Effect of high hydrostatic pressure on aroma components, amino acids, and fatty acids of Hami melon (Cucumis melo L. var. reticulatus naud.) juice

Abstract The changes and relationships between the volatile compounds and fatty acids, and between volatile compounds and free amino acids were analyzed after they were handled by 400 and 500 MPa (45°C/10 min) high hydrostatic pressure (HHP). The volatile components of 31, 30, and 32 were detected in the untreated, 400, and 500 MPa samples, respectively. Unlike the ketones and acids, the three contents, including ester (59.59%–71.34%), alcohol (5.95%–7.56%), and aldehyde (0.36%–1.25%), were greatly changed. While HHP treatment exerted a few effects on the contents of 12 kinds of fatty acids. With the increase in pressure, the contents of palmitic acid, linolenic acid, and α‐linolenic acid were remarkably reduced. The correlations between flavor compounds and amino acids, and between flavor compounds and fatty acids were studied by Pearson's correlation analysis and visualized with using the corrplot package in R software. The analysis showed that the amino acids were positively correlated with (E)‐6‐nonenal, (2E,6Z)‐nona‐2,6‐dienal and (Z)‐6‐nonen‐1‐ol, while they were negatively correlated with nonanal, (Z)‐3‐hexen‐1‐ol and ethyl caproate. Besides, the fatty acids were positively correlated with the esters of 2,3‐butanediol diacetate and 2‐methyl propyl acetate, while they were negatively correlated with (E)‐2‐octenal and (Z)‐6‐nonen‐1‐ol.

about the qualitative analysis of volatile compounds of the melons and identification of active fragrance components, but relative studies about the impacts of various nonthermal processing technologies on aroma were fewer.
Conversely, examination of the aroma compounds of HHPtreated strawberry coulis by GC-MS was to prove the changes of aromatic compounds after HHP treatment at 800 MPa/20°C/20 min (Lambert, Demazeau, Largeteau, & Bouvier, 1999). Bermejo-Prada, Vega, Pérez-Mateos, and Otero (2015) implemented a research to explored hyperbaric storage at room temperature how to affect aromatic volatiles of raw strawberry juice. In addition, mandarin juices were processed by HHP (0-500 MPa) and tested by electronic nose (Qiu, Wang, & Du, 2017). The data of E-nose were analyzed by local preserving projection that was compared with principal component analysis (PCA) as well as linear discriminant analysis to improve its diagnostic accuracy (Qiu et al., 2017).
In some literatures, esters, ketones, and alcohols were regarded as important factors that impacted the typical aroma of ripe fruits (Viljanen, Lille, Heiniö, & Buchert, 2011), and free amino acids and unsaturated fatty acids were viewed as prime precursors of aromatic synthesis in many fruits (Porretta, Birzi, Ghizzoni, & Vicini, 1995).
The literatures above indicated the volume added amino acids and the formation of isoamyl acetate, total esters, and 2-phenylethyl acetate were directly propotional, while this added amount and the synthesis of ethyl 3-hydroxybutyrate and diethyl succinate were inversely proportional (Esteve & Frígola, 2007). For the alcohols (except 2-phenylethanol and tyrosol), the formation of them had no indirect relation with the volume added amino acids (Dos-Santos et al., 2013). The addition of amino acids was favored for the formation of the total acids (Garde-Cerdán & Ancín-Azpilicueta, 2008). However, the report has been insufficient about the correlation between amino acids, fatty acids, and aroma compounds until now.
According to our preliminary experiments (Chen, Zheng, Dong, & Song, 2013;Pei, Hou, Wang, & Chen, 2018), Hami melon juice could maintain a better quality after processing by HHP at 400 MPa/45°C/10 min and 500 MPa/45°C/10 min. In these experiments, changes of fatty acids, amino acids, and aromatic compounds caused by HHP were determined, and the correlations between fatty acids and the aromatic compounds, and between the aromatic compounds and the free amino acids were explored.

| Chemicals
The four chemicals including heptanol (chromatographic grade), methanol (chromatographic purity), ethanol (chromatographic grade), and n-heptane (chromatographic purity) were purchased from Sigma-Aldrich. Besides, amino acid standard solution, including 16 individual amino acid standards (solid, purity ≥98%), single fatty acid methyl ester standards, and a standard mixture of fatty acid methyl esters, was purchased from China's J&K Chemical Ltd. And other reagents were bought from Beijing Chemical Reagent Co., Ltd.

| Preparations of melon samples and melon juice
Late maturing Jiashi muskmelons growing in Jiashi county of Xinjiang Uygur Autonomous Region, China, were harvested by the experienced melon farmers in late August 2017. Apart from the best eating quality, they were selected based on other several harvesting indexes, such as dark green skin, rich aroma, about 75% dry and firm netting skin, peduncle with a ring, and no decay or no quality deterioration. Moreover, harvested melons were immediately precooled by refrigerators under 4°C for 24 hr, followed by the conveyance to Shihezi University (Shihezi city, Xinjiang Uygur Autonomous Region, China), and they were stored in refrigerators before being analyzed.
In accordance with the method of Ma et al. (2007), the melons were washed under clear tap water; then calyx sections, pedicels, seeds, and peels were removed. Next, these melons were cut into pieces (about 1 × 2 × 3 cm 3 ), thoroughly mixed, and frozen in liquid nitrogen barrels (30 L) for about 5-7 min. Finally, the pieces of melons were sealed in vacuum and sterile laminated compound bags (20 × 15 cm 2 ) and then stored at −18°C. And the melons' pieces were thawed overnight in 4°C refrigerators. Bags containing melon pieces were selected randomly to be juiced through a Philips HR2838 juicer for analysis. The acquired melon juice was filtered by using an eightlayer sterile gauze, vacuumized, sealed in sterilized laminated compound bags, and stored at 4 ± 1°C.

| HHP treatment
High hydrostatic pressure processing equipment (Baotou Kefa High Pressure Technology Co., Ltd.) with maximum operating pressure and capacity of 600 MPa and 6 L, respectively, was respectively adopted to pressurize the Hami melon juice samples. The time to reach the designated pressure was <10 s, and the time of pressure release was <5 s. Before pressurization, deionized water was heated to the required temperature of 45°C, as the pressure transmitting medium. The samples were subject to the treatment under 400 and 500 MPa, respectively, at 45°C for 10 min.

| Extraction of volatiles
After the high-pressure processing, the volatiles were extracted from the samples by HS-SPME. Three types of samples were extracted for three times by using SPME (purchased from Supelco, Inc.) after preprocessing 50/30 μm carboxen/divinylbenzene fiber or polydimethylsiloxane (PDMS) fiber at 250°C for 30 min. To avoid browning and minimize the loss of volatiles, 8 ml melon juice was transferred quickly to a headspace bottle (15 ml) containing 2.1 g NaCl, and then, the juice was equilibrated through a hot plate/laboratory stirrer (PC-420, Corning, Inc., Life Science) at 40°C for 10 min.
The volatile components in the headspace were extracted at 40°C for 30 min by a stable flex PDMS fiber placed at 1 cm away from the liquid surface under 250 rpm/min magnetic stirring. Every one of the three analytical samples was experimented for three times, respectively.

| GC-MS analysis
After extraction, GC-MS analysis was carried out through a GC The SPME fiber head was inserted into the sample hole at 250°C with desorpting for 1 min, running time for 30 min. And fiber head was maintained 5 min to remove impurity in the sample hole, and this hole employed no crack (hole) mode. The flow rate of the carrier gas (99.9995% He) was 40 cm/s. The temperature program of the GC oven was kept at 50°C for 1 min, then was raised to 100°C at 5°C/ min, and increased to 250°C at the rate 10°C/min, and it was kept for 9 min. Helium at 1.2 ml/min acted as the carrier gas, and splitless injection mode was adopted. The quadrupole mass filter and the ion source were at 106 and 200°C, respectively. The mass spectrometer was used at 70 eV in the electron ionization mode, and the total ion current in the scanning range of 33-350 m/z was monitored to record the chromatograms. The NIST library was used for preliminary identification and combined with the retention time, the actual components, and the retention index to further determine most of the components.

| Analysis of free amino acids
An amino acid analyzer (L-8900, Hitachi) was adopted to analyze the amount of amino acids in the samples of Hami melon juice according to the Chinese Standard GB/T 5009. 124-2016 (2016).
After being hydrolyzed into free amino acids by HCl, the proteins of the juice samples were derivatized with ninhydrin solution, separated on an ion-exchange column, and recorded spectrophotometric absorbance detection at 440 or 570 nm. To identify the amino acids (g/100 g), the retention time of these amino acids was measured and then compared with that of the standard mixed solution.

| Fatty acid analysis
The third method of the Chinese Standard GB/T 5009. 168-2016 (2016) was used to determine the fatty acids. For total lipid extraction, 50 ml of sample was mixed with 10 ml of 95% ethanol. And the mixed liquid was transferred by a separatory funnel containing 50 ml ethyl ether/petroleum ether mixture and shaken for 5 min, and then it was left steady for 10 min. The ether layer extract was collected into a 250-ml flask and concentrated to dryness by rotary evaporation, so as to obtain the fat extract. Eight milliliter and 2% NaOH-methanol solution was added to the fat extract. The total fatty acids were converted into fatty acid methyl esters, and nitrogen stream was utilized to dry the organic phase. Next, 20 ml n-heptane was added and shaken for 2 min, and then, saturated aqueous NaCl was added. After leaving the mixture to stand for stratification, about 5 ml of the upper heptane extraction solution was pipetted into a 25 ml test tube. Approximately 3-5 g of anhydrous sodium sulfate was added, then shaken for 1 min, and stood for 5 min, before pipetting the upper phase of the solution into the sample bottle.
A Shimadzu GC-2010 equipped with a capillary column (100 m length × 0.2 μm film thickness × 0.25 mm i.d.) was adopted for fatty acid analysis. The temperatures of the detector and the sample injector were set at 280 and 270°C, respectively. The heating procedure was as the follows: started temperature at 100°C, kept for 13 min; raised to 180°C at 10°C/min, held for 6 min; increased to 200°C at 1°C/min, maintained for 20 min; and then raised to 230°C at 4°C/min, and kept for 10.5 min. The fatty acid results were expressed as g/100 g.

| Statistic analysis
Every experiments were conducted in triplicate, and the data analysis was analyzed, by using software SPSS (Statistical Program for Social Science) 17.0 purchased from Chicago, IL, USA. One-way analysis of variance (ANOVA) was used to evaluate the data. The correlation was analyzed by Pearson's correlation and visualized by using the corrplot package in R (Gu, Eils, & Schlesner, 2016;Wei & Wei, 2017). The PCA was performed with employing R package ade4 (Dray & Dufour, 2007).

| Change in volatile flavor compounds
The processed Hami melon juices under different high pressure processed were analyzed to decipher the resultant effects on the major aroma substances. The Hami melon juice odor was mostly consisted of various esters, alcohols, aldehydes, acids, and ketones (Chen et al., 2010). Table 1 listed the main odor compounds as well as their odor description. Ethyl acetate (intense, fruity) was an important aroma substance in Hami melon. The other substances were largely represented by various fruit odors, but ethyl 2-methylbutanoate, heptanal, (2E,6Z)-nona-2,6-dienal, and (Z)-3-hexen-1-ol showed cantaloupe-like flavors (Lambert et al., 1999). And the Table 2 presented the volatile flavor compounds in the control and processed Hami melon juice. Thirty-one volatile compounds were identified in fresh juice, whereas 30 and 32 volatile compounds were identified in the 400 and 500 MPa HHP-treated Hami melon juices, respectively. No significant differences could be seen in the ketones, and acids between HHP and untreated samples were observed.
Among the volatile components, esters were in majority (Table 2).

TA B L E 2 (Continued)
F I G U R E 1 Change of esters, alcohols, aldehydes, acids, and ketones grassy odors. The observed result showed that esters, acids, and ketones were not significantly changed by the different treatments.
When compared to the nontreated sample, the samples exposed to HHP treatments showed evidently lower and higher alcohol and aldehyde contents, respectively (p < .05). Earlier studies showed that under the effects of pressure, the content of aldehydes in juice was increased (Liu et al., 2014;Viljanen et al., 2011). From the perspective of flavor, it is well known that the most abundant volatiles may not be the most significant sensory elements. Some volatiles, even those at low concentrations, which were often viewed as crucial flavor compounds, play a decisive role in the aroma perceived (Bermejo-Prada et al., 2015). Although ultra-high-pressure treatment slightly increased the cucumber-like odor, HHP-treated melon juice still showed its unique Hami melon flavor. which both explained more than 80% (63.4% and 17.9%, respectively) of the data variance. Thus, the data interpretation rate was good. It could be seen from the distribution in the PCA loading plot that the 400 and 500 MPa were separated from the untreated group by a large distance, indicating a significant difference between the control group and the HHP groups in the content and type of the overall aroma components. Table 3 showed the changes in amino acids of Hami melon juice under the different treatment conditions. It was known that amino acids differed in their sensitivity to the processing conditions, depending on the processing methods and materials (Boye, Wijesinhabettoni, & Burlingame, 2012). The main free amino acids of the Hami melon juice in this study included arginine (Arg), lysine (Lys), phenylalanine (Phe), leucine (Leu), isoleucine (Iso), valine (Val), alanine (Ala), glycine (Gly), proline (Pro), glutamate (Glu), serine (Ser), threonine (Thr), and aspartic acid (Asp). For all these amino acids, six amino acids were essential and the rest seven were nonessential amino acids. According to Table 3, there were many kinds of amino acids in Hami melon juice, but the total content was low. The concentration of total amino acids increased greatly after HHP treatment, compared to the samples without treatment (p < .05). Lys was absent from the untreated sample, and Iso was found only after HHP treatment at 400 MPa. In comparison with the control, the content of each amino acid in the 400 MPa-treated juice increased, except Pro, which did not change.

| Change in amino acid contents
When the pressure rose from 400 to 500 MPa, Asp and Leu were evidently increased while Thr, Ser, and Arg decreased (p < .05), and Glu, Gly, Ala, Val, and Phe had no obvious change. Seven-day-old Brussel sprout seedlings were exposed to HHP treatment up at 800 MPa for 3 min, and Barba, Poojary, Wang, Olsen, and Orlien (2017)

| Change in fatty acids
The fatty acids of palmitic, palmitoleic, linoleic, α-linolenic, and oleic acids were the main fatty acid components in Hami melon juice, and they were key precursor substances of aroma esters (Chen et al., 2013). As shown in Table 4, after the fresh melon juice was treated at 400 MPa, palmitic, linoleic, and α-linolenic acids all showed a decreasing trend (p < .05) while oleic acid and palmitoleic acid have nonsignificant change. After HHP treatment at 500 MPa, however, the concentrations of all five fatty acids mentioned above decreased apparently compared with HHP at 400 MPa. During ultra-high-pressure treatment, it was reported that free fatty acids were mainly oxidized to form n-hexanal and hexenal (Porretta et al., 1995). Similar reports were seen in other experiments (Aganovic et al., 2014;Liu et al., 2014).

| Correlation analysis of aroma components with free amino acids
Apart from an attractive color, various amino acids could produce different tastes, like umami, bitter, and sweet (Tian et al., 2016).

F I G U R E 2 Principal component analysis (PCA) plot of aroma compounds of Hami melon juice at different treatments
Thus, if the content of amino acids was changed, the aroma substance and the quality of Hami melon juice might be affected. From the dark red area in the lower part of Figure 3, it could be seen that many aroma components (e.g., 2-methyl propyl acetate,  et al. (2013), the closest correlation was found in the volatiles and Phe, Leu, Val, Iso, or methionine. Zhang et al. (2018) noted that many amino acids (e.g., Ala) were closely related to aldehydes, pyrazine, and pyrrole.

| Correlation analysis of aroma components with fatty acids
Fatty acids were usually considered to be closely related to some odors (Braddock & Kesterson, 1973;Liu et al., 2014;Shi et al., 2018).    Viljanen et al. (2011) found that compared with standard tomatoes, the content of linolenic acids in the neutral lipid fraction of cherry tomatoes was higher, possibly leading to a higher concentration of (E)-2-hexenal and hexanal. Takahashi and Goto-Yamamoto (2011) found the content of medium-chain fatty acids was correlated with ethyl hexanoate concentrations in moromi (a fermented blend of barley, barley koji, yeast, and water), and hexanoic acid was correlated with fatty odor.
The results of the current investigation demonstrated esters and aldehydes were important aroma substances in Hami melon juice, and the levels of esters and aldehydes were closely associated with those of the amino acids and fatty acids, which provided a good basis for preserving the aroma of Hami melon drinks.

| CON CLUS IONS
The high-pressure treatment impacted on certain volatile odor compounds. The PCA diagram of aroma confirmed this conclusion. Thirtyone volatile compounds in total compounds were identified in fresh juice, whereas 30 and 32 volatile compounds were identified in treated Hami melon juices under 400 and 500 MPa HHP, respectively.
Compared with the control, the type and content of aldehydes, esters, and alcohols in the HHP Hami melon juice were changed, but ketones and acids did not get changed. However, only the aldehydes were clearly showed that amino acids were mainly related to 12 esters (hexyl acetate, heptyl acetate, methyl phenylacetate, ethyl caproate, 2-methyl butyl acetate, among others) and two alcohols, namely (Z)-3-hexen-1-ol and nonan-1-ol, and 2 aldehydes (heptanal, nonanal), whereas the fatty acids were mainly related to some esters and aldehydes, such as ethyl 2-methylpropanoate, ethyl butanoate, 2,3-butanediol diacetate, 2-butanol-2 methyl acetate, heptanal, and (E)-non-6-enal. The above analysis showed that the flavor of the 400 MPa sample was better than the 500 MPa one.

ACK N OWLED G M ENTS
This work was supported by the National Science Foundation of China (No. 31560461).

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

E TH I C A L A PPROVA L
This study does not involve neither human nor animal testing.