Variable Secondary Metabolite Profiles Across Cultivars of Curcuma longa L. and C. aromatica Salisb.

Background: Curcuma spp. (Zingiberaceae) are used as a spice and coloring agent. Their rhizomes and essential oils are known for medicinal properties, besides their use in the flavoring and cosmetic industry. Most of these biological activities were attributed to volatile and nonvolatile secondary metabolites present in the rhizomes of Curcuma spp. The metabolite variations among the species and even cultivars need to be established for optimized use of Curcuma spp. Objectives: We compared the phytochemical profiles of rhizomes and their essential oils to establish the variability among seven cultivars: five of Curcuma longa L. (Alleppey Supreme, Duggirala Red, Prathibha, Salem, Suguna) and two of C. aromatica Salisb. (Kasturi Araku, Kasturi Avidi). The GC-MS and LC-MS-based analyses were employed to profile secondary metabolites of these selected cultivars. Methods: Rhizomes of Curcuma spp. were subjected to hydro-distillation to collect essential oil and analyzed by GC-MS. The methanol extracts of fresh rhizomes were subjected to LC-MS analyses. The compounds were identified by using the relevant MS library databases as many compounds as possible. Results: The essential oil content of the cultivars was in the range of 0.74–1.62%. Several compounds were detected from the essential oils and rhizome extracts by GC-MS and LC-MS, respectively. Of these, 28 compounds (13 from GCMS and 15 from LCMS) were common in all seven cultivars, e.g., α-thujene, and diarylheptanoids like curcumin. Furthermore, a total of 39 new compounds were identified from C. longa L. and/or C. aromatica Salisb., most of them being cultivar-specific. Of these compounds, 35 were detected by GC-MS analyses of essential oils, 1,2-cyclohexanediol, 1-methyl-4-(1-methylethyl)-, and santolina alcohol, to name a few. The other four compounds were detected by LC-MS of the methanolic extracts of the rhizomes, e.g., kaempferol-3,7-O-dimethyl ether and 5,7,8-trihydroxy-2′,5′-dimethoxy-3′,4′-methylene dioxyisoflavanone. Conclusions: We identified and recorded the variability in the metabolite profiles of essential oils and whole rhizome extracts from the seven cultivars of Curcuma longa L. and C. aromatica Salisb. As many as 39 new metabolites were detected in these seven Indian cultivars of Curcuma spp. Many of these compounds have health benefits.


INTRODUCTION
Turmeric (Curcuma longa L.) is a perennial rhizomatous herb that belongs to the family Zingiberaceae (Prasath et al., 2018). It has been used traditionally in India for its medicinal value and as a spice (Srinivasan et al., 2004;Aggarwal et al., 2007;Esatbeyoglu et al., 2012). In Ayurvedic medicine, turmeric is used internally (as a stomachic, tonic, and blood purifier) or externally (prevention and treatment of skin diseases) (Gounder and Lingamallu, 2012). Turmeric was scientifically validated for several pharmacological benefits, including antioxidant, antiinflammatory, and chemoprotective properties (Miquel et al., 2002;Krup et al., 2013;Kanase and Khan, 2018;Umar et al., 2020). The rhizomes of turmeric are enriched with several bioactive metabolites, though the attention was mostly on curcuminoids. Besides curcumin (a curcuminoid), the essential oil of C. longa L. showed antimicrobial activity and ability to suppress aflatoxins production (Ferreira et al., 2013).
Out of 110 species of genus Curcuma, only ∼20 species were used so far for phytochemical studies (Nahar and Sarker, 2007). Curcuma longa L. is popularly known as turmeric, while C. aromatica Salisb. and C. caesia Roxb. are known as wild turmeric and black turmeric, respectively. C. longa L. and a few other species, including C. aromatica Salisb., produce curcumin, a yellow colored curcuminoid. So far, at least 235 compounds, primarily phenolics, terpenoids, and alkaloids, were identified from Curcuma spp. . About 70 varieties of C. longa L. are cultivated in India (Sasikumar, 2005;Parthasarathy and Chempakam, 2008), but very few are chemically profiled.
The essential oil of Curcuma spp. is used in traditional medicine for many ailments (Dosoky and Setzer, 2018). The volatile component of C. longa's rhizome is responsible for its aromatic flavor and odor (Gounder and Lingamallu, 2012). Its essential oil is considered safe for human use (Tisserand and Young, 2013). The oils of C. longa L. and C. aromatica Salisb. have applications in the food and pharmaceutical industries due to their antioxidant, antibacterial, and anti-inflammatory properties (Dosoky and Setzer, 2018). The essential oil also improved the bioavailability of curcumin, thereby its bioactivity (Shishu and Maheshwari, 2010). Preliminarily clinical trials indicated that the essential oil from C. longa L. and C. aromatica Salisb. was helpful against cancer, asthma, and other ailments (Cheng et al., 1999;Joshi et al., 2003;Li Y. et al., 2009). Thus, there is a need to identify high-yielding cultivars containing curcuminoids and essential oil.
Since the pharmacological properties of Curcuma spp. are dependent on their chemical profiles, studies on the chemical constituents of turmeric/wild turmeric and their essential oils gained significance. Thin-layer chromatography (TLC) is one of the methods employed to quantify curcumin (Setyaningsih et al., 2016) and other curcuminoids (Phattanawasin et al., 2009) in Curcuma longa L. A few different techniques used were HPTLC (Pathania et al., 2006;Paramasivam et al., 2009), nuclear magnetic resonance (NMR) spectroscopy (Li W. et al., 2009), and the HPLC method (Kulyal et al., 2016).
Our present study on metabolite profiles would pave the way for metabolomics by providing the identity of several metabolites. Metabolomics is a practical approach for the comprehensive profiling and comparison of metabolites in plant systems (De Vos et al., 2007). It is crucial for quality evaluation and scientific validation of medicinal plants and their products (Mukherjee et al., 2016). Mainly information on secondary metabolites of medicinal plants/spices is of great importance in health, food, and nutrition sectors, due to the antioxidant nature, color, or flavor of these secondary compounds (Beekwilder et al., 2005;Dixon et al., 2006;Hall, 2006). The quality of turmeric and other spices depends on factors, such as cultivation, collection, storage, milling, and processing, apart from genetics and adulteration issues. Therefore, metabolomics provides a practical approach for quality control (Mukherjee et al., 2016;Tetali et al., 2021).
Over the past decade, several methods suitable for largescale analysis of metabolites in plant extracts were developed (Dixon et al., 2006;Hall, 2006). However, to date, no single analytical method can successfully detect the entire metabolome of higher plants, especially of medicinal and aromatic plants, as they are highly rich in chemically diverse metabolites (Tetali et al., 2021). The GC-MS and LC-MS techniques mutually complement each other in unraveling secondary metabolomes comprising a wide range of volatile and nonvolatile compounds. These compounds belonged to terpenes, phenolic acids, phenylpropanoids, saponins, alkaloids, polyamines, and their derivatives (Huhman and Sumner, 2002;Moco et al., 2006).
Essential oils from different Curcuma species, including C. longa L. and C. aromatica Salisb, were studied for their chemical constituents (Choudhury et al., 1996;Angel et al., 2014;Nampoothiri et al., 2015;Dosoky and Setzer, 2018) to establish their variability. Variation in the volatile compositions of Curcuma spp. such as C. longa L. and C. zedoaria, was done using GC-MS (Dosoky et al., 2019). A combination of GC-MS and LC-MS techniques was used for metabolite analysis of C. domestica L. (C. longa L.) (Herebian et al., 2009). In the present study, the volatile (essential oil) and nonvolatile (total extract) components of the fresh rhizome of the seven cultivars of Curcuma spp. were analyzed by the GC-MS and LC-MS techniques. The present study is the first report revealing such detailed metabolite profiles of the selected cultivars to the best of our knowledge. These cultivars, except Alleppey Supreme, are typically cultivated in Telangana and Andhra Pradesh, and these states are among the largest producers of turmeric in India (Parthasarathy and Chempakam, 2008).
Most of the studies worldwide on Curcuma spp., for their curative properties, were with C. longa L., followed by C. aromatica Salisb, C. aeruginosa Roxb. (Simoh and Zainal, 2015), and C. kwangsiensis S. K. Lee & C. F. Liang (Zeng et al., 2009). Several cultivars exist within these species, which vary in their chemical profiles. The present article is the first attempt to characterize both volatile (essential oil) and nonvolatile (crude extract) components of fresh rhizomes of seven cultivars of Curcuma spp. by the GC-MS and LC-MS techniques. Our results using GC-MS and LC-MS analyses revealed high variability in their metabolite profiles of seven cultivars of genus Curcuma. We emphasize that such an approach could be exploited to distinguish cultivars for a specific application based on their metabolite profile.
Fresh rhizomes of four cultivars of Curcuma longa L. Isolation of Essential Oil by Hydrodistillation for GC-MS Analysis 50 g each of fresh turmeric rhizome of five cultivars of C. longa L. cvs. Alleppey Supreme, Duggirala Red, Prathibha, Salem, Suguna, and two cultivars of C. aromatica Salisb. cvs. Kasturi Araku and Kasturi Avidi were taken out from a −80°C freezer, made into pieces, and ground in a pestle with a mortar to a fine powder under liquid nitrogen. The powder was subjected to hydrodistillation in a Clevenger-type apparatus for 7 h. The essential oil obtained after distillation was dried over anhydrous sodium sulfate and kept at −80°C until GC-MS analysis.

GC-MS Running Conditions and Metabolite Identification
The chemical composition of the Curcuma spp. essential oil was analyzed by the GC-MS technique using Agilent 7890 A gas chromatograph coupled with a Leco Pegasus HT TOF mass spectrometer equipped with a 29.8 m × 320 µm HP-5MS 5% phenyl methyl siloxane capillary column with 0.25 µm film thickness. The oven temperature was programmed at 65°C for 2 min and then increased from 65 to 90°C at 5°C/min (held for 3 min). Then the temperature was increased from 90 to 103°C (held for 3 min) and from 103 to 150°C (held for 15 min) at 20°C/ min and 8°C/min, respectively. The temperature was raised finally from 150 to 280°C at 20°C/min. The injector, interphase, and ion source were maintained at 250°C, 280°C, and 250°C, respectively. The detector voltage was 1500 V. A solvent delay of 2 min was selected. One microliter (diluted with n-hexane; 1:10) of essential oil sample was injected into the GC-MS system using split mode (50: 1). Helium was used as a carrier gas at a flow rate of 1 ml/min. GC-MS data were measured at 70 eV; mass scan 40-1000 amu.

Preparation of Rhizome Extracts for LC-MS Analysis
Samples for LC-MS analysis were prepared by grinding the fresh rhizome to a fine powder in a mortar and pestle under liquid nitrogen. 1 g of the rhizome powder was suspended in 2 ml of MeOH (LC-MS grade). The samples were sonicated for 30 min and centrifuged for 25 min at 1500 rpm, and the supernatants were separated by filtering through a 0.45-µm Nylon filter disk. These extracts were freshly prepared for the analysis. A 200 µl aliquot of the extract was diluted quantitatively with internal standard (IS) 200 µl 4-fluoro-4′hydroxy benzophenone solution. It was prepared freshly for each analysis by dissolving in methanol for a final concentration of 0.58 mg/ml. The samples were subjected to LC-MS analysis for the complete metabolite profile. The data obtained with the samples collected in 2012 were presented in this article.

LC-MS/MS Conditions and Metabolite Identification
LC-MS analyses of the crude extract of fresh rhizome of Curcuma spp. were performed according to Jiang et al. (2006) using Agilent 6520 Accurate Q-TOF (Agilent Santa Clara, CA), and the column used was Zorbax Eclipse XDB-C 18, 4.6 × 50 mm, 1.8 µ; Mobile phase: A) buffer (5 mM ammonium formate, 0.1% formic acid, in deionized and distilled H 2 O) and B) acetonitrile; gradient (in buffer A): 0-2 min, 5% B; 2-57 min, 5-100% B; 57-60 min, 100% B; 60-65 min, 100-5% B; flow rate: 0.25 ml/min; temperature, 40 o C; injection volume 5 µl. For the MS detection, Agilent MSD-Trap-SL was equipped with electrospray ionization (ESI) interface as the ion source. The acquisition parameters for the negative mode were: drying N 2 temperature, 350 o C, 8 l/min; nebulizer pressure 40 psi; HV capillary 4000 V; skimmer 65.0 V; mass range measured: 110-1700 m/z; Spray voltage: 4 kV; scan rate 1.4. We analyzed the results in both the positive and the negative ion mode acquired by Agilent TOF/Q-TOF mass spectrometry and full MS scan, in the form of total ion current (TIC) chromatogram, and the metabolites were identified based on their MS/MS spectra and fragmentation rules reported previously (Jiang et al., 2006).

Essential Oil Content
The oil was obtained by hydro-distillation, in a Clevenger-type apparatus, of the fresh rhizomes of five cultivars (Alleppey Supreme, Duggirala Red, Prathibha, Salem, and Suguna) of Curcuma longa L. and two cultivars (Kasturi Araku and Kasturi Avidi) of C. aromatica Salisb. The yield of essential oil from the seven cultivars was in the range of 0.74-1.62% on a fresh weight basis, with the highest yield of 1.62% in cv. Kasturi Avidi (C. aromatica Salisb.) followed by cv. Alleppey Supreme (C. longa L.) with an amount of 1.42% and the lowest yield of 0.74% in cv. Duggirala Red (C. longa L.). The essential oil yields from the other five rhizomes were in between these values ( Table 1). The oil yields of C. longa L. varieties were higher than those of C. aromatica Salisb.

GC-MS Analysis of Essential Oil
Essential oils of seven cultivars of Curcuma spp. were subjected to GC-MS analysis, and the results from one of such studies for each cultivar are presented in this article. The representative TIC chromatograms of these cultivars are shown in Figure 1. Several compounds were detected in each cultivar's essential oil (Table 1). Only a few of the identified compounds were confirmed based on their match with the compound profiles found in the NIST databases and Leco-Fiehn Rtx5 library. Up to 44 compounds were identified from the five cvs. of C. longa L. and 31 compounds from two cvs. of C. aromatica Salisb. (Table 1).
Frontiers in Pharmacology | www.frontiersin.org June 2021 | Volume 12 | Article 659546 aromatica Salisb., cv. Kasturi Araku had the least ( Table 2). Of these, 23 cultivar-specific compounds were reported for the first time from the genus Curcuma. The chemical structures of these 23 compounds ( Table 2, Sl. Nos. 1-23) are presented in Figure 2 (panels 1-23). In addition to 41 cultivar-specific compounds, 26 compounds were present in more than one cultivar (Table 3). Among these, 12 were detected first time in the genus Curcuma ( Table 3, Sl. Nos. 1 to 12; Figure 2, panels: 24-35). The remaining 14 were already known in C. longa L. A total of 13 compounds were common in all seven cultivars of C. longa L. and C. aromatica Salisb. (Table 4) and the representative structures of two of these compounds are given in Figure 3 (panel numbers 14-15, corresponding to serial numbers 10 and 13 respectively of Table 4). Most of these compounds belong to mono, di, and sesquiterpene. A summary of all 80 compounds identified by GC-MS in the essential oils of the seven cultivars is presented in Supplementary Table S1.

LC-MS Analysis of Methanol Extracts
Methanolic extracts of rhizomes from the seven cultivars of Curcuma spp. were subjected to LC-MS analysis. The results from one of such analyses for each cultivar are presented in this article. The use of "positive" and "negative" modes of LC-MS was quite helpful. TIC chromatograms of all seven cultivars of Curcuma longa L. and C. aromatica Salisb. were shown in Supplementary Figure S1A for negative mode and Supplementary Figure S1B for positive mode. A typical LC-MS analysis of methanolic extracts from rhizomes of C. longa L. cv. Alleppey Supreme revealed the presence of up to 86 compounds. Out of these, 43 were identified, and the remaining 43 compounds remained unknown. The (-) ESI-LC-MS detected 30 known compounds, and the (+) ESI-LC-MS detected 23 known compounds with an overlap of 10 compounds, detected by both negative and positive ion modes. A similar assessment of data was done with all seven cultivars of Curcuma spp. (Table 5). Altogether 62 compounds were identified, as presented in Supplementary Table S2. These compounds were grouped into three categories: cultivar-specific, detected in more than one cultivar, and common. There were 23 cultivar-specific compounds present in any one cultivar of C. longa L. or C. aromatica Salisb. (Table 6). 24 compounds were present in more than one cultivar of C. longa L. and/or C. aromatica Salisb. ( Table 7). The remaining 15 were common in all seven cultivars ( Table 8). Of these 15 common compounds found in the LC-MS/MS chromatograms, only one was a "Bisabolane" sesquiterpene (Parthasarathy et al., 2009) and all other 14 were diarylheptanoids. These were identified based on the MS/MS spectra reported by Jiang et al. (2006), including curcumin (CU), demethoxycurcumin (DMC), and bisdemethoxycurcumin (BDMC). Among the other diarylheptanoids, 1-(4-hydroxyphenyl)-7-(4-hydroxy-3-methoxyphenyl)-1,4,6-heptatrien-3-one; 1,5-bis(4-hydroxy-3-methoxyphenyl)-1,4-pentadien-3-one, etc., were common in all the cultivars of C. longa L. and C. aromatica Salisb. ( Table 8). The structure of the five common compounds was given in Figure 4 (panels 13, 14-15, 16, and 17 corresponding to serial numbers 10, 5-6, 15, and 1, respectively, of Table 8). In addition to diarylheptanoids, several other classes (phenolic acids, flavonoids, ketonic sesquiterpenes, and fatty acid derivatives) were also detected in the turmeric rhizomes.

Compounds Reported First Time From the Genus Curcuma Using GC-MS and LC-MS Analysis
A total of 39 compounds were detected (Figure 2) for the first time from the genus Curcuma. Out of these, 35 and 4 compounds were identified respectively in the essential oils and whole rhizome extracts of C. longa L. and C. aromatica Salisb. by the GC-MS and LC-MS techniques. Details of the compounds, including the class of compound, molecular weight, are given in Tables 2, 3, 5, and 6; structures of all these compounds are shown in Figure 2 (panels: 1-39). The MS and MS/MS spectra of these compounds are presented in Supplementary Figure S2 (panels: 1-39). These compounds were reported earlier from plants belonging to any genus other than Curcuma, and this is the first report from genus  6 | Cultivar-specific compounds identified by LC-MS in the rhizome extracts from one of the seven cultivars of Curcuma longa L. and C. aromatica Salisb. The structure of the compound with the serial number "1" is given in Figure 2 (panel number: 36) and compounds with Sl. Nos. 2, 3-9 are given in Figure 4 with the corresponding panel nos. 3, 5-11, respectively. Curcuma. Out of the total of 62 compounds detected by LC-MS analyses of rhizome extracts, four compounds were reported for the first time from the Curcuma genus. One was cultivar-specific ( Table 6; Sl. Nos. 1; Figure 2, panels: 36), and three were present in more than one cultivar (Table 7; Sl. Nos. 1-3). The structures of these first-time reported compounds are given in Figure 2 (panels: 37-39), and their corresponding mass fragmentation spectra are shown in Supplementary Figure S2 (

DISCUSSION
In one of our previous studies, we reported that the HPLC method could be a valuable tool to differentiate the cultivars of Curcuma spp. based on their curcuminoids content ratios (Kulyal et al., 2016). Curcuminoids play a significant role in food, cosmetics, and medicinal compounds. But there are several other secondary metabolites such as terpenoids (e.g., mono-, sesqui-, di-, tri-, so on), alkenes, aromatic compounds, flavonoids, coumarins, etc. that are responsible for various biological activities. All these secondary metabolites are present in either the volatile essential oil or the nonvolatile fraction of the Curcuma spp. Employing untargeted metabolomics would be the ideal way to identify as many metabolites as possible. Therefore, in the present study, we analyzed these secondary compounds using GC-MS and LC-MS/MS.

Versatility of GC-MS and LC-MS Techniques to Identify a Large Number of Metabolites
GC-MS analysis is an appropriate technique for analyzing volatile compounds, whereas LC-MS is for detecting polar compounds, and thus, these two techniques are mutually complementary to each other. In the present study, several of the volatile compounds present in the cultivars of C. longa L. and C. aromatica Salisb. belonging to mono-and sesquiterpenoids were detected by GC-MS (Tables 2-4). On the other hand, LC-MS analysis detected phenolic (Tables 6-8) compounds, including several diarylheptanoids in the methanolic extracts of both C. longa L. and C. aromatica Salisb. (Figure 4). Electrospray ionization (ESI), coupled with LC/MS/MS, turned out to be a powerful tool in metabolite profiling and metabolomics research. Studies on chemical derivatization and quantification of several metabolites in turmeric powders and fresh rhizome extracts by LC-MS or LC-MS/MS were made. But the rapid screening within the cultivars of C. longa L. of fresh turmeric rhizome has not yet been reported. To the best of our knowledge, we were able to record the presence of several metabolites, which were not reported so far in the C. longa L. and C. aromatica Salisb. (Tables 2, 3, 6, 7, and Figure 2), using the available literature search, Metlin library, mass bank, and NIST library. TABLE 6 | (Continued) Cultivar-specific compounds identified by LC-MS in the rhizome extracts from one of the seven cultivars of Curcuma longa L. and C. aromatica Salisb. The structure of the compound with the serial number "1" is given in Figure 2 (panel number: 36) and compounds with Sl. Nos. 2, 3-9 are given in Figure 4 with the corresponding panel nos. 3, 5-11, respectively.  7 | Compounds identified by LC-MS in rhizome extracts of more than one cultivar of Curcuma longa L. and C. aromatica Salisb. The compounds from serial numbers 1-3 are reported first time from the genus Curcuma, the structures of these compounds along with few others are given in Figure 2 (panels: 37-39; panels 1-2, 4, 12 corresponding to serial numbers 4-5, 6, 7).

Cultivar Variability Based on Secondary Metabolites
Based on the presence or absence of metabolites identified by GC-MS and LC-MS analyses, there was a need to authenticate cultivar variability. Thus, the metabolite library can be constructed based on the cultivar-specific and compounds found in more than one cultivar. There are very few reports on cultivar-specific secondary metabolite variation. Out of a total of 142 compounds identified by both GC-MS and LC-MS, only 28 compounds (13 from GCMS and 15 from LCMS) were common (Tables 4, 8) present in all the cultivars of C. longa L. and C. aromatica Salisb. Ten of 13 common compounds (GCMS) were reported earlier from C. longa L. rhizome. Two compounds, namely sabinene and α-caryophyllene, were reported from the leaves of C. longa L. The remaining one compound, i.e., 2-heptadecanone, was detected for the first time from these two Curcuma species. This compound was earlier reported in the essential oil of Curcuma angustifolia Roxb. rhizome (Srivastava et al., 2006). As per our analyses, 64 compounds (Tables 2, 6) out of 142 compounds were cultivar-specific. Of these 64 compounds, 41 were identified in essential oils by GC-MS (e.g., carvacrol, endo-borneol) and 23 (e.g., tumerone, methyl-7-methoxycoumarin,4-) in fresh rhizome extracts (LC-MS) of any one of the cultivars of C. longa L. or C. aromatica Salisb. In addition, 50 compounds (Tables 3, 7) were identified to be present in some of the cultivars, present in more than one cultivar but not common to all the cultivars of C. longa L. and C. aromatica Salisb. Out of these 50 compounds, 26 were identified in essential oils through GC-MS. For example, 1,3,5-cycloheptatriene was detected in all six cultivars except cv. Duggirala Red, whereas 12-oxabicyclo(9.1.0)dodeca-3,7-diene, 1,5,5,8-tetramethyl-, [1R-(1R*,3E,7E,11R*)]-, was detected only in cvs. Duggirala Red and Kasturi Araku. The rest 24 compounds were detected in rhizome extracts by LC-MS (e.g., chavicol detected in cvs. Alleppey Supreme and Suguna). The present extensive analyses of both essential oils and whole rhizome secondary metabolome of seven cultivars of C. longa L. and C. aromatica Salisb. established cultivar variability. Variability of the compounds within or/and in between the cultivars of C. longa L. and C. aromatica Salisb. will give a better understanding of their selection. The current study will help select cultivars for use in pharmacology or the food industry.

Discovery of First-Time Reported
Metabolites in C. longa L. and C. aromatica Salisb.
Some of the compounds identified in our study were present in more than one cultivar. For example, 6-(p-tolyl)-2-methyl-2heptenol ( Table 3) was detected in three cvs.: Alleppey supreme, Suguna of C. longa L., and Kasturi Avidi of C. aromatica Salisb. This compound was earlier reported from Zingiber officinale Roscoe (Zingiberaceae), used as a spice, food products, and beverages (Choudhari and Kareppa, 2013).

Limitations and Strengths of the Present Study
There is significant variability within and between the cultivars of C. longa L. and C. aromatica Salisb, which can be exploited to differentiate the cultivars of Curcuma spp. The feasibility of studies without using any standard compounds was pointed out by Núñez et al. (2020). Similarly, reference compounds were not used in our study to derive arithmetic indices under the experimental conditions. Despite the dilution made in the essential oil sample before injecting into the GC-MS system, the sample was still too concentrated. The high concentration of oil might have restricted the resolution due to overloading the detector. This could be the reason that we could not identify several compounds. We would ensure the further dilution of the oil sample in our future studies. However, the technology employed, GC-TOFMS and LC-QTOFMS, and MS-spectral database/literature search enabled us to establish the cultivar variability of Curcuma spp. The detailed information on the metabolite variability within or/and between the cultivars of C. longa L. and C. aromatica Salisb. may assist us in selecting the cultivars for a specific purpose, like culinary use, coloring, or pharmacological purpose. The studies such as the present one can help to select cultivars, particularly for use in pharmacology or the food industry. Metabolite variability poses a challenge in the use of turmeric in therapy. The practitioners need to be quite careful and use the identified cultivar and avoid mix-up. The caution applies to commercial/industrial use. Once standardized, the protocol should ensure the use of a specific cultivar. Our GC-MS and LC-MS-based metabolite identification is distinct from chemophenetic studies but is a complementary approach to characterize the Curcuma metabolome.

Importance of Curcuma spp. Metabolites for Human Health
Curcuminoids (CU, DMC, and BDMC) were identified as the main bioactive compounds of genus Curcuma and proved to have a broad spectrum of biological activities based on pharmacological studies. However, rhizomes and their essential oils of Curcuma spp. contained several other 8 | (Continued) Compounds commonly detected by LC-MS analysis of rhizomes extract of all seven cultivars of Curcuma spp.: five of Curcuma longa L. (cvs. Alleppey Supreme, Duggirala Red, Prathibha, Salem, and Suguna) and two of C. aromatica Salisb. (cvs. Kasturi Araku, Kasturi Avidi). The structures for the compounds in the serial numbers 10, 5, 6, 15, and 1 are given in Figure 4 (panels 13, 14, 15  bioactive (volatile and nonvolatile) compounds. A summary of the pharmacological studies with the metabolites detected in the present study is given in Table 9. Some of the studies demonstrated therapeutic activity with the isolated metabolites, e.g., carvacrol (Suntres et al., 2015), p-cymene (De Oliveira et al., 2015), which are commonly found in essential oils of Curcuma spp. A few other reports correlated anti-inflammatory and antioxidant properties of C. longa L. essential oil with its chemical components ar-tumerone, α-santalene (Singh et al., 2010) (Table 9). Several compounds detected in the present study in the essential oil or rhizome extracts of C. longa L. or C. aromatica Salisb. were also found in the essential oil of other medicinal plants, traditionally used for their health benefits. The examples of such compounds are 5,9-tetradecadiyne, a cultivarspecific compound of Duggirala Red (C. longa L.), earlier reported in Ferula vesceritensis Coss. & Durieu ex Trab. leaf essential oil, exhibiting antibacterial activity; 3-octen-5-yne, 2,7-dimethyl-, (Z)-, a hydrocarbon monoterpene, identified from cv. Kasturi Avidi (C. aromatica Salisb.) was earlier reported from fruit essential oil of the medicinally important plant, Litsea glutinosa (Lour.) C.B. Rob. (Chowdhury et al., 2008a). We suggest that the medicinal use of the genus Curcuma can be not only species but also cultivar-specific.

CONCLUDING REMARKS
Essential oils from spices and aromatic plants are enriched with bioactive metabolites, easily isolated and used, unlike the difficulties encountered with synthetic chemical products. The low mammalian toxicity and biodegradable nature of the natural secondary products provide an attractive option to develop them also for crop protection. Metabolomics is a practical and dynamic approach to make a comprehensive study. Both GC-MS and LC-MS techniques should be used to characterize the metabolite profiles of as many cultivars as possible for building a reference library. Preparative LC can be helpful to collect individual metabolite fractions and establish their identity. Several metabolites detected in 7 selected cultivars of Curcuma spp. by GC-MS and LC-MS analyses are reported first time in Curcuma spp. We suggest that the seven Indian cultivars of Curcuma spp. employed in our study can be used as sources of such compounds. Highthroughput analysis of cultivar-specific and first-time detected compounds in the present study may lead to new drug candidates. The metabolites validated for their medicinal or other users can be quantified using simple techniques such as HPLC or TLC to ensure their presence in the herbal preparations.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ebi.ac. uk/metabolights/MTBLS2790.

AUTHOR CONTRIBUTIONS
ST and AR planned and designed all the experiments. PK and PKK did the experiments and preliminary data interpretation related to GC-MS and LC-MS/MS, respectively. PK, ST, and AR re-analyzed the data and wrote the draft of the manuscript. SA prepared tables and reference search. AA drew the figures and helped in editing the manuscript. All the authors read and approved the final version of the manuscript.