Phytotoxicity induced by engineered nanomaterials as explored by metabolomics: Perspectives and challenges

Ecotoxicology and Environmental Safety 184 (2019) 109602

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv

Review

Phytotoxicity induced by engineered nanomaterials as explored by metabolomics: Perspectives and challenges

T

Xiaokang Lia, Ting Penga, Li Mub,∗, Xiangang Hua a

Key Laboratory of Pollution Processes and Environmental Criteria (Ministry of Education), Tianjin Key Laboratory of Environmental Remediation and Pollution Control, College of Environmental Science and Engineering, Nankai University, Tianjin, 300350, China b Tianjin Key Laboratory of Agro-environment and Safe-product, Key Laboratory for Environmental Factors Control of Agro-product Quality Safety (Ministry of Agriculture and Rural Affairs), Institute of Agro-environmental Protection, Ministry of Agriculture and Rural Affairs, Tianjin 300191, China

A R T I C LE I N FO

A B S T R A C T

Keywords: Engineered nanomaterials Phytotoxicity Metabolomics Metabolic pathway

Given the wide applications of engineered nanomaterials (ENMs) in various fields, the ecotoxicology of ENMs has attracted much attention. The traditional plant physiological activity (e.g., reactive oxygen species and antioxidant enzymes) are limited in that they probe one specific process of nanotoxicity, which may result in the loss of understanding of other important biological reactions. Metabolites, which are downstream of gene and protein expression, are directly related to biological phenomena. Metabolomics is an easily performed and efficient tool for solving the aforementioned problems because it involves the comprehensive exploration of metabolic profiles. To understand the roles of metabolomics in phytotoxicity, the analytical methods for metabolomics should be organized and discussed. Moreover, the dominant metabolites and metabolic pathways are similar in different plants, which determines the universal applicability of metabolomics analysis. The analysis of regulated metabolism will globally and scientifically help determine the ecotoxicology that is induced by ENMs. In the past several years, great developments in nanotoxicology have been achieved using metabolomics. However, many knowledge gaps remain, such as the relationships between biological responses that are induced by ENMs and the regulation of metabolism (e.g., carbohydrate, energy, amino acid, lipid and secondary metabolism). The phytotoxicity that is induced by ENMs has been explored by metabolomics, which is still in its infancy. The detrimental and defence mechanisms of plants in their response to ENMs at the level of metabolomics also deserve much attention. In addition, owing to the regulation of metabolism in plants by ENMs affected by multiple factors, it is meaningful to uniformly identify the key influencing factor.

1. Introduction Engineered nanomaterials (ENMs) have been applied for use in all aspects of life (e.g., construction materials, cosmetics, food additives, sewage treatment and antimicrobial agents) (Gajbhiye and Sakharwade, 2016; Hincapie et al., 2015; Lu et al., 2015; Peters et al., 2016). Recently, nano-enabled fertilizers, nano-enabled crop growth regulators, nano-enabled pesticides, nanostructured biosensors, remediation and biological imaging have attracted considerable attention (Antonacci et al., 2018; Monreal et al., 2015; Wang et al., 2016; Yin et al., 2018). The applications of ENMs in all areas, especially in agricultural production, leads to the contact of ENMs with plants (Sun et al., 2017). Plants, as an essential component of the environment, are the primary producers in ecological systems and critical for ecosystem functions (Wang et al., 2017; Zhou and Hu, 2017). Once ENMs contact with plants, they would be absorbed and transported by plants, ∗

decreasing plant biomass production/quality and transferring by food chains (Majumdar et al., 2016; Zhang et al., 2017a). Therefore, understanding the effects of ENMs on plants is critical for the design of safe ENMs and the scientific assessment of the risks that are associated with the application of nanomaterials (Li et al., 2018b). Numerous studies have been dedicated to assessing the risk of nanomaterials to plants with plant physiological activity, such as plant growth, reactive oxygen species (ROS), and antioxidant enzymes (Rizwan et al., 2017; Tripathi et al., 2017). However, theses plant physiological activity might not provide the global biological information that is important for nanosafety or the comprehensive information that can explain the molecular mechanisms of biological responses (Nel et al., 2012). Metabolites are small molecules that are produced or consumed by biological metabolic processes, which consist of carbohydrates, amino acids, lipids, nucleotides and others (Baker, 2011; Fuhrer and Zamboni, 2015) Metabolites are typical biomarkers

Corresponding author. E-mail address: [email protected] (L. Mu).

https://doi.org/10.1016/j.ecoenv.2019.109602 Received 4 July 2019; Received in revised form 20 August 2019; Accepted 21 August 2019 0147-6513/ © 2019 Elsevier Inc. All rights reserved.

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

2018). NMR has a greater quantitative capability and reproducibility than mass spectrometry, but its sensitivity is lower and quantification is more difficult (Gunsolus and Haynes, 2016; Markley et al., 2017).

that are applied as useful tools for the assessment of adverse effects induced by ENMs at the molecular level (Pagano et al., 2016). Moreover, metabolites are downstream of gene and protein expressions and are directly related to biological phenomena (Baker, 2011). The perturbation of metabolites in response to stress was a sensitive marker (Zhang et al., 2019). Importantly, the dominant metabolites and metabolic pathways are similar in different plants, representing the universal applicability of metabolomics analyses (Kang et al., 2019b; Li et al., 2018a). Herein, the influence of metabolism (i.e., carbohydrate and energy metabolism, amino acid metabolism, lipid metabolism and secondary metabolism) on plants by ENMs will be discussed, and the relationship between metabolism and ecotoxicological end points will be expounded upon. Finally, the factors influencing metabolic regulation will also be analysed to provide an important basis for the future design and application of environmentally friendly ENMs.

2.2. GC-(TOF)-MS To detect metabolites by GC-MS, the metabolites were extracted by a solvent mixture (e.g., methanol, chloroform and water), and ribitol was added as the internal standard. Subsequently, the mixed samples were sonicated, centrifuged and freeze-dried. The samples were derivatized in two steps. First, the samples were incubated with methoxyamine hydrochloride and then with N-methyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA) or bis(trimethylsilyl)trifluoroacetamide (BSTFA, containing 1% TMCS). Subsequently, the samples were analysed by GC-MS (Chen et al., 2019; Zhang et al., 2018b). The derived metabolites were pushed by carrier gas and were then separated by a capillary column. The separated metabolites are sequentially passed into the ion source of the high vacuum mass spectrometry system for ionization. According to the different movement behaviours of the different fragment ions in the electromagnetic field, the information of mass-to-charge ratio (m/z) was acquired to achieve a quantitative and qualitative analysis of the metabolites. The metabolites were identified by comparison with the NIST 14 library by Mass Hunter Qualitative Analysis software or other databases, and only the metabolites with matching scores of > 70% were considered reliable (Chen et al., 2019; Li et al., 2018b). The advantage of GC-(TOF)-MS analysis for metabolomics is that it has high sensitivity, high resolution and a mature standard database (Maag et al., 2015; Tsugawa et al., 2014; Vinaixa et al., 2016). However, the complicated pretreatment (e.g., derivatization) and the bias in the determination of volatile substance and derivable substances with active hydrogen groups has limited the application of GC-(TOF)-MS in metabolomics (Zhang et al., 2017b; Zou et al., 2019).

2. Development of the analytical methods for metabolomics Metabolomics is defined as scientific study of endogenous low molecular weight metabolites (relative molecular weight less than 1000) in various biological systems (Beyoglu and Idle, 2013). Unlike genes and proteins, these small molecule metabolites serve as direct signatures of biochemical activities and are easy to correlate with cellular biochemistries and biological phenomena (Baker, 2011). Compared to the traditional physiological state toxicity test (e.g., germination, biomass, oxidative stress, or enzyme activity), which are limited in that they probe one specific process, metabolomics provides novel modes of nanomaterial-biological interactions to globally judge ecotoxicology and mechanistic information by nanoparticles (NPs) (Gunsolus and Haynes, 2016; Tripathi et al., 2017). Metabolomics captures a snapshot of the cellular metabolic responses to a stressor (Zhao et al., 2018). The primary objective of metabolite analysis is to identify differences in metabolism, and the changes in metabolome in a tissue reflect an inhibition or activation of specific metabolic pathways related to biological responses (Hu et al., 2015a; Zhao et al., 2018). Consequently, metabolite analysis could establish a relationship between metabolic alterations and biological responses and identify the underlying molecular mechanisms of biological responses (Hu et al., 2015a). Before metabolite analysis, the metabolites should be extracted from fresh plant tissues exposed to ENMs. In general, the plant tissues (leaves/roots) are washed with phosphate buffered saline (PBS). Subsequently, the plant leaves/roots are cut using a sharp stainless-steel knife, rapidly frozen in liquid nitrogen to abolish metabolic conversations in tissues (De Vos et al., 2007). Then the metabolites are extracted by various solvents according to different analytical techniques, which are shown as follows. The various analytical techniques that have been employed for metabolomics analysis as listed in Table S1 and mainly involve nuclear magnetic resonance spectroscopy (NMR) and mass spectrometry (MS). MS mainly involves gas chromatography mass spectrometry (GC-MS) and liquid chromatography mass spectrometry (LC-MS) (Salehi et al., 2018; Sun et al., 2018).

2.3. LC-MS and integration of multiple techniques To detect metabolites by LC-MS, the metabolites were vortexed with 70–80% methanol. Subsequently, the mixture was extracted by sonication, centrifugation, filter membrane, evaporation and dissolution by 80% methanol for further LC-MS analysis (Gu et al., 2019; Negrin et al., 2019). The extracted metabolites were separated by a chromatographic column, and the mobile phase and then the isolated metabolites were ionized by the ion source. The metabolites were quantitatively and qualitatively analysed based on their mass spectral information (e.g., m/z, retention time). The major metabolite databases for plant metabolites that we generally used was METLIN (Garcia et al., 2016; Kite, 2018). LC-MS analysis for metabolomics possesses a high sensitivity and selectivity for the metabolites that are difficult to derivatize and volatilize (Chitarrini et al., 2017; Tamura et al., 2018). Given the advantages and disadvantages of a single technique, NMR, GC-(TOF)-MS and LC-MS are frequently employed together. 2.4. Perspectives and challenges in vivo, in situ and in spatial imaging

2.1. NMR The above studies on metabolomics analyses are based on the extraction of tissue metabolites and the subsequently further analysis using NMR or MS. The complicated extraction process will inevitably lead to the loss of key metabolites. To resolve the above issues, more attention should be paid to analysis in vivo or in situ. The in vivo sampling mode of solid phase microextraction (SPME, direct immersion/ headspace) on-line coupled to GC-MS in a fully automatic approach could be used to capture the metabolites of living plant specimens (Risticevic et al., 2016; Sgorbini et al., 2019). Moreover, imaging MS (e.g., Fourier-transform ion cyclotron resonance-mass spectrometry) and live single-cell MS are novel methods for detecting metabolites in situ at the cellular level, which could elucidate the localization of

For the NMR analysis, plant samples were washed, ground and freeze-dried and were then extracted with a solvent system made of a CH3OH-d4 and D2O KH2PO4 buffer, with the addition of a sodium salt of 3-(trimethylsilyl)propionic-2,2,3,3-d 4 acid (TSP, 0.01% w/w) as the internal standard for the NMR analysis. Subsequently, the extracting solution was vortexed, sonicated and centrifuged, and the supernatant was transferred to an NMR tube for further analysis (Cristina Abreu et al., 2018; Ji et al., 2018). The NMR spectra data (mainly standard 1H and 1H-1H correlation spectroscopy) were analysed by comparing the NMR data with references or by structure elucidation using two-dimensional NMR (Kim et al., 2010; Pontes et al., 2017; Pullagurala et al., 2

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

Fig. 1. Metabolite analysis of plants exposed to ENMs. Principal component analysis (PCA) (a) and partial least-squares discriminate analysis (PLS-DA) (b) score plots of metabolic profiles in cucumber leaves treated with different dosage of AgNPs and Ag+. A-E represent control, 0.04 mg and 0.4 mg of Ag ions, and 4 and 40 mg of AgNPs, respectively (Zhang et al., 2018a). (c) VIP scores from the PLS-DA analysis showing the discriminating metabolites induced the group separation. Control, low, medium and high in the figure indicates concentrations of CuNPs of 200, 400, and 800 mg/kg, respectively (Zhao et al., 2017a); (d) summary of the pathway analysis of rice exposed to nano-TiO2 on the MetaboAnalyst 3.0. Every point represents a metabolic pathway that is coloured according to the log of P value. The redder colour represents a more significant effect, and the size is determined by the degree of influence (Wu et al., 2017).

PCA score plot revealed that AgNPs were clearly separated from the control along the first principle axis (PC1), which explained 25.4% of the total variability (Fig. 1a) (Zhang et al., 2018a). A supervised PLS-DA analysis maximizes the separation between groups using multiple linear regression technique and helps in understanding which variables carry the class separating information. As shown in Fig. 1b, the PLS-DA score plot showed that all exposure groups (AgNPs and Ag+ groups) were generally separated from the control along component 1 in a dose-dependent manner. The above results indicated that metabolic profiles were perturbed by both AgNPs and Ag+ (Zhang et al., 2018a). VIP is the weighted sum of the squares of the PLS-DA analysis and indicates the importance of a variable to the entire model. To further determine the metabolites that were responsible for the separation, the metabolites with VIP scores > 1 after treatment with different concentrations of CuNPs were identified as the

metabolites in plant tissues (Stopka et al., 2019; Yamamoto et al., 2016). 2.5. Performance of metabolite data Metabolite perturbations in plants occur in response to stress (Ghorbanpour and Hadian, 2015; Hong et al., 2016). Principal component analysis (PCA) and partial least-squares discriminant analysis (PLS-DA) are often used to visually identify the metabolites with altered expression in plants following exposure to NPs. PCA and PLS-DA are generally conducted by SIMCA-P software and online resources (e.g., http://www.metaboanalyst.ca/) (Chen et al., 2019; Zhao et al., 2019). Unsupervised PCA analysis was conducted to integrally exhibit the clustering information of metabolites among groups and to further measure the component's contribution to variation in a data set. The 3

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

Fig. 2. Effects of graphene oxide (GO) and carboxylated single-walled carbon nanotubes (C-SWCNT) on the main metabolic pathways of algal cells at 96 h. The metabolic alterations that are labelled with blue and red arrows were determined based on comparisons of the experimental groups (GO and C-SWCNT) with the control. The directions of the arrows represent the upregulation or downregulation of the metabolites compared to the control. The black solid and dotted arrows represent direct and indirect reactions, respectively (Hu et al., 2015a). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.1. Two sides of carbohydrate and energy metabolism

discriminating metabolites in Fig. 1c (Zhao et al., 2017a). Subsequently, pathway enrichment analysis was used to identify the significantly affected biological pathways that contributed most significantly to the differentially expressed metabolites (Fig. 1d). For plant tissues, the pathway analysis was usually was conducted based on the KEGG database by MetaboAnalyst (Ke et al., 2018; Li et al., 2018a; Yuan et al., 2018). Except for the influence of biological pathways, it is also useful to consider a more global view of the impacts to the interconnected network of these metabolism pathways (Fig. 2). The network of metabolism pathways helps to provide a comprehensive vision of the metabolism that is impacted by ENMs (Zhao et al., 2018).

Carbohydrates are the primary energy source for plant life and are a significant component of plant biomass (Rossi et al., 2015; Thalmann and Santelia, 2017). In general, NPs downregulate carbohydrate metabolism and energy metabolism. ENMs (e.g., CdONPs and WS2NPs) reduce photosynthetic activities in plants and considerably downregulated the contents of the primary products of plant photosynthesis (i.e., glucose/fructose/sucrose in barley and starch in Chlorella vulgaris) (Vecerova et al., 2016; Yuan et al., 2018). By contrast, the inhibition of carbohydrate metabolism (e.g., starch and sucrose metabolism, and glyoxylate and dicarboxylate metabolism) indicates that energy expenditure exceeded the substance accumulation in plants, explaining the decrease in biomass (Wu et al., 2017). Some carbohydrates are important components of the cell wall; for example, a reduction in mannose is related to cell wall damage upon MWCNT exposure (Zhang et al., 2018b). Except for the downregulation of carbohydrates, ENMs also reduced energy metabolism in plants. CdONPs and Cu(OH)2 nanopesticides downregulated the TCA cycle (tricarboxylic acid cycle, e.g., citric acid, isocitric acid and fumaric acids) in barley roots and lettuce leaves (Vecerova et al., 2016; Zhao et al., 2016c). One reason for the reduction in citric acid could be the inhibition of malate dehydrogenase and/or citrate synthase enzymes by NPs (Vecerova et al., 2016). Moreover, the reduction of citric acid inhibited the dissolution of metal iron from metal ENMs and was a response of the plant in defence against metal ENMs (Vecerova et al., 2016; Zhao et al., 2016a). However, the regeneration of citric acid in the TCA cycle was interrupted, thus leading to a reduction in ATP (adenosine triphosphate) production and subsequently to the inhibition of plant growth (Vecerova et al., 2016; Zhang et al., 2018b). The above results indicated that the downregulation in carbohydrate and energy metabolism was responsible for the decrease of biomass and the defence stimulus in plants (Fig. 4). In contrast, some studies have proposed that ENMs upregulate

3. Exploring the phytotoxicology of ENMs by metabolomics To understand the phytotoxicity induced by ENMs, the uptake and transport of ENMs should first be understood (Fig. 3). The previous studies have demonstrated that ENMs can be taken up by plant roots and translocated to the stem/leaves/fruits. For example, CuNPs were taken up by the roots (cucumber) from the soils/nutrient solution and translocated to plant stem/leaves and even fruits (Huang et al., 2019; Zhao et al., 2016a, 2016b). MWCNTs (multiwalled carbon nanotubes) translocated to fruits from the roots of tomato seedlings (McGehee et al., 2017). By contrast, NPs can also be taken up by plant leaves and translocated to root tissues. Cu(OH)2 nanopesticide translocates to the root tissues from leaves through phloem in lettuce (Zhao et al., 2016c). CuONPs were transported from roots to shoots via xylem and then further translocated from the shoots back to the roots via phloem in maize (Wang et al., 2012). Metabolite perturbations in plants occur in response to stress, and the uptake/transport of ENMs in plant tissues perturbs the metabolites of plants. The present literature supports the hypothesis that ENMs primarily disturb carbohydrate metabolism, energy metabolism, amino acid metabolism, lipid metabolism and secondary metabolism, as listed in Table S1. 4

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

Fig. 3. The application of ENMs and their uptake/transport in plants. The implication of ENMs in agriculture and other fields leads to they contacting with plants, which mainly involves in root and leaf exposure. ENMs could be translocated by phloem/xylem between different tissues of plants.

Energy metabolism based on sugar decomposition (e.g., glycolysis, TCA cycle, and PPP) is a defence system and was also significantly upregulated by some NPs (e.g., Cu(OH)2 nanopesticide, TiO2NPs and AgNPs) (Ke et al., 2018; Wu et al., 2017; Zhao et al., 2016c). Energy metabolism was upregulated by ENMs, indicating that the plants had attempted to produce energy to maintain their normal physiological processes or increase defence-related activities (e.g., repair of DNA/ RNA/other biomolecules and manufacture of defence compounds) (Ke et al., 2018; Zhang et al., 2018a; Zhao et al., 2018). Oxidized-MWCNTs elevated the PPP in algal cells, as indicated by increased pathway intermediates such as ribulose 5-phosphate. The PPP can generate NADPH (nicotinamide adenine dinucleotide phosphate) and provides the raw materials for the synthesis of nucleotides and nucleic acids, and the generation of excess NADPH would assist in the defence against oxidative stress (Zhang et al., 2018b). Moreover, except for providing energy, the TCA cycle also provide raw materials for the synthesis of amino acids (Du et al., 2017). The elevated carbohydrate and energy metabolism was a plant defence mechanism against the stress of ENMs, and the elevated carbohydrates were also beneficial in the accumulation of biomass in plants (Fig. 4). 3.2. Alterations of amino acid metabolism in coping with stress of ENMs

Fig. 4. The defence and toxicity mechanisms of ENMs on plants by the regulation of carbohydrate and energy metabolism. Energy metabolism based on sugar decomposition is a defence system, which involves in glycolysis, TCA cycle, and PPP (pentose phosphate pathway). The blue and red arrows represent downregulation and upregulation of metabolites. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Amino acids are essential for protein biosynthesis and play crucial roles in signalling processes as well as in plant stress response (Hildebrandt et al., 2015). Exposure to ENMs could perturb amino acid metabolism in plants. The expression of amino acid related metabolism was upregulated in most studies and exhibited a stress defence mechanism. The five pathways of stress defence are shown in Fig. 5. (1) Amino acids are metal ion chelating agents. Metal ENMs (e.g., CuNPs, CdONPs and Cu(OH)2 nanopesticides) upregulated the expression of amino acid related metabolism (e.g., proline, glycine, asparagine, tryptophan, aspartic acid and histidine), and the amino acids acted as chelators of metal ions to decrease the bioavailability of metal ENMs in plant tissues (Vecerova et al., 2016; Zhao et al., 2016b, 2016c). Moreover, CuNPs upregulated the expression of amino acid related metabolism (e.g., alanine, β-alanine, glycine, leucine, lysine, phenylalanine, proline, serine, threonine and valine) in cucumber root exudate and then provided binding sites for copper and hindered its translocation across the root cell membrane (Zhao et al., 2016a). As a

carbohydrate metabolism and energy metabolism. Hydrated graphene ribbon upregulated carbohydrate metabolism (glucopyranoside, glucopyranose, glucose, mannose, and maltose), which have osmoprotective, antioxidant, and cell membrane phospholipid roles in wheat seed germination (Hu and Zhou, 2014). The Cu(OH)2 nanopesticide upregulated the primary products of photosynthesis (e.g., sucrose and glucose) in spinach leaves by foliar spray, indicating that spinach leaves attempt to fix more carbon in response to the Cu(OH)2 nanopesticide (Zhao et al., 2017b). These inconsistent conclusions probably depend on the tested ENMs, species and dosages (Hu et al., 2015a; Zhao et al., 2018). A standard proposal is a necessity for nanotoxicological tests. 5

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

Fig. 5. The toxicity mechanisms of ENMs and the defence of plants based on amino acid metabolism. Up-regulated and down-regulated expressions of amino acid-related metabolisms reflect the stress defence and toxicity mechanisms, respectively. The red arrows represent the upregulation of metabolites. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Nanocolloids in the natural water downregulated the expression of glycine, serine, threonine, alanine, aspartate and glutamate metabolism in Chlorella vulgaris and rice, and these processes were related to the inhibition of cell division (Ouyang et al., 2018; Hu et al., 2018). Moreover, the downregulated expression of amino acid (e.g., alanine, aspartate and glutamate) induced by nanocolloids could lead to a decrease in the osmotic pressure in the cytoplasm, which is associated with plasma membrane shrinkage (Ouyang et al., 2018; Kang et al., 2019a). The downregulated expression of amino acid related metabolism possibly indicated that excessive stress by ENMs induced the imbalance between ROS and the antioxidant defence system, leading to the impairment of the defence system.

multifunctional amino acid, proline also decreased the release of metal ions by reducing stress-induced cellular acidification (Wu et al., 2017). (2) Amino acids were involved in plant antioxidant defence. The generation of ROS is the primary toxic mechanism of NPs in plant. To protect plant cells from oxidative damage, two antioxidant defence systems (enzyme and non-enzyme) would be activated. Some amino acids are the precursors of antioxidant enzymes, and the increase of amino acids could promote the production of antioxidant enzymes to resist oxidative stress (Wu et al., 2017; Zhao et al., 2016b). For example, the increase of glycine, glutamine and proline promoted the biosynthesis of glutathione (GSH) and superoxide dismutase (SOD), which are essential antioxidants and vital for the maintenance of the intracellular redox state (Wu et al., 2017; Zhao et al., 2016b, 2016c). GSH also sequesters heavy metals (Zhao et al., 2016c). Some amino acids (e.g., glycine, proline and glutamine) are effective antioxidants for the scavenging of ROS (Wu et al., 2017; Zhao et al., 2016a, 2016b). (3) Amino acids served as energy sources. Cu(OH)2 nanopesticides and CuNPs increased branched-chain amino acids (BCAA, i.e., alanine, valine, and leucine) in maize leaves and cucumber fruits, and the proteins were oxidative phosphorylation energy sources that acted an adaptation process of the plants to stress (Taylor et al., 2004; Zhao et al., 2016b, 2018). (4) Amino acids were the precursors of secondary metabolites. Phenylalanine, tryptophan and tyrosine served as the precursors for the synthesis of protective secondary metabolites for plant defence, signalling and reproduction (e.g., flavonoids, phenolics, isoflavonoids, anthocyanins, tocopherols, and plastoquinone) and were upregulated by CdONPs in barley (Vecerova et al., 2016; Yoo et al., 2013). (5) Amino acids protected the cell membrane. Phenylalanine and tryptophan posed protective effects on membranes and reduced lipid peroxidation (tryptophan) to enhance tolerance to cadmium (Cd) and were upregulated by CdONPs in barley (Vecerova et al., 2016). These above results indicated that increased expression of amino acid related metabolism was also an adaptation process against the toxicity of NPs in diversified processes. Interestingly, some studies also demonstrated that expression of amino acid related metabolism could also be downregulated in plants by ENMs (Zhang et al., 2018a). As indicated, AgNPs dramatically decreased glutamine and asparagine, indicating that the process of inorganic nitrogen fixation may be disturbed; besides, the glycine/serine ratio was also decreased by AgNPs, indicating that the photorespiratory activity was inhibited and leaves entered senescence after exposure to AgNPs. Moreover, GO downregulated the expression of amino acid related metabolism by increasing the content of ROS, which induced growth inhibition of the chlorella vulgaris (Hu et al., 2014a).

3.3. Lipid metabolism reflects changes of cell membrane under the exposure of ENMs Fatty acids are critical components of cellular membranes, and the alteration of fatty acids may affect cell survival under stress conditions (Zhao et al., 2018). The regulation of lipid metabolism is related to the following two functions. (1) The upregulation of fatty acids protected the cell membrane. TiO2NPs, Cu(OH)2 nanopesticides and hydrated graphene ribbon elevated the content of fatty acids (e.g., linoleic acid, linolenic acid, stearic acid, palmitic acid, hexadecanoic acid and octadecadienoic acid) and the precursors (e.g., 1-monostearin and 1monopalmitin) for membrane lipids in rice, maize and wheat, protecting the cell membrane under abiotic stress (membrane lipid peroxidation) (Hu and Zhou, 2014; Wu et al., 2017; Zhao et al., 2018). Correspondingly, fatty acids were decreased by C60 fullerols in cucumber leaves, and the composition of the cell membrane was perturbed (Zhao et al., 2019). (2) Membrane fluidity is positively associated with the level of unsaturated fatty acids and is critical to maintain cellular metabolism and function (Hoekstra et al., 2001; Zhao et al., 2018). The deterioration of membrane fluidity and osmotic stress induced by GO and SWCNTs is associated with the decrease in unsaturated fatty acids in Chlorella vulgaris (Hu et al., 2015a). Reduced membrane fluidity was also considered as a response to prevent the transport of NPs to the cytosol (Vecerova et al., 2016). Interestingly, oxidized MWCNTs elevated the levels of unsaturated fatty acids and decreased the levels of saturated fatty acids in Chlorella pyrenoidosa, promoting membrane fluidity (Zhang et al., 2018b). However, the increase in membrane fluidity was linked to cell plasmolysis and a destabilized and disrupted membrane structure.

6

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

3.4. Correlation between secondary metabolism and oxidative damage induced by ENMs

4. Integrating metabolomics and other omics to analyse phytotoxicity

Secondary metabolism is a non-enzymatic antioxidant defence system to resist oxidative damage in plant (Kasote et al., 2015; Tattini et al., 2015). Numerous studies have demonstrated that secondary metabolites are upregulated in plants under the stress of ENMs. TiO2NPs, CuNPs and AgNPs significantly upregulated secondary metabolites (e.g., ascorbic acid, tocopherol, salicylic acid, benzoic acid, 4hydroxybenzoate, phytosphingosine, nicotinamide, itaconic acid, 2furoic acid, 3-cyanoalanine and phytol) in rice and cucumber to combat oxidative stress (Wu et al., 2017; Zhang et al., 2018a; Zhao et al., 2016a). Correspondingly, Cu(OH)2 nanopesticides and CeO2NPs significantly impaired the antioxidant system by downregulating a variety of secondary metabolites (e.g., phenolic compounds, carotenoids and ascorbic acid) in lettuce leaves, spinach leaves and bean leaves (Salehi et al., 2018; Zhao et al., 2016c, 2017b).

‘Omics’ technologies are primarily involved in genomics (genes), transcriptomics (mRNA), proteomics (proteins) and metabolomics (metabolites) and have become powerful tools to detect the holistic plant response to environmental stresses (Feussner and Polle, 2015; Kumar et al., 2015; Romero et al., 2006). Metabolites are the downstream products of gene and protein expression, and multi-omics approaches could provide a comprehensive and multilevel perspective to understand plant biochemical responses to ENMs, compared to a metabolomics analysis alone (Kang et al., 2019b; Zhao et al., 2019). Genes and proteins possess more specific functional descriptions in the present database than metabolites. The combination of metabolomics (what has changed in the physiological state of plants) with proteomics (What is happening in the physiological state of plants) or metabolomics with genomics/transcriptomics (What might happen to the physiological state of the plants) could globally explain the effect mechanism (Kang et al., 2019b; Li et al., 2018a; Salehi et al., 2018; Zhao et al., 2019). For example, the metabolomics results in Fig. 7 revealed that C60 fullerols upregulated antioxidant metabolites (e.g., methyl trans-cinnamate, 3-hydroxyflavone, and 1,2,4-benzenetriol) and downregulated cell membrane metabolites (palmitoleic and linolenic acid). Moreover, the proteomics analysis revealed that C60 fullerols upregulated the chloroplast proteins that are involved in water photolysis (PSII protein), light-harvesting (CAB), ATP production (ATP synthase), pigment fixation (Mg-PPIX), and electron transport (Cyt b6f). The integration of metabolomics and proteomics suggested that C60 fullerols significantly accelerated the electron transport rate and further induced ROS overproduction in chloroplast thylakoids. To resist ROS damage, plants upregulated antioxidant molecules and defence-related proteins. The overproduction of ROS also perturbed carbon metabolism. The multi-omics databases could support and complement each other to completely explain the response of plants to ENMs. The one problem of a multi-omics approach is that the regulation trends between metabolomics and other omics processes did not exactly match. Once the trends of regulation by multi-omics approaches are inconsistent, the metabolites and DNA/mRNA/proteins need to be further verified (e.g., by mass spectrometry for metabolites, quantitative realtime PCR for gene analysis, or Western blot for proteins).

3.5. Integrating the metabolism processes to comprehensively explain the cellular stress response induced by ENMs The aggregation of ENMs on the plant cell surfaces can partition/ penetrate and disrupt the cell wall/membrane, inducing direct physical damage (Deng et al., 2017; Long et al., 2012; Wang et al., 2015). To resist damage, fatty acids and some amino acids were upregulated, thus protecting the cell membrane. Moreover, unsaturated fatty acids were downregulated to prevent the transport of ENMs to the cytosol. Once the excess stress overwhelmed the above protective responses, damaging responses could occur. For example, unsaturated fatty acids were downregulated and induced an increase in membrane fluidity that was responsible for cell plasmolysis and the disruption of the membrane structure. Moreover, the aggregation of ENMs on the surface of phototrophic organisms could explain the reduction of photosynthesis and inhibit plant growth, namely, by a shading effect (Deng et al., 2017; Li et al., 2015; Long et al., 2012; Middepogu et al., 2018). The reduction of photosynthesis was reflected in the downregulation of the content of its primary products (e.g., glucose/fructose/sucrose). For metal-based ENMs, free metals are also the reason for phytotoxicity in plant cells. To decrease the free metal content in plant tissues, some amino acids were upregulated to act as chelators of metal ions, and citric acid was downregulated to lower the dissolution of metal iron from metal ENMs. Oxidative stress due to the overproduction of ROS has been recognized as a primary influencing factor of ENM toxicity (Middepogu et al., 2018; Zhang et al., 2018b). To address the excessive oxidative stress by ENMs, antioxidant defence mechanisms were activated. From the metabolomic perspective, the antioxidant defence mechanisms include: 1) upregulated expression of amino acid related metabolism, which promotes the production of antioxidant enzymes and acts as antioxidants and precursors of secondary metabolites; 2) upregulated membrane lipids to protect the cell membrane from lipid peroxidation; and 3) upregulated secondary metabolism to resist oxidative damage in the plant. Moreover, additional protective responses could also be activated by exposure to ENMs. These responses include upregulated carbohydrate metabolism to fix more carbon and upregulated energy related metabolism (e.g., TCA cycle, PPP and glycolysis) to produce enough energy to resist the stress. Once these protective responses were overwhelmed, damaging responses could occur (Wang et al., 2015; Zhang et al., 2018b). The damaging responses (e.g., decline of biomass, lipid peroxidation, cell plasmolysis, and the disruption of membrane structure) were reflected in the downregulation of carbohydrate/energy/ amino acid/secondary metabolism and the upregulation of unsaturated fatty acids. The above damaging and protective responses are shown in Fig. 6.

5. Factors affecting metabolic regulation The regulation of metabolism was affected by multiple factors, such as plant species/life cycle, nanomaterial properties, exposure concentration/method/duration and environmental modification. Cu(OH)2 nanopesticides differentially regulated metabolic pathways in cucumber and corn (upregulated energy-related and antioxidant defence related pathways in maize and upregulated N metabolism in cucumber) (Zhao et al., 2018). Moreover, plants in different stages also exhibited different metabolic perturbations when exposed to the same ENMs (Lin and Xing, 2007; Mamyandi et al., 2012). The nanoparticle properties that were vital factors that regulated the metabolites in plants included type of NPs (e.g., carbon-based and metal-based nanomaterials), size, oxidation and the phases of ENMs. Compared to carbon-based ENMs, metal-based nanomaterials were involved in the release of metal ions, and the metal ions could affect the metabolism of plants (Hu et al., 2015a; Ke et al., 2018; Sharma and Dietz, 2009). GOQD and rGO posed more obvious disturbances in biological metabolism than GO (Li et al., 2018b; Ouyang et al., 2015). The phases of ENMs also influence the plant metabolism. Yuan et al. (2018) revealed that Ce-WS2 (chemically exfoliated WS2, mainly in the 1T phase) induced more obvious alterations in metabolites (e.g., amino acids and fatty acids) and metabolic pathways (e.g., starch and sucrose metabolism) than Ae-WS2 (annealed exfoliated WS2, 2H phase). These alterations correlated with cell membrane damage, oxidative stress and inhibition of photosynthesis. It 7

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

Fig. 6. Metabolomics reveals the protective and damaging response in plant cells when they are exposed to ENMs. The black oval box represents a plant cell. The green and red arrows represent the downregulation and upregulation of metabolism. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7. Summary scheme showing the main changes induced by C60 fullerol in cucumber leaves, as detected at the metabolite and protein levels (Zhao et al., 2019). The proteins and metabolites with red and green italics represented its upregulated and downregulated differential expression, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

of metabolites (Salehi et al., 2018). Moreover, the exposure duration also affected the plant response to ENMs (Tripathi and Sarkar, 2014; Wang et al., 2014). The above results will help us choose a rational application scope and dosage of ENMs. Environmental modifications also pose crucial influence on plant metabolism, such as natural organic matter, hydration, irradiation and biological secretions. Humic acid (HA) significantly relieved the changes in the metabolic profiles caused by exposure to single-layer molybdenum disulphide (SLMoS2) and GO (Hu et al., 2014b; Zou et al., 2018). Moreover, long-time hydration and irradiation reduced the nanotoxicity of graphene to algal cells by reducing the generation of reactive oxygen species, diminishing protein carbonylation and decreasing tail DNA (Hu et al., 2015b). CuNPs that are modified by biological secretions significantly decrease Cu uptake and bioaccumulation in plants (Huang et al., 2017). The metabolic profiles that are

is difficult to uniformly identify the key influencing factor, although establishing the relationship between phytotoxicity and nanoparticle properties will help the design of environmentally friendly ENMs. For the phytotoxicity of ENMs, the exposure concentration/method/ duration to plants also play an important role. For the concentrations of ENMs, Zhao et al. (2018) indicated that low doses of Cu(OH)2 nanopesticides may accelerate the synthesis of aromatic amino acids (precursors for a wide range of secondary metabolites), but higher doses may inhibit the shikimate-phenylpropanoid pathway, indicating a change in the balance between ROS and defence systems. Different exposure methods also affected the regulation of metabolism in plants (Salehi et al., 2018; Vecerova et al., 2016). For CeO2NPs, spraying had a more marked effect towards bean leaves than soil application (e.g., higher Ce content, relative water content and electrolyte leakage, and induction of oxidative stress), which was consistent with the regulation

8

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

References

changed by the environmentally transformed ENMs are more meaningful than those by pristine ENMs.

Antonacci, A., et al., 2018. Nanostructured (Bio)sensors for smart agriculture. TrAC Trends Anal. Chem. 98, 95–103. Baker, M., 2011. Metabolomics: from small molecules to big ideas. Nat. Methods 8, 117–121. Beyoglu, D., Idle, J.R., 2013. Metabolomics and its potential in drug development. Biochem. Pharmacol. 85, 12–20. Chen, X., et al., 2019. Nanoparticle TiO2 size and rutile content impact bioconcentration and biomagnification from algae to daphnia. Environ. Pollut. 247, 421–430. Chitarrini, G., et al., 2017. Identification of biomarkers for defense response to Plasmopara viticola in a resistant grape variety. Front. Plant Sci. 8, 1524. Cristina Abreu, A., et al., 2018. NMR-based metabolomics approach to study the influence of different conditions of water irrigation and greenhouse ventilation on zucchini crops. J. Agric. Food Chem. 66, 8422–8432. De Vos, R.C., et al., 2007. Untargeted large-scale plant metabolomics using liquid chromatography coupled to mass spectrometry. Nat. Protoc. 2, 778–791. Deng, X.Y., et al., 2017. Biological effects of TiO2 and CeO2 nanoparticles on the growth, photosynthetic activity, and cellular components of a marine diatom Phaeodactylum tricornutum. Sci. Total Environ. 575, 87–96. Du, C., et al., 2017. Biological effect of aqueous C60 aggregates on Scenedesmus obliquus revealed by transcriptomics and non-targeted metabolomics. J. Hazard Mater. 324, 221–229. Feussner, I., Polle, A., 2015. What the transcriptome does not tell - proteomics and metabolomics are closer to the plants' patho-phenotype. Curr. Opin. Plant Biol. 26, 26–31. Fuhrer, T., Zamboni, N., 2015. High-throughput discovery metabolomics. Curr. Opin. Biotechnol. 31, 73–78. Gajbhiye, S., Sakharwade, S., 2016. Silver nanoparticles in cosmetics. J. Cosmet. Dermatol. Sci. Appl. 06, 48–53. Garcia, C.J., et al., 2016. Untargeted metabolomics approach using UPLC-ESI-QTOF-MS to explore the metabolome of fresh-cut iceberg lettuce. Metabolomics 12, 138. Ghorbanpour, M., Hadian, J., 2015. Multi-walled carbon nanotubes stimulate callus induction, secondary metabolites biosynthesis and antioxidant capacity in medicinal plant Satureja khuzestanica grown in vitro. Carbon 94, 749–759. Gu, R., et al., 2019. Metabolic profiling of different parts of Acer truncatum from the Mongolian plateau using UPLC-QTOF-MS with comparative bioactivity assays. J. Agric. Food Chem. 67, 1585–1597. Gunsolus, I.L., Haynes, C.L., 2016. Analytical aspects of nanotoxicology. Anal. Chem. 88, 451–479. Hildebrandt, T.M., et al., 2015. Amino acid catabolism in plants. Mol. Plant 8, 1563–1579. Hincapie, I., et al., 2015. Use of engineered nanomaterials in the construction industry with specific emphasis on paints and their flows in construction and demolition waste in Switzerland. Waste Manag. 43, 398–406. Hoekstra, F.A., et al., 2001. Mechanisms of plant desiccation tolerance. Trends Plant Sci. 6, 431–438. Hong, J., et al., 2016. Plant metabolomics: an indispensable system biology tool for plant science. Int. J. Mol. Sci. 17, 767. Hu, X., et al., 2014a. Interactions between graphene oxide and plant cells: regulation of cell morphology, uptake, organelle damage, oxidative effects and metabolic disorders. Carbon 80, 665–676. Hu, X., et al., 2014b. Humic acid acts as a natural antidote of graphene by regulating nanomaterial translocation and metabolic fluxes in vivo. Environ. Sci. Technol. 48, 6919–6927. Hu, X., et al., 2015a. Effects of graphene oxide and oxidized carbon nanotubes on the cellular division, microstructure, uptake, oxidative stress, and metabolic profiles. Environ. Sci. Technol. 49, 10825–10833. Hu, X., et al., 2015b. Ambient water and visible-light irradiation drive changes in graphene morphology, structure, surface chemistry, aggregation, and toxicity. Environ. Sci. Technol. 49, 3410–3418. Hu, X., et al., 2018. Characterization and toxicity of nanoscale fragments in wastewater treatment plant effluent. Sci. Total Environ. 626, 1332–1341. Hu, X., Zhou, Q., 2014. Novel hydrated graphene ribbon unexpectedly promotes aged seed germination and root differentiation. Sci. Rep. 4, 3782. Huang, Y., et al., 2019. Antioxidant response of cucumber (Cucumis sativus) exposed to nano copper pesticide: quantitative determination via LC-MS/MS. Food Chem. 270, 47–52. Huang, Y., et al., 2017. Interactions, transformations, and bioavailability of nano-copper exposed to root exudates. Environ. Sci. Technol. 51, 9774–9783. Ji, H.-G., et al., 2018. Diverse metabolite variations in tea (Camellia sinensis L.) leaves grown under various shade conditions revisited: a metabolomics study. J. Agric. Food Chem. 66, 1889–1897. Kang, W., et al., 2019a. Nanoscale colloids induce metabolic disturbance of zebrafish at environmentally relevant concentrations. Environ. Sci. Nano 6, 1562–1575. Kang, W., et al., 2019b. Study of the persistence of the phytotoxicity induced by graphene oxide quantum dots and of the specific molecular mechanisms by integrating omics and regular analyses. Environ. Sci. Technol. 53, 3791–3801. Kasote, D.M., et al., 2015. Significance of antioxidant potential of plants and its relevance to therapeutic applications. Int. J. Biol. Sci. 11, 982–991. Ke, M., et al., 2018. Phytotoxic effects of silver nanoparticles and silver ions to Arabidopsis thaliana as revealed by analysis of molecular responses and of metabolic pathways. Sci. Total Environ. 644, 1070–1079. Kim, H.K., et al., 2010. NMR-based metabolomic analysis of plants. Nat. Protoc. 5, 536–549.

6. Perspectives and challenges Plants are the primary producers in ecological systems and are essential to the activity of human life. The wide application of ENMs has attracted much attention to the phytotoxicity of ENMs. Metabolites are typical and sensitive biomarkers. Metabolomics could provide a comprehensive assessment of the adverse effects that are induced by ENMs at the molecular level, compared to single traditional ecotoxicological end points. This review summarized the damaging and protective response from the metabolite perspective (carbohydrate, energy, amino acid, lipid and secondary metabolism) in plants that were exposed to ENMs. There are some gaps that should be highlighted in the future works. Metabolomics that are measured by a single type of instrument (GC-MS, LC-MS or NMR) would miss some metabolites, owing to the bias of any single type of instrument. Multiple instrument detection on the same samples makes the metabolites more comprehensive. Furthermore, ENMs affect various metabolic pathways; however, it is hard to estimate their key influences on metabolic pathways and the direct network of relationships among the different pathways. The key influencing pathways should be identified by controlling (i.e., upregulating or inhibiting) some influencing pathways, but the related studies are lacking. Moreover, the positive control in the present metabolomics analysis does not attract much attention. In addition, studies on the threshold of metabolite influence should be conducted to identify the levels of recoverable metabolite disturbance. The metabolic profile of plants to stress is very sensitive. To improve the repeatability of sample analysis and avoid the occurrence of false negatives or false positives, strict control of experimental conditions, internal references, and a large number of parallels should be conducted. More attention should be paid to the metabolomics analysis from in vivo, in situ and spatially imaging studies. Compared to gene and protein databases, the metabolite database is relatively scarce, and the annotation of metabolite function needs to be further developed. For a prospective study of nanotoxicity, it is essential to excavate the relationship between metabolites and a biological response instead of repeating the work of traditional ecotoxicological end points. Employing metabolomics to study the phytotoxicity of nanomaterials, the multiple factors (e.g., plant species/life cycle, nanomaterial properties, exposure concentration/ method/duration and environmental modification) should be considered. Metabolite analysis also contributes to designing and applying the environmentally friendly ENMs, although the related information is rare. A recent work that employed metabolomics to screen the biocompatible surface modifications for NPs was reported (Sun et al., 2018). Expanding the application of metabolomics to control nanotoxicity will be attractive in the future. Notes The authors declare no competing financial interests. Acknowledgments This work was financially supported by the National Natural Science Foundation of China (grant nos. 21876092, 21722703, 31770550 and 21577070), the Central Public Research Institutes Basic Funds for Research and Development (Agro-Environmental Protection Institute, Ministry of Agriculture), the Natural Science Foundation of Tianjin City (grant no. 18JCYBJC23600) and a 111 program (grant no. T2017002). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.ecoenv.2019.109602. 9

Ecotoxicology and Environmental Safety 184 (2019) 109602

X. Li, et al.

Mediterranean summers. New Phytol. 207, 613–626. Taylor, N.L., et al., 2004. Lipoic acid-dependent oxidative catabolism of alpha-keto acids in mitochondria provides evidence for branched-chain amino acid catabolism in Arabidopsis. Plant Physiol. 134, 838–848. Thalmann, M., Santelia, D., 2017. Starch as a determinant of plant fitness under abiotic stress. New Phytol. 214, 943–951. Tripathi, D.K., et al., 2017. An overview on manufactured nanoparticles in plants: uptake, translocation, accumulation and phytotoxicity. Plant Physiol. Biochem. 110, 2–12. Tripathi, S., Sarkar, S., 2014. Influence of water soluble carbon dots on the growth of wheat plant. Appl. Nanosci. 5, 609–616. Tsugawa, H., et al., 2014. Highly sensitive and selective analysis of widely targeted metabolomics using gas chromatography/triple-quadrupole mass spectrometry. J. Biosci. Bioeng. 117, 122–128. Vecerova, K., et al., 2016. Changes of primary and secondary metabolites in barley plants exposed to CdO nanoparticles. Environ. Pollut. 218, 207–218. Vinaixa, M., et al., 2016. Mass spectral databases for LC/MS- and GC/MS-based metabolomics: state of the field and future prospects. TrAC Trends Anal. Chem. (Reference Ed.) 78, 23–35. Wang, P., et al., 2016. Nanotechnology: a new opportunity in plant sciences. Trends Plant Sci. 21, 699–712. Wang, S., et al., 2015. The effect of CuO NPs on reactive oxygen species and cell cycle gene expression in roots of rice. Environ. Toxicol. Chem. 34, 554–561. Wang, S., et al., 2014. Foliar application with nano-silicon alleviates Cd toxicity in rice seedlings. Environ. Sci. Pollut. Res. 22, 2837–2845. Wang, Y., et al., 2017. Agglomeration determines effects of carbonaceous nanomaterials on soybean nodulation, dinitrogen fixation potential, and growth in soil. ACS Nano 11, 5753–5765. Wang, Z., et al., 2012. Xylem- and phloem-based transport of CuO nanoparticles in maize (Zea mays L.). Environ. Sci. Technol. 46, 4434–4441. Wu, B., et al., 2017. Metabolomics analysis of TiO2 nanoparticles induced toxicological effects on rice (Oryza sativa L.). Environ. Pollut. 230, 302–310. Yamamoto, K., et al., 2016. Cell-specific localization of alkaloids in Catharanthus roseus stem tissue measured with imaging MS and single-cell MS. Proc. Natl. Acad. Sci. U.S.A. 113, 3891–3896. Yin, J., et al., 2018. Opportunities to advance sustainable design of nano-enabled agriculture identified through a literature review. Environ. Sci. : Nano 5, 11–26. Yoo, H., et al., 2013. An alternative pathway contributes to phenylalanine biosynthesis in plants via a cytosolic tyrosine: phenylpyruvate aminotransferase. Nat. Commun. 4, 2833. Yuan, P., et al., 2018. The phases of WS2 nanosheets influence uptake, oxidative stress, lipid peroxidation, membrane damage, and metabolism in algae. Environ. Sci. Technol. 52, 13543–13552. Zhang, H., et al., 2018a. Metabolomics reveals how cucumber (Cucumis sativus) reprograms metabolites to cope with silver ions and silver nanoparticle-induced oxidative stress. Environ. Sci. Technol. 52, 8016–8026. Zhang, H., et al., 2019. Metabolomics reveals the “invisible” responses of spinach plants exposed to CeO2 nanoparticles. Environ. Sci. Technol. 53, 6007–6017. Zhang, L., et al., 2018b. Cellular response of Chlorella pyrenoidosa to oxidized multiwalled carbon nanotubes. Environ. Sci. : Nano 5, 2415–2425. Zhang, P., et al., 2017a. Phytotoxicity, uptake and transformation of nano-CeO2 in sand cultured romaine lettuce. Environ. Pollut. 220, 1400–1408. Zhang, X., et al., 2017b. Molecular mechanisms of developmental toxicity induced by graphene oxide at predicted environmental concentrations. Environ. Sci. Technol. 51, 7861–7871. Zhao, L., et al., 2017a. 1H NMR and GC-MS based metabolomics reveal nano-Cu altered cucumber (Cucumis sativus) fruit nutritional supply. Plant Physiol. Biochem. 110, 138–146. Zhao, L., et al., 2017b. Metabolomics reveals Cu(OH)2 nanopesticide-activated anti-oxidative pathways and decreased beneficial antioxidants in spinach leaves. Environ. Sci. Technol. 51, 10184–10194. Zhao, L., et al., 2016a. 1H NMR and GC-MS based metabolomics reveal defense and detoxification mechanism of cucumber plant under nano-Cu stress. Environ. Sci. Technol. 50, 2000–2010. Zhao, L., et al., 2018. Comparative metabolic response between cucumber (Cucumis sativus) and corn (Zea mays) to a Cu(OH)2 nanopesticide. J. Agric. Food Chem. 66, 6628–6636. Zhao, L., et al., 2016b. GC-TOF-MS based metabolomics and ICP-MS based metallomics of cucumber (Cucumis sativus) fruits reveal alteration of metabolites profile and biological pathway disruption induced by nano copper. Environ. Sci. : Nano 3, 1114–1123. Zhao, L., et al., 2016c. Metabolomics to detect response of lettuce (Lactuca sativa) to Cu (OH)2 nanopesticides: oxidative stress response and detoxification mechanisms. Environ. Sci. Technol. 50, 9697–9707. Zhao, L., et al., 2019. C60 Fullerols enhance copper toxicity and alter the leaf metabolite and protein profile in cucumber. Environ. Sci. Technol. 53, 2171–2180. Zhou, Q., Hu, X., 2017. Systemic stress and recovery patterns of rice roots in response to graphene oxide nanosheets. Environ. Sci. Technol. 51, 2022–2030. Zou, W., et al., 2018. Environmental transformations and algal toxicity of single-layer molybdenum disulfide regulated by humic acid. Environ. Sci. Technol. 52, 2638–2648. Zou, W., et al., 2019. Dissolved oxygen and visible light irradiation drive the structural alterations and phytotoxicity mitigation of single-layer molybdenum disulfide. Environ. Sci. Technol. 53, 7759–7769.

Kite, G.C., 2018. Comment on METLIN: a technology platform for identifying knowns and unknowns. Anal. Chem. 90, 13126–13127. Kumar, S., et al., 2015. Omics and biotechnology of arsenic stress and detoxification in plants: current updates and prospective. Environ. Int. 74, 221–230. Li, F., et al., 2015. Toxicity of nano-TiO2 on algae and the site of reactive oxygen species production. Aquat. Toxicol. 158, 1–13. Li, X., et al., 2018a. Integrating proteomics, metabolomics and typical analysis to investigate the uptake and oxidative stress of graphene oxide and polycyclic aromatic hydrocarbons. Environ. Sci. : Nano 5, 115–129. Li, X., et al., 2018b. Effects of the size and oxidation of graphene oxide on crop quality and specific molecular pathways. Carbon 140, 352–361. Lin, D., Xing, B., 2007. Phytotoxicity of nanoparticles: inhibition of seed germination and root growth. Environ. Pollut. 150, 243–250. Long, Z., et al., 2012. Systematic and quantitative investigation of the mechanism of carbon nanotubes' toxicity toward algae. Environ. Sci. Technol. 46, 8458–8466. Lu, P.-J., et al., 2015. Analysis of titanium dioxide and zinc oxide nanoparticles in cosmetics. J. Food Drug Anal. 23, 587–594. Maag, D., et al., 2015. Metabolomics in plant-herbivore interactions: challenges and applications. Entomol. Exp. Appl. 157, 18–29. Majumdar, S., et al., 2016. Cerium biomagnification in a terrestrial food chain: influence of particle size and growth stage. Environ. Sci. Technol. 50, 6782–6792. Mamyandi, M.M., et al., 2012. Effect of nano-iron spraying at varying growth stage of sugar beet (Beta vulgaris L.) on the size of different plant parts. Intl. J. Agric. Crop Sci. 4, 740–745. Markley, J.L., et al., 2017. The future of NMR-based metabolomics. Curr. Opin. Biotechnol. 43, 34–40. McGehee, D.L., et al., 2017. Multiwalled carbon nanotubes dramatically affect the fruit metabolome of exposed tomato plants. ACS Appl. Mater. Interfaces 9, 32430–32435. Middepogu, A., et al., 2018. Effect and mechanism of TiO2 nanoparticles on the photosynthesis of Chlorella pyrenoidosa. Ecotoxicol. Environ. Saf. 161, 497–506. Monreal, C.M., et al., 2015. Nanotechnologies for increasing the crop use efficiency of fertilizer-micronutrients. Biol. Fertil. Soils 52, 423–437. Negrin, A., et al., 2019. LC-MS metabolomics and chemotaxonomy of caffeine-containing holly (Ilex) species and related taxa in the Aquifoliaceae. J. Agric. Food Chem. 67, 5687–5699. Nel, A., et al., 2012. Nanomaterial toxicity testing in the 21st century: use of a predictive toxicological approach and high-throughput screening. Acc. Chem. Res. 46, 607–621. Ouyang, S., et al., 2015. Envelopment-internalization synergistic effects and metabolic mechanisms of graphene oxide on single-cell Chlorella vulgaris are dependent on the nanomaterial particle size. ACS Appl. Mater. Interfaces 7, 18104–18112. Ouyang, S., et al., 2018. Nanocolloids in natural water: isolation, characterization, and toxicity. Environ. Sci. Technol. 52, 4850–4860. Pagano, L., et al., 2016. Molecular response of crop plants to engineered nanomaterials. Environ. Sci. Technol. 50, 7198–7207. Peters, R.J.B., et al., 2016. Nanomaterials for products and application in agriculture, feed and food. Trends Food Sci. Technol. 54, 155–164. Pontes, J.G.M., et al., 2017. NMR-based metabolomics strategies: plants, animals and humans. Anal. Methods 9, 1078–1096. Pullagurala, V.L.R., et al., 2018. ZnO nanoparticles increase photosynthetic pigments and decrease lipid peroxidation in soil grown cilantro (Coriandrum sativum). Plant Physiol. Biochem. 132, 120–127. Risticevic, S., et al., 2016. Capturing plant metabolome with direct-immersion in vivo solid phase microextraction of plant tissues. Anal. Chem. 88, 1266–1274. Rizwan, M., et al., 2017. Effect of metal and metal oxide nanoparticles on growth and physiology of globally important food crops: a critical review. J. Hazard Mater. 322, 2–16. Romero, R., et al., 2006. The use of high-dimensional biology (genomics, transcriptomics, proteomics, and metabolomics) to understand the preterm parturition syndrome. BJOG 113 (3), 118–135. Rossi, M., et al., 2015. Crop yield: challenges from a metabolic perspective. Curr. Opin. Plant Biol. 25, 79–89. Salehi, H., et al., 2018. Morphological, proteomic and metabolomic insight into the effect of cerium dioxide nanoparticles to Phaseolus vulgaris L. under soil or foliar application. Sci. Total Environ. 616–617, 1540–1551. Sgorbini, B., et al., 2019. Evaluation of volatile bioactive secondary metabolites transfer from medicinal and aromatic plants to herbal teas: comparison of different methods for the determination of transfer rate and human intake. J. Chromatogr. A 1594, 173–180. Sharma, S.S., Dietz, K.J., 2009. The relationship between metal toxicity and cellular redox imbalance. Trends Plant Sci. 14, 43–50. Stopka, S.A., et al., 2019. Ambient metabolic profiling and imaging of biological samples with ultrahigh molecular resolution using laser ablation electrospray ionization 21 Tesla FTICR mass spectrometry. Anal. Chem. 91, 5028–5035. Sun, A., et al., 2018. Screening small metabolites from cells as multifunctional coatings simultaneously improves nanomaterial biocompatibility and functionality. Adv. Sci. 5, 1800341. Sun, T.Y., et al., 2017. Envisioning nano release dynamics in a changing world: using dynamic probabilistic modeling to assess future environmental emissions of engineered nanomaterials. Environ. Sci. Technol. 51, 2854–2863. Tamura, Y., et al., 2018. Metabolomic evaluation of the quality of leaf lettuce grown in practical plant factory to capture metabolite signature. Front. Plant Sci. 9, 665. Tattini, M., et al., 2015. Isoprenoids and phenylpropanoids are part of the antioxidant defense orchestrated daily by drought-stressedPlatanus × acerifoliaplants during

10