might have resulted from higher concentrations of methyl acetate, hexyl acetate, and benzaldehyde (Table 1). The ‘floral’ odor in BJ might be due to the high concentration of nonan-1-ol, whose concentration in BJ was three and four times higher than that in SD and HB. In addition, the peach in BJ exhibited the highest intensity of ‘green- and grassy-like’ odor, which might be due to the high concentration of [(Z)-hex-3-enyl] acetate (1567.83 µg/kg) compared with HB (524.58 µg/kg) and SD (482.25 µg/kg). Accordingly, BJ and SD had richer odor characteristics in contrast with HB.
As a consequence, it can be concluded that the sensory evaluation results might be closely related to the five screened potential critical aroma compounds with respect to the main flavors of ‘sweet’, ‘fruity’, ‘green and grassy’.

4. Conclusions

In this study, the volatile profile of peach was characterized using HS-SPME/GC-MS. Five aroma compounds of methyl acetate, (E)-hex-2-enal, benzaldehyde, [(Z)-hex-3-enyl] acetate, and 5-ethyloxolan-2-one were identified as differentially critical volatile components for the three cultivation regions of BJ, HB, and SD by taking into account both the primary aroma-active compounds with OAV > 1 and the critical compounds explored through chemometrics methods. The potentially significant compounds found using chemometrics methods were determined on the basis of significant difference (p < 0.05), FC, S-plot analysis, jack-knifing confidence interval, VIP, and the SUS plots. As visualized in the SUS-plots, the critical compound of [(Z)-hex-3-enyl] acetate was identified as the unique feature for BJ, whereas the compounds of methyl acetate, (E)-hex-2-enal, benzaldehyde, and 5-ethyloxolan-2-one were ascertained as the shared features. Moreover, the multi-classification model gave an outstanding performance, with an accuracy of 100%, and enabled the three cultivation regions to be distinguished using the five critical markers. The sensory evaluation results might be closely related to the five explored critical aroma compounds with respect to the main flavors of ‘sweet’, ‘fruity’, ‘green and grassy’. This research on potential aroma is extremely important for geographical origin traceability. Moreover, it also creates a guide to evaluate peach quality and provides a basis for understanding the formation of the characteristic aroma compounds in peaches

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/foods12040837/s1. Figure S1: The score plot of PCA analysis of BJ, SD, and HB. Figure S2: Raw data of the uncertain compounds with larger confidence intervals. (a) 1,2-xylene in the model of BJ vs. HB; (b) benzaldehyde in the model of SD vs. HB; (c) phenylmethanol in the model of SD vs. HB. Figure S3: The box plots of the five critical aroma compounds. (a) methyl acetate; (b) (E)-hex-2-enal; (c) benzaldehyde; (d) [(Z)-hex-3-enyl] acetate; (e) 5-ethyloxolan-2-one. Table S1: The screened critical aroma compounds for pair-wise comparisons of BJ, SD, and HB.

Author Contributions

Q.L.: data curation, formal analysis, investigation, methodology, writing-original draft; B.L.: data curation, formal analysis; R.Z.: data curation, formal analysis; S.L.: data curation, formal analysis; S.Y.: data curation, formal analysis; Y.L.: conceptualization, writing-review and editing; J.L.: methodology, investigation, data curation, funding acquisition, writing-review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. 32202183), and Key Laboratory of Tropical Fruits and Vegetables Quality and Safety for State Market Regulation (No. KF-2022005), and Central Public-interest Scientific Institution Basal Research Fund (No. S2022JBKY-15), and Key Research and Development Project of Hainan Province (ZDYF2020078).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All data presented within the article are available upon reasonable request from the corresponding authors.

Acknowledgments

The authors gratefully acknowledge the funds.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Wu, H.; Xu, Y.; Wang, H.; Miao, Y.; Li, C.; Zhao, R.; Shi, X.; Wang, B. Physicochemical characteristics, antioxidant activities, and aroma compound analysis of seven peach cultivars (Prunus persica L. Batsch) in shihezi, xinjiang. Foods 2022, 11, 2944. [Google Scholar] [CrossRef] [PubMed]
  2. Medeiros, A.; Tavares, E.; Bolini, H.M.A. Descriptive sensory profile and consumer study impact of different nutritive and non-nutritive sweeteners on the descriptive, temporal profile, and consumer acceptance in a peach juice matrix. Foods 2022, 11, 244. [Google Scholar] [CrossRef] [PubMed]
  3. Wang, X.; Guo, W.; Sun, B.; Li, H.; Zheng, F.; Li, J.; Meng, N. Characterization of key aroma-active compounds in two types of peach spirits produced by distillation and pervaporation by means of the sensomics approach. Foods 2022, 11, 2598. [Google Scholar] [CrossRef] [PubMed]
  4. Leccese, A.; Viti, R.; Bartolini, S. The effect of solvent extraction on antioxidant properties of apricot fruit. Open Life Sci. 2011, 6, 199–204. [Google Scholar] [CrossRef]
  5. Xu, S.; Zhan, P.; Tian, H.; Wang, P. The presence of kiwifruit columella affects the aroma profiles of fresh and thermally treated kiwifruit juice. LWT Food Sci. Technol. 2022, 165, 113756. [Google Scholar] [CrossRef]
  6. Dabbou, S.; Lussiana, C.; Maatallah, S.; Gasco, L.; Hajlaoui, H.; Flamini, G. Changes in biochemical compounds in flesh and peel from Prunus persica fruits grown in Tunisia during two maturation stages. Plant Physiol. Biochem. 2016, 100, 1–11. [Google Scholar] [CrossRef]
  7. Zhang, Y.; Liu, W.; Zhang, B.; Zhang, Y.; Cai, Z.; Song, H.; Ma, R.; Yu, M. Analysis of volatile compounds and their potential regulators in four high-quality peach (Prunus persica L.) cultivars with unique aromas. LWT Food Sci. Technol. 2022, 160, 113195. [Google Scholar] [CrossRef]
  8. Marsili, R. Flavor, Fragrance, and Odor Analysis; CRC Press: Boca Raton, FL, USA, 2001; Volume 115. [Google Scholar]
  9. Chen, C.; Liu, Z.; Yu, H.; Lou, X.; Huang, J.; Yuan, H.; Wang, B.; Xu, Z.; Tian, H. Characterization of six lactones in cheddar cheese and their sensory interactions studied by odor activity values and feller’s additive model. J. Agric. Food Chem. 2021, 70, 301–308. [Google Scholar] [CrossRef] [PubMed]
  10. Yang, Y.; Zhu, H.; Chen, J.; Xie, J.; Shen, S.; Deng, Y.; Zhu, J.; Yuan, H.; Jiang, Y. Characterization of the key aroma compounds in black teas with different aroma types by using gas chromatography electronic nose, gas chromatography-ion mobility spectrometry, and odor activity value analysis. LWT Food Sci. Technol. 2022, 163, 113492. [Google Scholar] [CrossRef]
  11. Riu-Aumatell, M.; Castellari, M.; López-Tamames, E.; Galassi, S.; Buxaderas, S. Characterisation of volatile compounds of fruit juices and nectars by HS/SPME and GC/MS. Food Chem. 2004, 87, 627–637. [Google Scholar] [CrossRef]
  12. Butkhup, L.; Jeenphakdee, M.; Jorjong, S.; Samappito, S.; Samappito, W.; Chowtivannakul, S. HS-SPME/GC-MS analysis of volatile aromatic compounds in alcohol related beverages made with mulberry fruits. Food Sci. Biotechnol. 2011, 20, 1021–1032. [Google Scholar] [CrossRef]
  13. Giannetti, V.; Mariani, M.B.; Mannino, P.; Marini, F. Volatile fraction analysis by HS-SPME/GC-MS and chemometric modeling for traceability of apples cultivated in the Northeast Italy. Food Control 2017, 78, 215–221. [Google Scholar] [CrossRef]
  14. Liu, Y.; Qian, X.; Xing, J.; Li, N.; Li, J.; Su, Q.; Chen, Y.; Zhang, B.; Zhu, B. Accurate determination of 12 lactones and 11 volatile phenols in nongrape wines through headspace-solid-phase microextraction (HS-SPME) combined with high-resolution gas chromatography-orbitrap mass spectrometry (GC-Orbitrap-MS). J. Agric. Food Chem. 2022, 70, 1971–1983. [Google Scholar] [CrossRef] [PubMed]
  15. Wold, S.; Sjöström, M.; Eriksson, L. PLS-Regression: A basic tool of chemometrics. J. Chemom. 2001, 58, 109–130. [Google Scholar] [CrossRef]
  16. Wold, S.; Esbensen, K.; Geladi, P. Principal component analysis. Chemom. Intell. Lab. Syst. 1987, 2, 37–52. [Google Scholar] [CrossRef]
  17. Lee, L.C.; Jemain, A.A. On overview of PCA application strategy in processing high dimensionality forensic data. Microchem. J. 2021, 169, 106608. [Google Scholar] [CrossRef]
  18. Han, L.; Cheng, Y.; Zhang, T.; Zhou, Q.; Zhang, W.; Li, Y.; Li, G. Targeted metabolomics with a chemometric study of oxygenated heterocyclicaglycones as a tool for preliminary authenticity assessment of orange and grapefruit juices. Front. Nutr. 2022, 9, 897982. [Google Scholar] [CrossRef] [PubMed]
  19. Soares, L.C.; Alves, J.D.O.; Linhare, L.A.; Egreja Filho, F.B.; Fontes, M.P. Vulnerability of tropical soils to heavy metals: A PLS-DA classification model for lead. Microchem. J. 2017, 133, 258–264. [Google Scholar] [CrossRef]
  20. Trygg, J.; Wold, S. Orthogonal projections to latent structures (O-PLS). J. Chemom. 2002, 16, 119–128. [Google Scholar] [CrossRef]
  21. Wiklund, S.; Johansson, E.; Sjöström, L.; Mellerowicz, E.J.; Edlund, U.; Shockcor, J.P.; Gottfries, J.; Moritz, T.; Trygg, J. Visualization of GC/TOF-MS-based metabolomics data for identification of biochemically interesting compounds using OPLS class models. Anal. Chem. 2008, 80, 115–122. [Google Scholar] [CrossRef]
  22. Budiene, J.; Guclu, G.; Oussou, K.F.; Kelebek, H.; Selli, S. Elucidation of volatiles, anthocyanins, antioxidant and sensory properties of cv. Caner Pomegranate (Punica granatum L.) juices produced from three juice extraction methods. Foods 2021, 10, 1497. [Google Scholar] [CrossRef] [PubMed]
  23. Burdock, G.A. Fenaroli’s Handbook of Flavor Ingredients; CRC Press: Boca Raton, FL, USA, 2016. [Google Scholar]
  24. Zhu, J.; Xiao, Z. Characterization of the key aroma compounds in peach by gas chromatography-olfactometry, quantitative measurements and sensory analysis. Eur. Food Res. Technol. 2019, 245, 129–141. [Google Scholar] [CrossRef]
  25. Pino, J.A.; Trujillo, R. Characterization of odour-active compounds of sour guava (Psidium acidum [DC.] Landrum) fruit by gas chromatography-olfactometry and odour activity value. Flavour Fragr. J. 2021, 36, 207–212. [Google Scholar] [CrossRef]
  26. Zhou, H.; Yu, Z.; Ye, Z.; Su, M. Multiplex analyses of the changes of aromatic compounds during the development of peach fruit using GC-MS and iTRAQ proteomic techniques. Sci. Hortic. 2018, 236, 96–105. [Google Scholar] [CrossRef]
  27. El Hadi, M.A.M.; Zhang, F.J.; Wu, F.F.; Zhou, C.H.; Tao, J. Advances in fruit aroma volatile research. Molecules 2013, 18, 8200–8229. [Google Scholar] [CrossRef]
  28. Liu, W.; Zhang, Y.; Ma, R.; Yu, M. Comparison of aroma trait of the white-fleshed peach ‘Hu Jing Mi Lu’ and the yellow-fleshed peach ‘Jin Yuan’ based on odor activity value and odor characteristics. Horticulturae 2022, 8, 245. [Google Scholar] [CrossRef]
  29. Feussner, I.; Wasternack, C. The lipoxygenase pathway. Annu. Rev. Plant Biol. 2002, 53, 275–297. [Google Scholar] [CrossRef]
  30. Zhang, J.; Zhao, J.; Xu, Y.; Liang, J.; Chang, P.; Yan, F.; Li, M.; Liang, Y.; Zou, Z. Genome-wide association mapping for tomato volatiles positively contributing to tomato flavor. Front. Plant Sci. 2015, 6, 1042–1053. [Google Scholar] [CrossRef] [Green Version]
  31. Yahia, E.M.; Carrillo-Lopez, A. Postharvest Physiology and Biochemistry of Fruits and Vegetables; Woodhead Publishing: Cambridge, UK, 2018. [Google Scholar]
  32. Peng, B.; Yu, M.; Zhang, B.; Xu, J.; Ma, R. Differences in PpAAT1 activity in high-and low-aroma peach varieties affect γ-decalactone production. Plant Physiol. 2020, 182, 2065–2080. [Google Scholar] [CrossRef]
  33. Yu, H.; Xie, T.; Xie, J.; Chen, C.; Ai, L.; Tian, H. Aroma perceptual interactions of benzaldehyde, furfural, and vanillin and their effects on the descriptor intensities of Huangjiu. Food Res. Int. 2020, 129, 108808. [Google Scholar] [CrossRef]
  34. Dudareva, N.; Klempien, A.; Muhlemann, J.K.; Kaplan, I. Biosynthesis, function and metabolic engineering of plant volatile organic compounds. New Phytol. 2013, 198, 16–32. [Google Scholar] [CrossRef] [PubMed]
  35. Li, Q.; Yang, S.; Zhang, R.; Liu, S.; Zhang, C.; Li, Y.; Li, J. Characterization of honey peach (Prunus persica (L.) Batsch) aroma variation and unraveling the potential aroma metabolism mechanism through proteomics analysis under abiotic stress. Food Chem. 2022, 386, 132720. [Google Scholar] [CrossRef] [PubMed]
  36. Andrew Clayton, T.; Lindon, J.C.; Cloarec, O.; Antti, H.; Charuel, C.; Hanton, G.; Provost, J.P.; Le Net, J.L.; Baker, D.; Walley, R.J.; et al. Pharmaco-metabonomic phenotyping and personalized drug treatment. Nature 2006, 440, 1073–1077. [Google Scholar] [CrossRef] [PubMed]
  37. Michell, A.W.; Mosedale, D.; Grainger, D.J.; Barker, R.A. Metabolomic analysis of urine and serum in Parkinson’s disease. Metabolomics 2008, 4, 191–201. [Google Scholar] [CrossRef]
  38. Lapins, M.; Eklund, M.; Spjuth, O.; Prusis, P.; Wikberg, J.E. Proteochemometric modeling of HIV protease susceptibility. BMC Bioinform. 2008, 9, 181. [Google Scholar] [CrossRef] [Green Version]
  39. Wei, J.; Xie, G.; Zhou, Z.; Shi, P.; Qiu, Y.; Zheng, X.; Chen, T.; Su, M.; Zhao, A.; Jia, W. Salivary metabolite signatures of oral cancer and leukoplakia. Int. J. Cancer 2011, 129, 2207–2217. [Google Scholar] [CrossRef]
  40. Madala, N.E.; Piater, L.A.; Steenkamp, P.A.; Dubery, I.A. Multivariate statistical models of metabolomic data reveals different metabolite distribution patterns in isonitrosoacetophenone-elicited Nicotiana tabacum and Sorghum bicolor cells. SpringerPlus 2014, 3, 254. [Google Scholar] [CrossRef] [Green Version]
Foods 12 00837 g001 550
Figure 1. The OPLS-DA analysis of the HS-SPME/GC-MS data. (a) score plot for BJ vs. SD; (b) validation plot of 200 permutation tests for BJ vs. SD; (c) score plot for BJ vs. HB; (d) validation plot of 200 permutation tests for BJ vs. HB; (e) score plot for SD vs. HB; (f) validation plot of 200 permutation tests for SD vs. HB. The slope of R2 > 0 together with the intercept of Q2 on the y-axis < 0 indicating a valid model.
Figure 1. The OPLS-DA analysis of the HS-SPME/GC-MS data. (a) score plot for BJ vs. SD; (b) validation plot of 200 permutation tests for BJ vs. SD; (c) score plot for BJ vs. HB; (d) validation plot of 200 permutation tests for BJ vs. HB; (e) score plot for SD vs. HB; (f) validation plot of 200 permutation tests for SD vs. HB. The slope of R2 > 0 together with the intercept of Q2 on the y-axis < 0 indicating a valid model.
Foods 12 00837 g001
Foods 12 00837 g002 550
Figure 2. The S-plot and loading plot with jack-knifed confidence intervals. (a) S-plot of BJ vs. SD; (b) loading plot of BJ vs. SD; (c) S-plot of BJ vs. HB; (d) loading plot BJ vs. HB; (e) S-plot of SD vs. HB; (f) loading plot of SD vs. HB. The highlighted pink solid circle in S-plot represents the potential metabolites (|p(cov)| ≥ 0.1 and |p(corr)| ≥ 0.5); the highlighted pink rectangle with confidence interval represents the corresponding potential aroma in S-plot; the blue arrow in the loading plot represents the compound with high confident interval; the blue circle in the S-plot represents the corresponding compound in the loading plot.
Figure 2. The S-plot and loading plot with jack-knifed confidence intervals. (a) S-plot of BJ vs. SD; (b) loading plot of BJ vs. SD; (c) S-plot of BJ vs. HB; (d) loading plot BJ vs. HB; (e) S-plot of SD vs. HB; (f) loading plot of SD vs. HB. The highlighted pink solid circle in S-plot represents the potential metabolites (|p(cov)| ≥ 0.1 and |p(corr)| ≥ 0.5); the highlighted pink rectangle with confidence interval represents the corresponding potential aroma in S-plot; the blue arrow in the loading plot represents the compound with high confident interval; the blue circle in the S-plot represents the corresponding compound in the loading plot.
Foods 12 00837 g002
Foods 12 00837 g003 550
Figure 3. The SUS-plots among BJ, HB, and SD. (a) BJ vs. SD and BJ vs. HB; (b) SD vs. BJ and SD vs. HB; (c) HB vs. BJ and HB vs. SD. The aroma compounds close to the diagonal line are equally affected in both lines and considered as common potential biomarkers; the aroma compounds close to the x axis are specific for cultivation in x axis and close to y axis are specific for cultivation in y axis.
Figure 3. The SUS-plots among BJ, HB, and SD. (a) BJ vs. SD and BJ vs. HB; (b) SD vs. BJ and SD vs. HB; (c) HB vs. BJ and HB vs. SD. The aroma compounds close to the diagonal line are equally affected in both lines and considered as common potential biomarkers; the aroma compounds close to the x axis are specific for cultivation in x axis and close to y axis are specific for cultivation in y axis.
Foods 12 00837 g003
Foods 12 00837 g004 550
Figure 4. (a) The confusion matrix of PLS-DA model; (b) the sensory evaluation of the aroma profiles of peach from three cultivation regions of BJ, SD, and HB.
Figure 4. (a) The confusion matrix of PLS-DA model; (b) the sensory evaluation of the aroma profiles of peach from three cultivation regions of BJ, SD, and HB.
Foods 12 00837 g004
Table
Table 1. The volatile compounds of peach from three cultivation regions of BJ, SD, and HB. nd represents not detected.
Table 1. The volatile compounds of peach from three cultivation regions of BJ, SD, and HB. nd represents not detected.
No.Volatile CompoundsRT (min)Concentration (μg/kg)
BJSDHB
1Isocyanic acid4.4751.92 ± 6.5287.67 ± 5.3592.08 ± 7.68
2Methyl acetate4.93123.67 ± 12.57602.58 ± 86.4097.50 ± 8.99
32-Methylpropane5.974.83 ± 0.83ndnd
4Pentan-3-one6.4513.17 ± 1.70ndnd
52-Butan-2-yloxybutane8.39129.67 ± 11.48116.92 ± 23.45188.00 ± 10.82
61-(Furan-2-yl)pentan-1-one8.733.67 ± 0.9814.00 ± 1.866.08 ± 1.62
7(E)-Hex-2-enal9.72606.17 ± 52.162688.33 ± 331.8365.75 ± 8.85
85-Methylhept-3-yne9.77nd5.33 ± 1.15nd
9Ethylbenzene9.85166.08 ± 15.3787.00 ± 7.8388.83 ± 7.32
101,2-Xylene10.18264.67 ± 35.01339.33 ± 46.64425.92 ± 16.56
11Methyl hexanoate11.24ndnd45.00 ± 7.77
12Benzaldehyde12.421546.42 ± 203.314858.00 ± 178.883628.92 ± 266.85
13(2R)-2-Hydroxy-3-methylbutanenitrile12.66ndnd12.08 ± 1.62
14[(Z)-Hex-3-enyl] acetate13.251567.83 ± 248.83482.25 ± 95.83524.58 ± 98.45
15Hexyl acetate13.401915.08 ± 161.472290.92 ± 144.761782.25 ± 294.10
16(Z)-Hex-2-enyl acetate13.451821.17 ± 185.551826.42 ± 161.331592.25 ± 308.70
17(E)-Hex-3-en-1-yne13.757.67 ± 1.56ndnd
18Phenylmethanol14.27nd196.42 ± 12.44nd
195-Ethyloxolan-2-one14.49135.50 ± 32.1011.92 ± 3.29384.83 ± 45.06
20Octan-1-ol14.8535.67 ± 8.97ndnd
21Nonan-1-ol17.19143.42 ± 26.4944.58 ± 6.2938.17 ± 10.31
22[(E)-Hex-2-enyl] 3-Methylbutanoate18.79nd28.00 ± 3.30nd
232,3-Dimethylbut-3-en-2-ol21.43nd15.42 ± 2.61nd
Table
Table 2. Odor description, odor threshold value of the identified volatiles in peach from three cultivation regions of BJ, SD, and HB in northern China.
Table 2. Odor description, odor threshold value of the identified volatiles in peach from three cultivation regions of BJ, SD, and HB in northern China.
No.Volatile CompoundsRI aRI bOdor Description cThreshold (μg/L) dID eCategory
1Isocyanic acid /n.fMS, RIacid
2Methyl acetate sweet, fruity3MS, RI, Sester
32-Methylpropane605-/n.fMS, RI, Shydrocarbon
4Pentan-3-one703700ethereal acetone40MS, RIketone
5Sec-Butyl ether801-/n.fMS, RI, Sethers
61-(2-Furyl) pentan-1-one815-/n.fMS, RIketone
7(E)-Hex-2-enal858854green, leaf30MS, RI, Saldehyde
85-Methylhept-3-yne860-/n.fMS, RIhydrocarbon
9Ethylbenzene863864/1200MS, RIbenzene
101,2-Xylene877-geranium450MS, RI, Sbenzene
11Methyl hexanoate922924fruity70MS, RI, Sester
12Benzaldehyde970970sweet750MS, RIaldehyde
13(2R)-2-Hydroxy-3-methylbutanenitrile980-/n.fMS, RIalcohol
14[(Z)-Hex-3-enyl] acetate10041009green, leaf8MS, RIester
15Hexyl acetate10111010fruity10MS, RI, Sester
16(Z)-Hex-2-enyl acetate10131005/n.fMS, RI, Sester
173-Hexen-1-yne1025-/n.fMS, RIhydrocarbon
18Phenylmethanol10471042floral, rose2500MS, RI, Salcohol
195-Ethyloxolan-2-one10561056sweet, coconut50MS, RIlactone
20Octan-1-ol10711068green130MS, RIalcohol
21Nonan-1-ol11711172floral rose50MS, RI, Salcohol
22[(E)-Hex-2-enyl] 3-methylbutanoate12421244/n.fMS, RIester
232,3-Dimethylbut-3-en-2-ol1366-/n.fMS, RIalcohol
a RI, retention indexes was calculated refereed to the retention time of C7–C40 n-alkanes under the same condition. b RI, retention indexes was found in the NIST Chemistry WebBook, SRD 69, URL: https://webbook.nist.gov/chemistry/cas-ser/ accessed on 26 November 2022. c Odor description found in the literature with database (Flavornet; The LRI and Odor Database). d Odor threshold values from (Burdock, 2010). e Identification method. MS, identification based on the NIST 2017 mass spectral database; RI, retention index; S, the compounds were identified using authentic standard compounds. n.f, data were not found in the literature.
Table
Table 3. The volatile compounds with OAVs > 1 in peach from three cultivation regions. OAV, odor activity value. BJ, SD, and HB represent Beijing, Shandong and Hebei provinces, respectively.
Table 3. The volatile compounds with OAVs > 1 in peach from three cultivation regions. OAV, odor activity value. BJ, SD, and HB represent Beijing, Shandong and Hebei provinces, respectively.
Volatile CompoundsOAV Values
BJSDHB
Methyl acetate41.22200.9032.50
(E)-Hex-2-enal20.2189.612.19
Benzaldehyde2.066.484.84
[(Z)-Hex-3-enyl] acetate195.9860.2865.60
Hexyl acetate191.51229.10178.00
Nonan-1-ol2.870.890.76
5-Ethyloxolan-2-one2.710.247.70
Table
Table 4. OPLS-DA model parameters for the three cultivation regions of BJ, SD, and HB. R2p(X) represents the variance related to class separation, R2(X) represents the total explained variance, R2(Y) represents the model developing ability, Q2(Y) represents the cross-validated predictive ability.
Table 4. OPLS-DA model parameters for the three cultivation regions of BJ, SD, and HB. R2p(X) represents the variance related to class separation, R2(X) represents the total explained variance, R2(Y) represents the model developing ability, Q2(Y) represents the cross-validated predictive ability.
Modeling ParametersR2p(X)R2(X)R2(Y)Q2(Y)ComponentsIntercept R2
BJ vs. SD0.7370.8610.9880.9831 predictive + 1 orthogonal0.078
BJ vs. HB0.6800.8460.9800.9691 predictive + 1 orthogonal0.058
SD vs. HB0.6410.7290.9940.9841 predictive + 1 orthogonal0.126
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Li, Q.; Li, B.; Zhang, R.; Liu, S.; Yang, S.; Li, Y.; Li, J. Flavoromics Approach in Critical Aroma Compounds Exploration of Peach: Correlation to Origin Based on OAV Combined with Chemometrics. Foods 2023, 12, 837. https://doi.org/10.3390/foods12040837

AMA Style

Li Q, Li B, Zhang R, Liu S, Yang S, Li Y, Li J. Flavoromics Approach in Critical Aroma Compounds Exploration of Peach: Correlation to Origin Based on OAV Combined with Chemometrics. Foods. 2023; 12(4):837. https://doi.org/10.3390/foods12040837

Chicago/Turabian Style

Li, Qianqian, Bei Li, Rong Zhang, Shuyan Liu, Shupeng Yang, Yi Li, and Jianxun Li. 2023. "Flavoromics Approach in Critical Aroma Compounds Exploration of Peach: Correlation to Origin Based on OAV Combined with Chemometrics" Foods 12, no. 4: 837. https://doi.org/10.3390/foods12040837

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop