Keywords

1 Introduction

According to Food and Agricultural Organization, more than 50,000 plant species have medicinal values across the globe (Schippmann et al. 2002). The biological diversity of medicinal plants and their vivid chemical capabilities have been used by man for the treatment of various diseases. This brought forth to traditional medicinal practices such as Ayurveda, Unani, Chinese, Middle Eastern, and African systems of medicine (Mamedov 2012). The conventional detection methods of plant metabolites mainly relied on phytochemical screening tests followed by simple techniques of separation such as thin layer chromatography (TLC). Isolation of metabolites by these methods is often tedious and not much efficient. Also the poor and low selectivity of detection makes hard the detection of trace amount of metabolites in the sample (Srivastava et al. 2014). A thorough understanding of the entire metabolites of plants became inevitable for discovering novel metabolites, along with their concerned biosynthetic genes. Thus, metabolomics, an ongoing extension of “omics”, has emerged as a powerful tool (Sumner 2010).

Currently, there is an increased attraction toward herbal remedies, as the green medicine is assumed to be safe and eco-friendly. Rising incidences of harmful side effects from using synthetic drugs also persuade human beings to utilize natural products for various ailments (Chanda 2014). Along with this, there is an increased attention toward natural products; novel approaches are also developed to make a clear understanding of the mode of action of these green remedies, as demonstrated by the integrated “omics” approaches, which provides a holistic approach in drug research (Gonulalan et al. 2020).

A large number of phytochemicals with pharmacological value have been isolated and play predictable therapeutic roles in the clinical world (Gonulalan et al. 2020). The therapeutic effects exhibited by many of the medicinal plants are not always due to a single major compound, but in many cases due to synergistic or antagonistic activity of various compounds. If active compounds are present in very minute quantities in plants, it can affect its chemical characterization, and the process is very challenging also (Williamson 2001). In order to address this task, metabolomics can contribute an important role; it provides insight into the metabolome of plants. Metabolomics is observed as extension of the ongoing high-throughput technologies to the comprehensive analysis of small molecule metabolites of biological system (Sumner 2010). Besides, metabolomics confers several benefits as compared to other “omics;” this is because the biochemical phenotypes are not always represented by transcriptomic or proteomic approaches (Fernie and Stitt 2012). Metabolomics aims to measure the metabolites in the sample and provides information about the interconnection of metabolic pathways and the phenotype of the biological system (Johnson et al. 2016).

Metabolomics is also employed for obtaining valuable data to aid in discovering novel genes and the concerned pathways. The advancements in the sequencing technologies and metabolome-based genome-wide association study (mGWAS) are effectively utilized to unveil the genetic mechanisms behind diverse metabolome and how they are associated to the complex traits seen in plants (Hong et al. 2016). Wen et al. (2014) conducted mGWAS study in maize kernels, through which they identified 1459 locus-trait associations among three environments. Resequencing and analysis of association among candidate genes led to the identification of causal variants of five genes concerned with metabolic traits. Likewise, by combining metabolomics and transcriptomics, all genes of seco-iridoid pathway in Catharanthus roseus have been identified (Miettinen et al. 2014; Salim et al. 2014).

As per the estimates of the World Health Organization (WHO), about 80% of world’s population use traditional medicines at a certain period of their lifetime (Chintamunnee and Mahomoodally 2012). Traditional herbal medicines are widely distributed across the globe and have been used for centuries without any rigorous rules (Efferth and Greten 2012). Together with this, there are chances for adulteration as products of medicinal plants are marketed and distributed (Perini et al. 2018). The bioactive compounds responsible for the therapeutic properties also need to be identified in order to do semi-synthesis and further development of novel new drug candidates from these lead compounds (Plazas et al. 2019).

In this context, metabolomics is a promising tool that can be effectively used to mine various bioactive metabolites from different medicinal plants (RaoGajula and Nanjappan 2021). In this chapter, we present how metabolomics approach can effectively be employed for medicinal plants. The chapter also discusses various strategies, technical advancements, data processing methods, and databases. In addition to this, various applications of metabolomics in connection with medicinal plants are also addressed.

2 Strategies Employed in Plant Metabolomics

Before the advent of “omics” era, investigation and extraction of bioactive metabolites from plants mainly relied on conventional extraction techniques like Soxhlet extraction, hydrodistillation, and maceration. The efficiency of these methods mainly depends on the solvents selected, polarity of selected compound, etc. (Azmir et al. 2013). Also, traditional methods of separation and bioassay-guided fractionation are too laborious as well as expensive. Hence, they are not cost-effective in an industrial approach for drug development (Yuliana et al. 2013). Revolutionizing changes have been made by the metabolomics technologies, which aim for targeted and global profiling of metabolites of the sample. The metabolomic profiling of medicinal plants is becoming crucial and aids in the development of novel phytotherapeutics (Shyur and Yang 2008). As herbal formulations are multi-compound medicines, metabolomic approach is vital for screening of multi-compounds. Hence, metabolomic approaches are gaining an upper hand in medicinal plant research (Lee et al. 2017). In addition to this, the technical advancements in mass spectrometry (MS) and nuclear magnetic resonance (NMR) make possible the estimation of a wide array of compounds and comparing the support data of novel compounds with natural products library (RaoGajula and Nanjappan 2021). The entire process of metabolomic analysis to derive meaningful data from the metabolome of medicinal plants is outlined in Fig. 11.1. Some of the major strategies employed in metabolomics are briefly discussed below.

Fig. 11.1
A flow diagram as a process of metabolomic analysis contains a multi-step workflow of metabolomics. It starts from sample preparation to data preprocessing with the number of applications and databases.

The overall process of metabolomic analysis to mine useful data from medicinal plants

Full size image

2.1 Metabolite Profiling

This involves the identification and quantification of a large group of compounds, sharing similar chemistry or related by metabolic pathways (Shafi and Zahoor 2021). Metabolic profiling includes nontargeted and targeted approaches.

2.1.1 Nontargeted Approach

The nontargeted/untargeted approach involves simultaneous measurement of a possible large number of metabolites in the sample (Shah et al. 2012; Schrimpe-Rutledge et al. 2016). This approach is appropriate for biomarker discovery and is employed for generating hypothesis. It is also called undirected or unbiased metabolomics (Wang et al. 2010). However, it cannot be regarded as truly unbiased because the researcher has to choose a combination of ionization mode and stationary phase, which in turn facilitates the detection of certain compounds and lowers the detection of some others (Ribbenstedt et al. 2018). Together with, it faces certain other challenges such as complicated protocols, longtime consumption to handle huge amount of raw data, problems of identification and characterization of unknown metabolites, reliability of the platform chosen, and greater chances of detecting more abundant metabolites (Roberts et al. 2012).

2.1.2 Targeted Approach

Metabolic profiling of known metabolites with distinct identities (already defined compounds) is done for the targeted approach (Shah et al. 2012). Also called directed or biased metabolomics, it mainly focuses on metabolites of a particular chemical class or any other predetermined group of compounds. This approach is hypothesis-driven and aims to verify biological pathways or authenticate an untargeted metabolomic study (Wang et al. 2010). Here, analysis can be carried out either in a quantitative or semiquantitative manner. Some advantages of targeted analyses like the downstream analysis are almost set free from analytical artifacts, and novel metabolic associations can be deciphered at particular physiological states (Roberts et al. 2012). Major metabolomic studies in medicinal plants, which employed targeted/untargeted approaches, are given in Table 11.1.

Table 11.1 List of metabolomic studies of medicinal plants via targeted/untargeted approaches (from 2010 onward)
Full size table

2.2 Metabolite Fingerprinting

This involves an untargeted approach, in which extensive identification and quantification of individual metabolites are not done. The data obtained were analyzed to recognize patterns specific to the fingerprint of metabolites at a given biological state, for example, stress response, etc., provided in a living system (Wolfender et al. 2015; Shafi and Zahoor 2021). It mainly finds applications in comparing fingerprints or patterns of metabolites that vary with response to a particular disease condition, environmental changes, etc. (Barderas et al. 2011). The major platforms employed here include NMR spectroscopy, Fourier transform infrared spectroscopy (FTIR), and Raman spectroscopy (Ellis et al. 2007). Gray and Heath (2005) examined cold acclimation effects on the metabolome of Arabidopsis using metabolite fingerprinting.

2.3 Metabolite Footprinting

This approach comprises the profiling of extracellular metabolites alone (Pope et al. 2007). It mainly focuses on the biochemical and chemical changes brought about by organisms due to the effect of its immediate environment. Every living cell modifies its medium by the secretion of various enzymes and other metabolites; later, a subsequent interaction of the secreted components with the constituents present in the medium happens. This will generate metabolic profiles that are specific to particular species and/or their genetic makeup (Villas-Bôas et al. 2006). As it involves secreted metabolites, sampling is relatively simple in which it only requires the procedure of separating the media from the cell/tissue. Further analysis was done by means of NMR (Filloux and Ramos 2014). The procedure and differences in metabolite footprinting and fingerprinting are outlined in Fig. 11.2.

Fig. 11.2
A flow diagram depicts the sample's metabolite fingerprinting and footprinting, which leads to spectroscopic analysis and data processing. Data processing leads to ecotypes 1 and 2 and a unique footprint.

Workflow depicting the procedure and differences of metabolite fingerprinting and footprinting

Full size image

2.4 Metabonomics

Metabonomics implies extensive evaluation of metabolic changes in a living system. It does quantitative measurement of the metabolic variations of a biological system in response to genetic modification or pathological/physiological stimuli (Bjerrum 2015).

2.5 Chemoprofiling

The chemical constitution of plants needs to be understood to ensure sufficient therapeutic evaluation (Efferth and Greten 2012). Chemoprofiling includes pattern-oriented/multicompound approach and compound-oriented/marker approach. Pattern-oriented approach focuses on all detectable compounds in a plant extract that forms a unique fingerprint, and it does not characterize all the compounds in a plant extract. In the case of compound-oriented approach, some of the bioactive or major constituents are considered for obtaining specific chemoprofiles (Govindaraghavan et al. 2012).

2.6 Metabolic Fluxomics

Rates of metabolic reactions (fluxes) provide information about metabolic phenotypes and cellular regulation mechanisms. The direct measurement of metabolic fluxes is not possible, and it needs to be obtained by measuring other observables like changes in concentration of enzymes, metabolites, carbon balance, etc. (Niedenführ et al. 2015). Flux analysis by means of 13C fluxomics (using 13C, 14C, 2H, 15N isotopic traces) has become a method of choice to decipher the regulation and constitution of metabolic networks (Zamboni 2011; Niedenführ et al. 2015).

3 Technological Advancements in Plant Metabolomics

3.1 Mass Spectrometry Imaging (MSI)

MSI works by ionizing the peptides/protein or metabolites from the biological sample in a two-dimensional or three-dimensional coordinate (Fletcher et al. 2008; Seeley and Caprioli 2012).

3.1.1 Gas Chromatography-Mass Spectrometry (GC-MS)

GC-MS is routinely used for the analysis of organic compounds that are volatile or can be derivatized to make them volatile (Hall 2006). Separation of compounds happens due to the difference in partition coefficients between stationary (solid) phase and mobile (gas) phase (Smedsgaard 2007; Roessner and Beckles 2009). By employing electron impact ionization (EI), analytes are ionized; it results in a unique fragmentation pattern for every constituent phytochemical. Identification of compounds can be made by comparing the GC-MS fragmentation patterns and retention time, along with the available information in the databases of GC-MS (Kopka et al. 2004). GC-MS method gains popularity because it helps in the determination of amino acids, sugars, and organic acids. One of the limitations of GC-MS is that the overlapping peaks of GC-MS chromatogram render the detection of individual metabolite signals much more difficult (Saito and Matsuda 2010).

3.1.2 Liquid Chromatography-Mass Spectrometry (LC-MS)

LC-MS forms a valuable tool to unveil the immense wealth of phytochemicals including both primary and secondary metabolites. As opposed to GC-MS, LC-MS can handle a wide range of compounds with different chemical properties, for example, they are either volatile or not (Hill and Roessner 2015). In addition to this, partially purified crude extracts can be directly fed into the LC/MS system. This eliminates various steps of sample treatment. The introduction of ultra performance liquid chromatography (UPLC) together with high-resolution Fourier transform (FT) MS, time-of-flight (TOF) MS, and Orbitrap-based MS made beneficial advancements in LC-MS-based metabolomics (Salem et al. 2020a).

3.1.3 Capillary Electrophoresis-Mass Spectrometry (CE-MS)

Capillary electrophoresis employs high voltage for the electrophoretic separation of different ions in a narrow-bore capillary (Ren et al. 2018). The initial separation of metabolites takes place according to their charge-to-size ratio followed by mass-to-charge ratio-based separation (Zhao et al. 2012; Klepárník 2015). Even though CE-MS is considered as a novel method in metabolomics, it suffers from drawbacks of having poor stability, and there are chances for capillary blockage by salt (Ren et al. 2018). However, it performs the ability to separate low volume of biological fluids, simple sample preparation protocol, capability of concentrating analytes and better separation efficiency, making it as a method of choice (Klepárník 2015; Ren et al. 2018).

3.2 Nuclear Magnetic Resonance (NMR) Spectroscopy

NMR spectroscopy is a chief tool for studying plant metabolomics. It gains an upper hand over other techniques by facilitating high-throughput analysis, easy sample preparation, rapid performance, and easy quantitation (Kim et al. 2011). NMR makes possible the analysis of profuse primary metabolites and heterogeneous secondary metabolites such as alkaloids, flavonoids, terpenoids, etc. As the signals in the NMR spectrum are indicative of the molar concentrations of the component compounds, concentrations of these compounds can be compared with other samples, and it eliminates the need for calibration curves for each compound (Kim et al. 2010). NMR also suffers from certain disadvantages. Its low sensitivity demands larger amount of sample than other methods. Also, there is considerable overlapping of signals in the NMR spectra impeding accurate signal identification and peak integration (Kim et al. 2010; Halabalaki et al. 2014).

3.2.1 2D NMR

Signal overlap is a real problem in 1D NMR, hindering both the identification and quantification of compounds in complex plant extracts. Hence, 2D NMR spectroscopy can be employed, which provides good signal resolution by distributing the resonance in a second axis. One limitation of this method is the lengthy acquisition time as compared to 1D NMR (Kim et al. 2011). J-resolved spectroscopy and heteronuclear single quantum coherence (HSQC) are the major useful 2D methods (Ludwig and Viant 2010; Salem et al. 2020a). HSQC is particularly employed as a confirmation tool to detect whether the usual suspected compounds (common primary and secondary plant metabolites) are present or not and for the quantification of metabolites. 1H-1H COSY (correlation spectroscopy), NOESY (nuclear overhauser effect spectroscopy), HMBC (heteronuclear multiple bond correlation), etc. are few important 2D NMR-based methods (Kim et al. 2011; Salem et al. 2020a).

3.2.2 Solid-State NMR

Solid-state NMR allows the examination of semisolid samples such as tissue samples via the use of high-resolution magic-angle spinning (HRMAS). This technique involves mixing of the sample with a minimal solvent volume followed by rotating the sample at 54.74° with high spinning rates (Pérez et al. 2010; Kruk et al. 2017). Better resolution can be achieved by HRMAS. Nevertheless, the faster spinning and high temperature may cause tissue distortion. Also, the quantification of metabolites poses a challenge in HRMAS, which can be overcome by a method called ERETIC (Electronic REference To access In vivo Concentration) (Kruk et al. 2017).

4 Data Processing Methods

Large amount of data is generated in metabolomics analysis through the analytical platforms such as NMR and MS (Liland 2011). Meaningful data can be mined from this complex data set through multivariate analysis (MVA) methods like discriminant analysis, discrimination map analysis, and principal component analysis (PCA). These methods are mainly focused on the reduction of data such as m/z values from MS analysis, chromatography data, NMR data, etc. The reduced data are presented as discrimination maps, score pots, load plots, etc. according to the MVA method employed. The raw data from the analysis are mathematically and statistically processed, and that is represented in the form of vectors in score plot. These vectors correspond to the metabolite fingerprint of the tested sample in the case of score plot, whereas in the loading plot, each signal’s contribution is represented. The discrimination map, in addition to visualizing the contribution of each signal, also represents the variation in signals. Thus, the marker metabolites can be selected by statistical means by connecting the loaded factor and the data from chemical analysis (Okada et al. 2010).

Another strategy to handle this enormous metabolomic data involves the utilization of software tools. These software packages process the raw data of spectra and also perform statistical analysis in order to find out the metabolites, which are significantly expressed in the sample; later, the metabolites are compared with metabolite databases, and subsequently, multiple “omics” data is integrated and analyzed; finally, it helps to visualize molecular interactions also (Krastanov 2010; Sugimoto et al. 2012). MetaboAnalyst, MetaCore™, and InCroMAP are some of the versatile software tools used for this purpose (Cambiaghi et al. 2016). Among these, only MetaboAnalyst is a comprehensive tool that can perform both the pre-processing of data and its statistical analysis (Xia et al. 2015). MetaCore™ is an integrated database, which makes possible the visualization and functional analyses of various kinds of omics data together with options for biological pathway analyses. This versatile tool also finds applications in biomarker identification to drug discovery process (Cambiaghi et al. 2016). InCroMAP is user-friendly software, which can generate global maps of metabolic processes in cell from the metabolic analysis (Wrzodek 2012; Wrzodek et al. 2012). It suffers a drawback, as it can’t perform data pre-processing (Cambiaghi et al. 2016).

5 Databases for Handling Metabolomics Data

The field of plant metabolomics is advancing rapidly, the ultimate aim of which is a holistic understanding of various functions and healing potential of medicinal plants (Afendi et al. 2012). Therefore, to achieve profile management, metabolite identification, effective data mining, and efficient platforms are required (Ferry-Dumazet et al. 2011). For making the metabolomics data meaningful, it has to be organized in a standard form that allows cross-referencing with other datasets. Several specialized metabolomics databases for plants are available, and these databases provide updated and comprehensive information (Shafi and Zahoor 2021). In order to make metabolomics as a valuable tool for functional genomics, it requires the availability of annotated metabolomics data, and it can be accessed through internet. Visualization tools (such as error and ratio plots), if integrated into the databases, would help in the comparison of metabolome at various conditions such as environmental changes, genetic disturbances, and variation in experimental parameters to which the biological system is subjected and would find the effective analytical platform that can effectively communicate the maximum metabolomic changes according to these changing conditions (Bais et al. 2010). The different databases, which handle metabolomics data of medicinal plants, are indicated in Table 11.2 (i.e., from 2010 onward). Various applications of databases in medicinal plant research are schematically represented in Fig. 11.3.

Table 11.2 List of metabolomics databases of medicinal plants (from 2010 onward)
Full size table
Fig. 11.3
A cyclic schema of plant metabolomics databases for medicinal plants. It contains various applications such as in-silico, identification, integration, three-dimensional structure, modeling, annotation, cross-species and cross-platform.

Various applications of plant metabolomics databases

Full size image

6 Applications of Metabolomics in Medicinal Plant Research

6.1 Investigating the Bioactive Compounds in Medicinal Plants

Inability to efficiently identify potential bioactive compounds from medicinal plants is a bottleneck in their extensive clinical applications, masking the wealth of efficient lead compounds for drug industry (Plazas et al. 2019). Metabolomics finds immense applications in unveiling the bioactive compounds responsible for the therapeutic effects of medicinal plants (Yuliana et al. 2013), and some of the examples are briefly outlined. The treatment of Alzheimer’s disease is posing a great challenge, because of the absence of efficient drug candidates (Rahman and Choudhary 2015). Only viable option is to focus on drugs that act as cholinesterase inhibitors. To screen for anti-cholinesterase alkaloids found in Zanthoxylum members of Rutaceae, metabolomics profiling of its nine species has been performed along with chemometric analysis. This experiment resulted in the detection of 11 isoquinoline alkaloids with potential anti-cholinesterase activity (Plazas et al. 2019). To explore the potential of Camptotheca acuminata for antineoplastic treatment, metabolite fingerprinting has been performed, and they identified the alkaloids related to camptothecin having antineoplastic action. Leaves at different growth stages were analyzed to find the optimum stage of growth for harvesting. Various camptothecin-related alkaloids revealed through the study offer promising compounds to be utilized as precursors in the synthesis of semisynthetic derivatives of camptothecin (Montoro et al. 2010). Li et al. (2013) performed metabolic profiling of Tussilago farfara to investigate the compounds responsible for its expectorant and antitussive activities. They also analyzed different plant parts to optimize the part with maximum activity; it revealed that flower buds and leaves possessed maximum bioactive principles. The study confirmed the role of 3,5-dicaffeoylquinic acid, rutin, and chlorogenic acid in the therapeutic properties of the plant.

Carica papaya has broad spectrum of therapeutic applications such as anti-inflammatory, antibacterial, anticancer, antioxidant, antimalarial, vasodilatory properties, etc. Metabolic fingerprinting performed through 1D and 2D 1H NMR revealed that high concentration of secondary metabolites occupied in younger leaves and UPLC-ESI-MS analysis verified and confirmed the active metabolites. Apart from attributing the therapeutic effects to one or two major compounds, the role of synergism of metabolites in its diverse clinical applications was revealed in this study (Gogna et al. 2015). Taha et al. (2020) investigated the antimicrobial, anticancer, and antioxidant activities of a desert medicinal plant Hyphaene thebaica. Past studies in this plant were mainly concentrated on the clinical applications of the fruit, while this study utilized leaves, fruit, and male parts of the plant. These three parts were evaluated in three groups according to the anthocyanin, flavonoid, flavonol, phenolic, saponin, and tannin content. Significant increase in antioxidant activity was found in male parts and leaves than fruits. Metabolic profiling by HPLC confirmed the dominance of chrysin, p-hydroxybenzoic acid, p-coumaric acid, protocatechuic acid, syringic acid, rosmarinic acid, and vanillic acid in male parts; chlorogenic acid in fruit; and apigenin-7-glucosides, catechins, and rutin in leaves. Some of the important bioactive compounds from medicinal plants studied using metabolomics are given in Table 11.3.

Table 11.3 Major bioactive compounds from medicinal plants studied using metabolomics
Full size table

6.2 Elucidate the Mode of Action of Herbal Medicine

There are concerns over toxicity, efficacy, quality, etc. of many traditional medicinal herbs. Metabolomics is a valuable tool to answer these concerns. Herbal formulations of traditional Chinese medicine (TCM) have wide clinical applications in treatment of liver diseases, which were largely limited due to queries over safety and quality. Metabolomics has been employed to decipher the mode of action of three herbal remedies (yin chen hao tang, xiaozhang tie, and silymarin) for liver disease in TCM through studies in cell culture systems, animal models, and clinical studies. This study provides insights to the efficacy and safety of herbal remedies and also underlines the importance of combinations of herbal medicines in efficient treatment methods (Beyoğlu and Idle 2020). Likewise, the ability of Suanzaoren decoction to treat insomnia has been delineated by assessing the metabolic changes taking place in a model insomnia drosophila after administering Suanzaoren decoction. This decoction is a formulation of five herbal medicines (Poria, seed of Ziziphus jujuba; roots and rhizome of Glycyrrhiza uralensis, G. inflata, and G. glabra; rhizome of Ligusticum chuanxiong; and rhizome of Anemarrhena asphodeloides). The analysis revealed that Suanzaoren decoction could significantly increase sleep activity. In addition to this, the hypnotic effect of this formulation is found to affect the global metabolomics of the test organism (Yang et al. 2012).

6.3 Quality Assessment of Herbal Products

The capabilities of natural product derived drugs can be utilized in full potential only if the chemical composition of herbal products is standardized to check the proper quality of natural product (Heyman and Meyer 2012). Quality of herbal formulations has been assessed earlier based on one or two major compounds present in it. However, herbal medicines comprise multiple compounds, and metabolic profiling can be performed to evaluate the multiple components, thereby meeting adequate quality standards (Lee et al. 2017). Metabolomics is gaining wide acclaim nowadays to assess phytochemical constituents, thus providing an acceptable method for quality control of herbal medicine. By employing chromatographic and spectroscopic methods, a valid metabolite fingerprint can be obtained, and further metabolite profiling helps to identify the individual constituents that form the unique fingerprint (RaoGajula and Nanjappan 2021).

For even properly authenticated plant material, there may be difference in quality between different batches due to a variety of factors such as inter or intra-species variations, environmental factors, harvesting stage, plant parts used, post-harvesting factors, etc. (Nafiu et al. 2017). Xiang et al. (2011) employed metabonomic analysis to characterize ecotypic variation in three species of Curcuma, namely, Curcuma kwangsiensis, C. phaeocaulis, and C. wenyujin as part of quality control measures. This study was focused on the essential oil composition, and PCA efficiently distinguished samples according to differences in species and ecotypes. Some of the medicinal herbs are difficult to distinguish between varieties as in the case of Ficus deltoidea (a popular medicinal herb of Malaysia), whose seven varieties are laborious to identify morphologically due to extensive heterophylly. The difficulty to identify the desired variety and variations of their chemical constitution is posing challenges in their commercialization. Untargeted metabolomics is performed by ultra-high-performance liquid chromatography time-of-flight mass spectrometry (UHPLC-TOFMS), and the subsequent data analysis was able to distinguish three chemotypes on the basis of differences in flavonoid content. The study also identified 15 glycosylated flavones and 1 furanocoumarin as chemical marker (Afzan et al. 2019). Saposhnikoviae Radix (a common crude drug) obtained from rhizome and root of Saposhnikovia divaricata was investigated for the differences in metabolome using the specimens from China and Mongolia. Metabolic profiling confirmed that these two regional groups are clearly distinct with respect to O-glucosylcimifugin, being more abundant in Mongolian group. They can also be distinguished based on the differences in content of eight chromones (Batsukh et al. 2020). Rastogi et al. (2020) investigated the interspecies variation among three medicinally important species of Ocimum, namely, O. gratissimum, O. kilimandscharicum, and O. sanctum by analyzing the temporal changes in metabolite composition.

Adulteration of medicinal herbs with certain other plants or any other foreign substance is a common practice and a major challenge of natural drug industry (Chanda 2014). Wallace et al. (2018) evaluated the adulteration in commercial plant-derived products by using supplements of Hydrastis canadensis (goldenseal) as the test case. An untargeted metabolomics analysis employing UPLC-MS helped to detect adulteration in 3 test samples among the 35 test samples. This analysis revealed the potential applications of untargeted metabolomics in detecting possible adulteration. Extensive adulteration is observed in highly priced oil of Serenoa repens (saw palmetto), which triggered the investigation for methods to detect designer blends of cheap fatty acids as adulterants. The combined use of metabolomic analysis and isotopic fingerprinting suggested the possible adulteration and whose source is mostly animal based fatty acids (Perini et al. 2018).

6.4 Safety and Toxicity Assessment of Natural Products

One of the factors that hinder the clinical applications of traditional phytomedicine into mainstream is their limited molecular level characterization. If there is proper recording of the side effects, reaction with other medicines, safety, hypersensitivity, tolerance, problems of overdose, etc. of natural products, then only the product can be commercialized (Cordell 2015). Nowadays, traditional Chinese medicine (TCM) is attracting wide attention due to debates over safety concerns. Even though TCM contains natural formulations, their complex nature and mechanism of action remain as a barrier to assess it by traditional methods. For this, metabolomics helps to get an idea about the possible toxicity of bioactive compounds in TCM. Cardiotoxicity, hepatotoxicity, nephrotoxicity, and reproduction toxicity have been revealed as the side effects many of the components of TCM by metabolomic studies (Duan et al. 2018). Traditional African medicine (TAM) also has much popularity among African communities with wide acceptance among both rural and urban populations. In this context, metabolomics provides an opportunity to make a holistic analysis of phytochemicals, biomarkers, and the mechanism by which TAM modifies metabolic pathways (Quansah and Karikari 2016).

In order to investigate the different levels of toxicity exhibited by Senecio scandens and S. vulgaris, a metabolomics study employing UPLC-MS was performed. Senecio scandens, which is an approved medicine of Chinese medicine, did not show toxic effects, while S. vulgaris belonged to the same genera exhibited significant hepatotoxicity. The metabolomics analysis revealed that senecionine, a particular marker of S. vulgaris remained the cause for its toxic effects (Xiong et al. 2012). Polygonum multiflorum has been a part of TCM, whose dry roots are widely used for therapeutic properties. Hepatotoxicity is a concern over its use, and different processing methods are documented for its processing to reduce the toxicity. With the aim of finding the most suitable method of its processing, variations in metabolomic profiles of P. multiflorum with respect to various processing methods were analyzed by UHPLC/Q-Orbitrap-MS. The study identified emodin-8-O-glucoside and torachrysone-O-hexose as the toxic markers of P. multiflorum. The toxic effects were decreased with every processing method and the best results being obtained by steaming with black soybean; the method was suggested by Chinese Pharmacopoeia (Han et al. 2019).

7 Conclusions and Future Prospects

A large proportion of world’s population still relies on medicinal plants for therapeutics and our traditional medicinal systems use herbal formulations, it was tested and evaluated by trial-and-error mechanisms. Even though these medicinal plants have such a history of long-term association with human therapeutics, systematic characterization of their active compounds and the intricate mechanisms underlying their action have been achieved with the advent of metabolomics. Metabolomics is a promising approach, which can revitalize the researches in herbal medicine by characterizing the active compounds in medicinal plants responsible for their clinical action. Metabolomics helps to study the synergism between various compounds in attaining the therapeutic effects, assessing the quality, safety, and toxicity of herbs and herbal formulations, etc. The rapid analysis of large number of metabolites in metabolomics is what makes it distinct from other analytical methods. It can also contribute much to the upcoming era of personalized medicine by metabolic profiling of the concerned individual in response to the herbal drug administration; this helps to monitor the efficacy and toxicity of the drug to each individual. Efficient analytical platforms, data handling methods, and associated databases are crucial to the success of metabolomics. Recent advancements in MS- and NMR-based methods have contributed much to the progress of medicinal plant metabolomics. If these two platforms are employed together, metabolomics has much potential to provide a non-biased quantification and characterization and can provide an integrated picture of component metabolites in the sample. Such a comprehensive knowledge regarding the medicinal plants and herbal formulations can open new avenues in herbal clinical industry. Thus, the treasure trove of phytochemicals as potential drug leads can be explored in future.