Untargeted UHPLC-Q-Exactive-MS-based metabolomics reveals associations between pre-and post-cooked metabolites and the taste quality of geographical indication rice and regular rice

SHl Shi-jie,ZHANG Gao-yu,CAO Cou-gui,JlANG Yang#

1 College of Plant Science and Technology,Huazhong Agricultural University,Wuhan 430070,P.R.China

2 Shuangshui Shuanglü Institute,Huazhong Agricultural University,Wuhan 430070,P.R.China

Abstract Geographical indication (GI) rice refers to the rice of specific geographical origin,which tends to have a good taste quality and a high commodity price. Rice is favored for its soft texture and chewiness after cooking. However,GI rice is also plagued by rice fraud. Understanding the reasons for the excellent taste quality of GI rice and identifying its geographical origin can help maintain the stability of the rice market and promote the development of the rice industry. In this study,we determined the taste quality of rice. Untargeted metabolomics based on UHPLC-Q-Exactive-MS was used to identify metabolites in GI and regular rice before and after cooking. Our findings suggested that GI rice showed lower protein and amylose content,resulting in higher starch gelatinization properties and taste quality. This study identified 520 metabolites,among which 142 and 175 were significantly different between GI and regular rice,before and after cooking,respectively. The increased variety of metabolites after cooking was significantly negatively correlated with the taste quality of rice. GI rice was lower in amino acids and lipid metabolite content before and after cooking,which may be the reason for the excellent taste quality. Through linear discriminant analysis,we found that the differential metabolites of rice after cooking were more accurate in discriminating rice from different geographic origins,up to 100%. This work gained new insights into the metabolites of GI rice,which explains its excellent taste quality. The rice metabolites after cooking could be used for more accurate geographical identification of rice.

Keywords: rice,metabolomics,taste quality,before cooking,after cooking

1.lntroduction

Rice is the staple food for two-thirds of the world’s population and provides over two-thirds of their caloric intake (Virmani 1996). In 2017,the world’s rice production accounted for 758.8 million metric tons,of which 70%was processed into milled rice (FSA 2017). With the improvement in people’s living standards,consumers’demand for high-quality rice is increasing. The taste quality of rice refers to the taste characteristics of cooked rice. Generally speaking,geographical indication (GI) rice is trusted by consumers because it tends to have better sensory properties and taste qualities. Some examples of GI rice include Wuchang rice from China,Basmati rice from India,and Jasmine rice from Thailand (Giraud 2013;Kukusamude and Kongsri 2018). These world-famous rice are well-received by consumers for their aroma,softer texture,and chewiness after cooking (Wangcharoenet al.2016;Custodioet al.2019). However,GI rice is often plagued by food fraud due to its high value and limited quantity. Food fraud,a growing problem worldwide,is widely believed to be a deliberate illegal act for financial gains through counterfeiting,counterfeiting,or deliberate mislabeling (Bogadiet al.2016). For example,the demand for Jasmine rice in the international market has gradually increased over the past decade,so some unscrupulous traders have mixed Jasmine rice with other rice (Fridez 2016). Some traders pass off Jasmine rice by using inferior rice grown elsewhere (Vemireddyet al.2015). These frauds often undermine consumer trust in GI rice and exporting countries. This necessitates the development of new methods to identify GI rice.

In recent years,metabolomics has increasingly become a useful tool for identifying rice fraud(Uawisetwathana and Karoonuthaisiri 2019). Rice is rich in primary metabolites,such as amino acids and organic acids,as well as a large number of secondary metabolites,such as flavonoids and alkaloids;these specific metabolites provide the feasibility of identifying rice in different geographical locations (Chenet al.2013). 1H NMR-based metabolomic analysis showed that metabolomics could well identify 106 rice samples rice in 9 origins (Huoet al.2017). Gas chromatographymass spectrometry (GC-MS)-based metabolomics analysis of rice from China,India,and Vietnam showed that even 100% accuracy could be achieved in identifying the geographical origin of rice (Chet al.2021). In addition,metabolomics was also used to understand the differences between rice qualities. The better appearance quality of regenerated rice is often related to the change in the rice sugar metabolism pathway (Linet al.2022). Liquid chromatography-mass spectrometr(LC-MS)-based metabolomics showed that metabolites orchestrate the relationship between rice starch,protein,and lipids resulting in superior quality characteristics of high-quality rice cultivars under drought conditions(Chenet al.2020). Long-term storage increases the levels of metabolites related to lipid oxidation,polyols,and energy production in rice,which eventually leads to rice deterioration (Leeet al.2019). These findings demonstrate the superiority of metabolomics in identifying the geographic location and quality of rice. Among these metabolomics techniques,untargeted metabolomics analysis based on high-resolution mass spectrometry(e.g.,Q-Exactive (QE) series or triple time-of-fligh (TOF))allows unbiased,systematic,and accurate determination of various metabolites in rice,thus reflecting the metabolic level of the rice. It remains understudied how to identify the geographic origin of rice through metabolomics and clarify the relationship between metabolites and the good taste quality of GI rice. Rice is often eaten after cooking. During the cooking process,a large amount of water and heat destroy the starch and protein structure of rice and change the metabolite composition (Sittipod and Shi 2016). However,most current studies have focused on metabolomics in raw rice,while less research has been done on the metabolite differences in cooked rice. Several recent studies have shown that researchers are interested in the metabolite composition of rice after cooking. The volatile compounds from rice cooking can be used for the geographical marking of rice (Zhaoet al.2022a,b). The 194 metabolites of rice after cooking were associated with different rice subspecies (Heubergeret al.2010). The anthocyanins among the flavonoids are the most retained compounds after rice cooking(Rocchettiet al.2022). Cooking is a necessary step before the rice is eaten. Compared with regular rice,GI rice tends to have better sensory properties and taste quality. However,the metabolites of GI rice before and after cooking have not been studied,and the relationship between metabolites and rice taste quality is unknown.

In this study,we collected GI and regular rice of the same rice variety to measure the taste quality.Untargeted metabolomics based on UHPLC-Q-Exactive-MS was used to identify metabolites in GI and regular rice before and after cooking. Further goals were to explore the relationship between different metabolites before and after cooking and the taste quality of rice and to determine the accuracy of metabolomics before and after cooking in identifying GI rice.

2.Materials and methods

2.1.Materials

Rice samples were collected in Jiangjiadayan Village,Sunqiao Town,Jingshan City,Hubei Province,China(31.12°N,112.97°E),and Shuangquan Village,Mingfeng Town,Yuan’an City,Hubei Province,China (31.04°N,111.67°E) in November 2019. Jingshan rice (JS) refers to rice grown in Sunqiao Town,Jingshan City,which is a national geographical indication product in China.In ancient times,Jingshan rice was dedicated to the emperor. The rice has good taste quality and a history of over 400 years (Jingmen Municipal People’s Government 2021). Yuan’an rice (YA) collected in Yuan’an City is not GI rice. The rice variety in the 2 regions was Ezhong 5.About 500 g of rice were collected in 6 different locations on a field in Jingshan and Yuan’an when rice was mature;the rice samples were then sun-dried to a moisture content of around 14%. A huller (JLG-2118,Nongao,China) was used to dehull the rice,and a rice mill (Pearlest,Kett,Japan) was used to grind brown rice into milled rice. The picture of milled rice is shown in Appendix A.

2.2.Rice taste quality

The taste quality of rice after cooking was determined by the rice taste quality analyzer (Shiet al.2021),a tool that has been proven effective in relevant studies (Chenet al.2021). In short,30-g head of milled rice was weighed in a stainless steel jar,then the rice was washed within 20 s,and finally,the ratio of moisture and rice mass was ensured to be 1.4:1 (Bergman 2019). After soaking for 30 min,the rice was steamed in a rice cooker for 40 min and then simmered for 10 min. The cooked rice was flash cooled in the quench box for 20 min and left at room temperature for 1 h. A rice taste quality analyzer (STA1B,Satake,Japan) was used to measure the taste value of rice. A portion of the cooked rice was placed in liquid nitrogen and then freeze-dried for metabolomic analysis.

The gelatinization properties of rice starch were determined by a rapid viscosity analyzer. A quantity of 3 g milled rice flour and 25 mL of pure water were thoroughly mixed in an aluminum can and then placed in a rapid viscosity analyzer (RVA-TecMaster,Perten,Australia).The RVA program was first held at 50°C for 1 min,then heated to 95°C and held for 2.5 min,and finally cooled to 50°C. The gelatinization properties of rice starch included peak viscosity,hold viscosity,final viscosity,breakdown(peak viscosity-hold viscosity),setback (final viscositypeak viscosity),and gelatinization temperature.

2.3.Rice protein content and amylose content

A quantity of 4 mg of milled rice flour was weighed in parts per million balance,and an elemental analyzer (Elementar,Langenselbold,Germany) was used to determine the nitrogen content of rice,which was then converted to rice protein content using a conversion factor of 5.95 (Julianoet al.1973). The amylose content of rice was determined using the iodometric method (Manet al.2012). Briefly,0.01 g of rice was weighed in a ten-thousandth balance,100 μL of 95% ethanol was added to ensure that the sample was fully dissolved,and then 900 μL of 1 mol L-1NaOH solution was added. The mixture was heated in a boiling water bath for 10 min,and after cooling,9 mL of pure water was added. Then pipetted 0.5 mL into a new centrifuge tube and added 9.2 mL of pure water. Finally,100 μL of 1 mol L-1acetic acid and 200 μL of 0.2% iodine solution were added,and the mixture was mixed and allowed to stand for 10 min. The absorbance of rice was determined at 620 nm,and a standard curve of amylose and amylopectin was used to determine the amylose content of rice.

2.4.Metabolite analysis

First,0.2 g of raw milled (JS/YA) and cooked (JSC/YAC) rice were accurately weighed. Then,0.6 mL of 2-chlorophenylalanine (4 ppm) methanol was added.Samples and 100 mg glass beads were placed in a tissue grinder for 90 s at 60 Hz and sonicated for 15 min at room temperature. The samples were centrifuged at 12 000 r min-1for 10 min,and 300 μL of the supernatant was filtered through a 0.22-μm membrane. The filtrate was added to the detection bottle for UHPLC-Q-Exactive-MS analysis(De Voset al.2007). Then,20 μL of each sample was aspirated to form a QC sample (quality control),which is used to correct for deviations in analytical results and errors caused by the analytical instrument itself. The remaining samples were used for UHPLC-Q-Exactive-MS analysis.Conditions for LC-MS are detailed in Appendix B.

2.5.Data analysis

Proteowizard (v3.0.8789) was used to convert the obtained raw data into mzXML format,and the XCMS package(R v3.3.2) was used to perform peak identification,peak filtration,and peak alignment. Principal component analysis (PCA),partial least squares-discriminant analysis (PLS-DA),and orthogonal partial least squaresdiscriminant analysis (OPLS-DA) were performed using the ropls package in R. According to VIP value≥1 orP≤0.05,t-test was performed to screen out significantly different metabolites. The OmicStudio tools were used to perform a correlation network (https://www.omicstudio.cn/tool). In the pathway impact analysis,hypergeometric tests were used to determine the importance of different metabolite pathways (Xia and Wishart 2010). IBM SPSS Statistics 25 (version 20.0,Chicago,USA) was used for linear discriminant analysis (LDA). The 10 most abundant and significantly different metabolites of rice before and after cooking were used to identify the geographic origin of rice,and the metabolites were normalized to remove the effects of orders of magnitude (Huanget al.2021). The tolerance test was also performed in the LDA analysis,and 2 variables before and after rice cooking were excluded because of their lower contribution to the model (Karabagiaset al.2019). The Origin 2023(OriginLab,Northampton,USA) was used for plotting.

3.Results and discussion

3.1.Changes in protein content,amylose content,and taste quality of rice from different geographical locations

The rice protein content,amylose content,starch gelatinization characteristics,and rice taste quality are shown in Fig.1. Jingshan rice had lower amylose content and lower protein content (Fig.1-A and B). This may be due to different environments across various geographic locations. For example,adequate light and water resources and local customary cultivation practices can affect the chemical composition of rice (Bryantet al.2009). The gelatinization curve of starch is shown in Fig.1-D. When water and heat enter the starch,the viscosity will gradually increase,and after the viscosity reaches a peak,the starch breaks,and the viscosity gradually decreases (Schirmeret al.2015). Jingshan rice exhibited significantly higher peak viscosity,hold viscosity,breakdown,and significantly lower final viscosity,setback,and gelatinization temperature (Fig.1-E-J). Generally speaking,rice with lower protein content and lower amylose content has better taste quality (Luet al.2009).Proteins will adhere to the surface of starch,thereby inhibiting starch gelatinization (Yeet al.2018). Amylose inhibits the viscosity change of starch during starch gelatinization,resulting in lower peak viscosity (Varavinitet al.2003). This was confirmed in the gelatinization curve of starch (Fig.1-D). The RVA characteristic of starch represents the viscosity change of starch during gelatinization,which is an important indirect method to evaluate the taste quality of rice (Champagneet al.1999). The peak viscosity indicates the degree of water absorption of starch during heating;breakdown indicates the shear resistance of starch to heat;setback indicates the degree of recrystallization of starch during cooling;gelatinization temperature is the temperature at which starch begins to gelatinize (Heet al.2015). Higher peak viscosity and breakdown and lower gelatinization temperature are often associated with good taste quality of rice,and lower setback indicates that the rice is not easy to regenerate after cooling (Gravois and Webb 1997). The taste value of Jingshan rice was higher,which was consistent with the results of starch gelatinization,indicating that Jingshan rice had a higher taste quality(Fig.1-C). These findings suggested that GI rice had a higher taste quality in its origin,which may be related to the protein content,amylose content,and other metabolites of rice. However,the specific differential metabolites of rice from different geographic origins remains unclear.

3.2.Differential metabolite analysis of rice from different geographical locations before and after cooking

To understand the differences between metabolites of rice from different geographic origins,we performed a metabolomic analysis. The results of the hierarchical clustering analysis among metabolites are shown in Appendix C. A total of 520 metabolites were identified(Appendix D). We performed the principal component analysis (PCA) of metabolites,and the results of PCA showed a strong separation of rice from different geographic origins before and after cooking,indicating the reliability of our analysis results (Fig.2). To explore the differential metabolites between rice from different geographic origins,we performed differential metabolite screening through the PLS-DA model. There were a total of 142 significant differential metabolites in JS/YA,of which 63 were significantly up-regulated and 79 were significantly down-regulated. There were 175 significant differential metabolites in JSC/YAC,of which 17 were significantly up-regulated and 158 were significantly down-regulated (Appendix E). These findings showed that there were more types of differential metabolites after cooking,and the content of most metabolites in Yuan’an rice was higher than that in Jingshan rice.

Fig. 1 Protein content (A),amylose content (B),taste quality (C) and starch gelatinization properties (D-J) of rice from different geographic origins. JS,Jingshan rice;YA,Yuan’an rice. Error bars indicate the standard errors of the mean (n=6). Different letters in the same column indicate significant differences (P＜0.05).

Fig. 3 The top 10 metabolites of rice from different geographical origins before (A) and after (B) cooking. JS,Jingshan rice;YA,Yuan’an rice. Error bars indicate the standard errors of the mean (n=6). ＊ and ＊＊ denote significant differences at the 0.05 and 0.01 levels,respectively.

Fig. 4 Metabolic pathway analysis of JS/YA. A,the number of metabolic pathways. B,the pathway impact analysis. JS,Jingshan rice;YA,Yuan’an rice.

Fig. 5 Correlations between differential metabolites contents and rice taste quality before (A) and after (B) cooking in rice from different geographic origins. The orange line segment represented a positive correlation,the blue line segment represented a negative correlation,and the thickness of the line segment represented the magnitude of the correlation. The small orange circles on the periphery of the circles were metabolites,the red circles were rice taste quality,and the darker the color,the more related metabolites. The absolute value of Pearson’s correlation coefficient value above the threshold value was shown (r=0.5,P＜0.05).

In JS/YA,the most diverse metabolites were others,carboxylic acids and derivatives,fatty acyls,organooxygen compounds,and phenols. In JSC/YAC,the most diverse metabolites were carboxylic acids and derivatives,others,fatty acyls,organonitrogen compounds,and benzene and substituted derivatives(Appendix F). The variety of carboxylic acids and derivatives and fatty acyls increased after rice cooking.The most common metabolites in rice before and after cooking were carboxylic acids and derivatives and fatty acyls. Among these metabolites,the top 10 metabolites of Yuan’an rice before cooking were L-asparagine,L-glutamic acid,pyroglutamic acid,L-glutamine,triethylamine,sorbitol,deltaline,L-valine,imidazol-5-yl-pyruvate,and sphinganine (Fig.3). The content of 8 metabolites in Yuan’an rice was significantly higher than that in Jingshan rice,most of which were amino acids(Appendix E). After rice cooking,the top 10 metabolites of Yuan’an rice were 9,10-epoxyoctadecenoic acid,oleic acid,10E,12Z-octadecadienoic acid,L-glutamic acid,L-glutamine,L-asparagine,pipecolic acid,L-aspartic acid,L-olivosyl-oleandolide,and L-phenylalanine (Fig.3).Among these metabolites,the content of all metabolites in Yuan’an rice after cooking was significantly higher than that in Jingshan rice,including 6 kinds of amino acids and 3 kinds of fatty acids (Fig.3;Appendix E). These results showed that amino acids were the most abundant metabolites before and after rice cooking (Appendix F).Yuan’an rice had poor taste quality,and the content of amino acids in the rice before and after cooking was higher (Appendix E). Amino acids are the basic substances for protein synthesis. Yuan’an rice had higher protein content,which means higher amino acid content.Some previous studies have shown that the amino acid content of rice is closely related to the taste quality of rice (Gonget al.2020). The higher amino acid content in rice may have a negative correlation with the taste quality of rice (Ninget al.2010;Shiet al.2022). In some experiments of adding amino acids,it was found that the peak viscosity of starch decreased after the exogenous addition of amino acids,and the amino acids inhibited starch gelatinization (Wanet al.2017;Gonget al.2020).Therefore,higher taste quality of geographical indication rice was associated with lower amino acid content.

Apart from amino acids,fatty acids are the substances that differ the most after the rice is cooked. During the cooking process,lipids can form amylose-lipid complexes with amylose,which can affect the water solubility of starch and the gelatinization properties of starch (Banks and Greenwood 1972;Seneviratne and Biliaderis 1991).The addition of oleic acid to wheat starch increased the gelatinization temperature of the starch (Banks and Greenwood 1972). Removal of lipids in wheat and corn starch increased the peak viscosity of the starch and decreased the gelatinization temperature of the starch,indicating a higher degree of starch gelatinization (Melvin 1979). Therefore,higher fatty acid content may mean poorer taste quality. In our study,Yuan’an rice was found to have a significant increase in the content of the three fatty acids after cooking (Fig.3;Appendix E),which is consistent with the study of chestnuts,where the fatty acid content increased after cooking (Gon?alveset al.2010).When fatty acid was added exogenously,the hardness of rice was significantly increased,and the stickiness and taste quality of rice was significantly decreased (Biet al.2019). These findings suggested that GI rice had lower amino acid and fatty acid content before and after cooking,which may be beneficial for rice to maintain a higher taste quality.

3.3.Differential metabolic pathways of rice from different geographic origins before cooking

To understand the differences in metabolic pathways in rice before cooking. We performed the pathway impact analysis based on Kyoto Encyclopedia of Genes and Genomes (KEGG) to understand the metabolic pathways of differential metabolites. The number of metabolite pathways is shown in Fig.4-A,with carbohydrate metabolism and amino acid metabolism having the highest number,followed by lipid metabolism. Among these metabolite pathways (Fig.4;Appendix G),the significantly different metabolite pathways included alanine,aspartate,and glutamate metabolism (impact=0.75,P＜0.05),phenylalanine metabolism (impact=0.26,P＜0.05),and cysteine and methionine metabolism (impact=0.20,P＜0.05). Alanine,aspartate,and glutamate metabolism was the most important metabolic pathway. These findings demonstrate the importance of amino acid metabolism in rice of different geographic origins.

3.4.Correlation between differential metabolites of rice from different geographic origins before and after cooking and rice taste quality

In order to understand the relationship between rice differential metabolites and rice taste quality,we performed an association analysis on the taste quality indicators of differential metabolites in rice (Fig.5).In JS/YA,a total of 126 metabolites contents were significantly associated with breakdown,including 53 positive correlations and 73 negative correlations. A total of 91 metabolites contents were significantly associated with gelatinization temperature,including 56 positive correlations and 35 negative correlations. A total of 119 metabolites contents were significantly associated with peak viscosity,including 49 positive correlations and 70 negative correlations. A total of 113 metabolites contents were significantly associated with setback,including 72 positive correlations and 41 negative correlations. A total of 120 metabolites contents were significantly associated with taste value,including 44 positive correlations and 76 negative correlations. Among the 120 metabolites contents significantly correlated with the taste value,16 of the 19 organic acids contents showed a significant negative correlation. After rice cooking,we found that the relationship between metabolites contents and rice taste quality was more prominent,and more metabolites contents showed significant negative correlations (Fig.5;Appendix H),such as in the breakdown (14 positive correlations,137 negative correlations),gelatinization temperature (143 positive correlations,10 negative correlations),peak viscosity (15 positive correlations,129 negative correlations),setback (129 positive correlations,14 negative correlations),and taste value (14 positive correlations,146 negative correlations). Among the 160 metabolites contents that were significantly correlated with the taste value,most organic acids and lipids contents showed a significant negative correlation. These findings suggested that the differential metabolites of geographical indication rice and regular rice after cooking were more closely related to the taste quality of rice. The contents of most amino acids and lipid metabolites were negatively correlated with the taste quality of rice.

3.5.Linear discriminant analysis of rice from different origins before and after cooking

LDA is a discriminant technique that can be used to distinguish group differences between different samples(Huanget al.2021). To obtain the classification and identification results of Jingshan rice and Yuan’an rice,we performed the discriminant analysis with the 10 most abundant and significantly different metabolites.According to Wilks’ lambda,a discriminant function was constructed before and after rice cooking,which explained 100% of the variance (Appendix I). The metabolites that had the most impact on the origin determination of rice before and after cooking were L-asparagine and 9,10-epoxyoctadecenoic acid,respectively (Appendix I).Two Fisher discriminant functions were used to distinguish rice in different geographical locations,as follows:

Different rice origins were well identified,and the LDA model achieved a classification rate of 100% before and after cooking (Table 1). The predictive ability was assessed by leave-one-out cross-validation. The LDA model had an accuracy rate of 91.7% in discriminating the origin of rice before cooking,and the accurate discrimination rate for Jingshan rice was 83.3%. One sample of Jingshan rice was identified as Yuan’an rice. However,the discrimination accuracy of both rice origins reached 100% after rice cooking. This result indicated that differential metabolites are more effective in identifying GI rice after cooking than before cooking.

Table 1 Classification and accuracy of rice in different origins according to the linear discriminant analysis (LDA) model1)

4.Conclusion

The taste quality and metabolites of GI and regular rice were often different,but the relationship between the specific metabolite composition difference and the taste quality remains to be studied. In this study,we compared the differences in metabolites and the taste quality of rice before and after cooking in 2 geographic locations,one of which was GI rice. Our findings suggested that GI rice had lower levels of amino acid metabolism,resulting in lower protein content. The lower protein content in rice promoted the gelatinization of starch,resulting in higher peak viscosity and breakdown,ultimately increasing the taste quality of rice. This study identified 520 metabolites,among which 142 and 175 metabolites were significantly different between GI and regular rice before and after cooking,respectively. After cooking,GI and regular rice exhibited more significant differences in the metabolite types. The content of more metabolite components was significantly negatively correlated with the taste quality of rice. Amino acids and lipids are important metabolites affecting rice’s taste quality. The results of the linear discriminant analysis showed that the metabolites of rice after cooking could be used to accurately determine the geographic origin of rice,reaching a 100% accurate discrimination rate. Therefore,inhibiting rice’s amino acid metabolism or lipid metabolism may be an effective way to improve taste quality in future breeding work. It is of practical significance to identify the adulteration of GI rice by using the metabolites of rice after cooking.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (32272204),the Key Project of Hongshan Laboratory of Hubei Province,China(2021hszd002),the Guangdong Provincial Key R&D Program of China (2021B0202030002),the Strategic Consulting Research Project of Chinese Academy of Engineering (2021-XZ-30),and the Fundamental Research Funds for the Central Universities,China (2662019QD049).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available at https://doi.org/10.1016/j.jia.2023.06.003