Bottle gourd [Lagenaria siceraria (Mol.) Standl.] (2n = 2x = 22), also known as long calabash, is an annual cucurbit crop, and its tender fruit is cooked as a vegetable because of its unique flavor (Wu et al., 2017). The bottle gourd contains essential nutrients, such as protein, vitamin, pectin, fibers, etc (Zhang et al., 2020). It is planted in Asia and Africa, as well as it is used in crafts, decorations, and medicinal (Xu et al., 2011). It can be used as the rootstock of other Cucurbitaceae, such as watermelon (Sari, 2003).

The fruits of different gourd varieties have different shapes, sizes and flavors (Wu et al., 2017; Zhang et al., 2020). These phenotypic differences are caused by gene transcription regulation (Zhang et al., 2020), and differences at gene level led to differences at metabolic level. There are few studies on the differences of metabolism of different gourd fruits.

Sugars, amino acids (Kader et al., 1978), organic acids (Robert and Franco, 2018), vitamins (Kafkas et al., 2006) components of plant organisms, but also the ratio of their contents to their contents affects the taste and flavor of edible parts of plants, such as fruits (Hobson and Bedford, 2015). The tastes that human beings can perceive are sweet, sour, bitter, salty and delicious. Generally speaking, the taste of the fruit is caused by its organic components. And traditionally, sugar content determines the sweetness of the fruit. However, with the continuous discovery of new organic compounds and technological progress, sugar is not the only substance with a sweet taste. For example, thaumatin is a kind of protein with a sweeter taste than sucrose (Healey et al., 2017). In addition, the content of organic acids and their derivatives in fruit pulp has important dietary considerations, which affect the taste of fruit and the applicability of processing into various fruit products in some cases. It can be seen that whether the fruit is sweet or sour has a great relationship with the proportion of sweet and sour substances in the fruit. The taste of tomato was the first concern and a balanced sugar-acid ratio was preferred in cherry tomatoes (Hobson and Bedford, 2015). Cucurbita pepo contains the toxic tetracyclic triterpenoid compound cucurbitacin, which is the cause of bitterness and toxicity (Verma and Jaiswal, 2015). The factors affecting fruit umami are free amino acids, nucleotides (monophosphates of inosinate or guanylate, inosine 5′-monophosphate, and guanosine-5′monophosphate), short peptides and organic acids (Kurihara, 2009; Wu et al., 2017). Previous studies have found that, free glutamate (Glu) was found to have the most significant association with umami taste, suggesting that free Glu was the main umami-conferring ingredient of bottle gourd (Wu et al., 2017).

Although the above-mentioned research shows that free glucose is an important factor affecting the fresh taste of bottle gourd fruit, there are many factors involved in the taste of the fruit, and the metabolomics analysis of bottle gourd with different tastes can reveal the metabolic reasons for this difference. However, there is no report on the metabolism of bottle gourd fruit taste. Therefore, in order to find out the difference of metabolism between two kinds of bottle gourd fruits with different tastes and provide a theoretical basis for quality breeding of bottle gourd, we collected and analyzed the metabonomic information of slightly bitter bottle gourd (BBG) and sweet bottle gourd (SBG).

Plant materials and chemicals

Five bottle gourd cultivars were cultivated in a randomized field plot according to standard agricultural practices in a field at the Zhuanghang comprehensive experimental station (E 121°28, N 30°57′) of the Shanghai Academy of Agriculture Sciences in 2019. Two bitter-tasted bottle gourd cv. BG27 and BG32, and three sweet-tasted bottle gourd cv. BG33, BG56, and BG58 were cultivated. They were sowed in March and planted in April. The photon flux density was range 650~850 W.m^-2. The temperature and relative humidity of the cultivation environment were about 10~25°C and 50~70%, respectively. The fruits were selected 14 days after simultaneous pollination, having a length of 20-35 cm and a diameter of 4.2-5.8 cm. Three marketable immature fruits were harvested for each cultivar and homogenized with equal weight to form a sample. All samples were stored at −80°C until analysis.

Chromatographic grade methanol and chloroform were purchased from Merck Chemicals (Germany). All of the chemicals, such as pyridine, methoxyamine hydrochloride, and N, O-bis (trimethylsilyl) trifluoroacetamide (BSTFA) containing 1% trimethylchlorosilane (TMCS), and the reference standards used in the study were purchased from Sigma-Aldrich (St. Louis, MO, USA). L-2-Chlorophenylalanine at 3.1 mg ml⁻¹ in water was prepared and used as internal standards (IS).

Sample preparation and extraction for metabolomic analysis

An aliquot of 1 g bottle gourd sample and 10 ml methanol-water (7:3 v/v) were mixed for metabolite extraction at 60 Hz for 180 s on a tissuelyzer. Further, the mixtures were ultrasonically extracted for 40 min and stored at 4°C for 24 h. After centrifugation at 12,000 × g for 10 min, 400 μl of the supernatant was spiked with 4 μl of the IS, L-2-chlorophenylalanine, and dried completely in a vacuum concentrator. An aliquot of 80 μl methoxyamine hydrochloride (15 mg/ml in pyridine) was added to the residue and incubated at 37°C for 90 min for methoxyamination. Subsequently, the sample was trimethylsilylated by adding 80 μl BSTFA (with 1% TMCS) and incubated at 70°C for 60 min. The derivatized samples were cooled to room temperature before being analyzed (Jaeger, 2008).

GC/MS conditions

For each cultivar, three biological replicates were independently analyzed. In total, 15 samples were randomly analyzed to reduce analysis bias. Under the adjusted initial pressure, all samples were analyzed in randomized order by a 7890B gas chromatography coupled with a 5977B mass spectrometer (Agilent Inc., CA, USA). A DB-5ms capillary column (30 m ×250 μm inner diameter, 0.25 μm film thickness; J&W Scientific, Folsom, CA, USA) was used to separate compounds. The injector port was heated to 280°C and injections (1 μl) were performed with a split ratio of 5:1. Helium (purity?99.999%) was used as the carrier gas at a constant flow of approximately 1 ml/min. The column temperature was held at 60°C for 1 min, then increased to 300°C at 5°C/min, and held for 11 minutes. The total run time was 60 min. The temperatures of the transfer line, ion source and quadruple were maintained at 280, 230 and 150°C, respectively. Electron impact ionization mass spectra were recorded with an ionization energy of 70 eV and EM voltage of 970 V. Mass spectra were scanned from 33 to 600 amu in total ion chromatogram mode after a solvent delay of 6.5 min (Weckwerth et al., 2004).

Data processing and statistical analysis

Metabolite identification was performed by the National Institute of Standards and Technology (NIST) mass spectral library (2017) in MSD ChemStation (version E.02.02.1431; Agilent Inc., CA, USA). The raw data acquired by Agilent GC/MS were imported into ChromaTOF (version 4.50.8.0; Leco Corporation, MI, USA) in NetCDF format. ChromaTOF could automatically compute baseline, find peaks above a signal to noise (S/N) of 100:1, deconvolute (identify overlapped peaks), integrate using specific masses, and align the same compound in different samples. The resulting three-dimensional dataset comprised sample information, peak retention time, and peak intensity. Some artificial peaks generated by noise, column bleed, and by-products in the silylation procedure were removed manually from the dataset. The resulting data were normalized to the area of the IS (L-2-chlorophenylalanine) for further statistical analysis. L-2-chlorophenylalanine was also utilized to assess process variability during sample preparation and data processing (Weckwerth et al., 2004).

For differential metabolites, we usually screen through VIP, such as screening out variables with VIP greater than 1.5, then calculating p-value, further screening out variables less than 0.05, and then observing these folds change was 2.0. Metabolites have been used for principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) by R (www.r-project.org/) to study metabolite cultivars-specific accumulation according to the described (Wang et al., 2018).

To further illustrate the biological significance associated with bitter/sweet taste, we used the Kyoto Encyclopedia of Genes and Genomes (KEGG) database to link differential metabolites to metabolic pathways in the SBGs compared with those in BBGs. Enrichment P-values were computed from a hypergeometric distribution. A p-value < 0.01 was selected to reduce the false discovery rate. The software used for drawing and data analysis was GraphPad Prism 7 (GraphPad Software Inc., La Jolla, CA, USA).

Differential metabolic pathways between BBG and SBG

To obtain detailed pathway information, all metabolites and the differential metabolites were mapped to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg/) (Figure 8A and B). In the mapping of KEGG, there were 12 different metabolites involved in metabolisms, 1 involved in genetic information processing and 1 involved in environmental information processing (Figure 8A). The differential metabolites were mainly involved in beta-alanine metabolism, indole alkaloid biosynthesis, galactose metabolism, ect (Figure 8B).

Figure 8. Enrichment of the differential metabolites to distinct KEGG pathways.

Source: Authors

Bottle gourd fruit varies in shape, size and flavor (Wu et al., 2017; Zhang et al., 2020). Previous studies have revealed the root causes of these differences at the genome and transcriptome levels (Konan et al., 2020). However, there is no research on the difference of metabonomics of gourd with different tastes. The metabolites of different gourd fruits are different. The metabolism of fruits with different tastes was different. GC-MS-based untargeted metabolomics of two groups of bottle gourd accessions yielded 16 differential metabolites. In previous reports, sugars were the  main factor affecting sweetness (Cameron, 2011). The sweetness of the saccharides differs. Furthermore, the sweetness of some saccharides varies with the solution concentration, while the sweetness of others does not (Cameron, 2011). Mannobiose, composed of D-mannose units, is a nondigestible disaccharide and rich in Cucurbitaceae crops. MNB can modulate intestinal and systemic immune responses in mice and prevent LPS-induced immune suppression as well as directly stimulate innate immune mechanisms in vitro as a TLR4 agonist (Kovacs-Nolan et al., 2013). Therefore, the difference in MNB contained in bottle gourds with different tastes will play a certain role in cultivating bottle gourds with high MNB content. D-talose is a variety of rare monosaccharides, different from the exception of isomers such as D-glucose and D-fructose, which exist in great abundance and showed considerable growth inhibition upon the nematode (Sakoguchi et al., 2016). Trehalose is a nonreducing disaccharide with two glucose units linked in an α, α-1,1 configuration, widespread in plant kingdoms and its phosphorylation forms play a vital role in the integration of plant metabolism and development (Richards et al., 2002). Evidence showed that exogenously-supplied trehalose can make winter wheat better cope with heat stress (Luo et al., 2018). D-fru is one of the most common and also the sweetest monosaccharide, but it is not the substance with the highest sweetness in nature. For example, protein thaumatin, first isolated from the katemfe fruit, Thaumatococcus daniellii, is 1600 times sweeter than sucrose on a weight basis (Wel and Loeve, 1972). Therefore, it is probably not only sugar that determines the strong sweet taste of the gourd.

4-Aminobutanoic acid was also known as the name of gamma-Aminobutyric acid (GABA), was a non-proteinogenic amino acid and was first isolated from potato tubers in 1949 (Ramesh et al., 2017). In plants, GABA acts as a signal substance to regulate plant growth, development and response to stress. In mammals, its action has been mainly described in the nervous system, where GABA receptors regulate brain function and development. L-aspartic acid is one of the 20 types of amino acids that make up a protein's structural unit (Akcay et al., 2012). Free amino acids have been found as the main umami ingredients in many vegetables, especially free glutamate and free aspartate are the two main umami ingredients (Zhang et al., 2013). Previous data has shown that free glutamic acid is the major factor affecting umami taste in bottle gourds (Wu et al., 2017). However, our data found that there were significant differences in aspartic acid metabolism, but there was no significant difference in glutamic acid content in different flavors of bottle gourd.

Besides sugars and amino acids, sour ingredients have an important influence on the flavor of fruits. Whether the taste of the fruit is sour or sweet depends on the ratio of sweet substances to sour substances. Shikimic acid is widely found in higher plants, with high content in the fruit of Illicium verum, which has an anti-tumor effect and is also an important synthetic raw material of Tamiflu, an anti-influenza drug (Coleman, 2005). Myo-Inositol belongs to B vitamins and is the precursor of various biological functional compounds containing inositol phosphate and lipid. Inositol and phosphatidylinositol play a key role in many metabolic pathways, and if affected, they will have adverse effects on human health (Benvenga and Antonelli, 2016). Ethylene is a plant hormone that can regulate fruit ripening. Ethylene glycol and Propane also exist in the flesh of bottle gourd, but their unique roles are not clear.

Multivariate statistics are used to further assess the differences in metabolic profiles among the two gourd groups. PCA is one of the most widely used chemometric tools to reveal the internal structure of several variables by a few principal components and extract and rationalize information from any multivariate description of a biological system (Park et al., 2013). The results showed that the samples from the two cultivars were grouped into two areas in the plot, indicating that different flavors had relatively distinct metabolic profiles. Especially, the dots of a gourd with sweet taste are more concentrated than the dots of a gourd with slightly bitter taste, which indicates that the metabolic regulation leading to the sweet taste of the gourd is more consistent, while the metabolic regulation related to bitter taste is more complex. The plot suggests that the differences in metabolic profiles detected in our study correlated with fruit flavors. In the PLS-DA models, BBGs are clearly separated from SBGs, indicating major distinctions in the metabolic profiles between the different fruit tastes of bottle guards. For BG32 and SBGs, there was some overlap in PCA (Figure 7A), but a clear difference was obtained in the PLS-DA score plots (Figure 7B), indicating that the BG32 with a slightly bitter taste could be grouped from the guard with a sweet taste.

The roles of carbohydrates, and their metabolism, in plants have a vast influence on both plant growth and development. During the storage of coffee beans in the pod, the change of flavor is related to the metabolism of sugars and free amino acids in coffee beans (Hinneh et al., 2018). Previous evidence has found that, free glutamate content was a key factor conferring umami taste in the bottle gourd (Wu et al., 2017).

However, in the study, we found that there were differences in metabolic substances and metabolic pathways between the two different tastes of bottle gourd, which lay the foundation to discover the metabolic architecture of the sweety/bitter taste in bottle gourd and facilitate the breeding of new varieties with better taste and quality.

The authors have not declared any conflict of interests.

This research was funded by the Shanghai Science and Technology Agriculture Promotion Project (2022-02-08-00-12-F01105). Thanks to the Instrumental analysis center, Shanghai Jiao Tong University for its instrument service.Thanks to the Institute of Vegetables and Flowers, Chinese Academy of Agricultural Sciences for providing the germplasm materials for the bottle gourd. Thanks to Guangzhou Gene Denovo Biotechnology Co., Ltd. for its Omicsmart cloud analysis platform.