Plants have the remarkable capability of responding to environmental cues thanks to different signal transduction cascades, which are triggered by signaling molecules that perceive and expand external and internal signals, resulting in plant adaptation reactions (Arisz et al., 2009; Sparks et al., 2013; Abd-El-Haliem et al., 2016). Some of the most important signaling molecules in plants involve lipids, including phosphoinositides (PPI), sphingolipids, lysophospholipids, oxylipins, N-acylethanolamines, and free fatty acids (Wang and Chapman, 2013). Though most of the leaf structural lipids in plant cells are galactolipids (approximately 70%), phospholipids, including phosphatidic acid (PA), play important roles in signal perception and transduction (Testerink and Munnik, 2005).

PA has been proposed to be a pivotal second messenger in plants and its synthesis has been reported to be induced in response to ethylene (Munnik, 2001), abscisic acid (Zhang et al., 2004), wounding and Nod factor (Munnik, 2001), osmotic pressure (Munnik et al., 2000; Testerink et al., 2004), cold (Ruelland et al., 2002), salinity (Zhang et al., 2012), temperature changes (Arisz et al., 2013), pathogen attack (Zhang and Xiao, 2015), and drought (Li et al., 2015). In all of those cases, PA synthesis has been associated with plant cell adjustments to overcome such stress events. Nevertheless, some positive-strand RNA viruses use PA in order to stimulate their replication (Hyodo et al., 2015), which has to be taken into consideration when designing strategies to apply PA as a potential biostimulator of adaptive responses in crop plants. Importantly, PA is a precursor to all phosphoglycerolipids as well as triacylglycerols and galactolipids, and its turnover is crucial in determining lipid metabolic fluxes and membrane compositions (Arisz et al., 2009, 2013).

In plant cells, PA can be generated in the plasma membrane from two different metabolic pathways: (1) as a product of the hydrolysis of structural phospholipids, such as phosphatidylcholine and phosphatidylethanolamine by the action of different isoforms of phospholipase D (PLD; Hong et al., 2014); or (2) through the combined activity of phospholipase C (PLC) and diacylglycerol kinase (DGK). There are two types of PLCs in plant cells: those that take PPI as substrate, the so called phosphatidylinositol (PI)-PLCs, and those that hydrolyze structural phospholipids, known as non-specific PLCs (NPCs). In any case, PLCs yield diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP₃). After phosphatidylinositol 4,5-bisphosphate hydrolysis, IP₃ diffuses into the cytosol and acts as a second messenger implicated in calcium (Ca²⁺) mobilization from intracellular compartments such as the vacuole, while DAG remains in the plasma membrane and may activate other physiological processes. DAG can then be phosphorylated by DGK to yield PA, which in turn can be further metabolized to PPI. Additionally, enzymes that dephosphorylate PA include lipid phosphate phosphatases and PA hydrolases (lipins). Moreover, PA can be further phosphorylated to diacylglycerol pyrophosphate by PA kinase, or metabolized to lyso-PA through PLA₂ activity (Testerink and Munnik, 2011). Most of these lipids and enzymes have been proved to have signaling functions, and the synthesis of signaling lipids in response to different environmental cues and stressors is essentially transient (Testerink and Munnik, 2011; Ge et al., 2012).

In eukaryotic cells, PA levels are typically low, corresponding to 0.67% of total phospholipids (Arisz et al., 2000) and PA formation depends on the extracellular stimuli perceived.

In the green unicellular algae Chlamydomonas moewusii, hyperosmotic stress triggers the activation of both PA biosynthetic pathways (i.e., PLD and PLC/DGK; Munnik et al., 2000). Indeed, with the activation of mastoparan (a potent activator of PLC and PLD signaling in plants), 5–17% of PA is generated by PLD, while the rest is assumed to be generated by the PLC/DGK (Munnik et al., 1998). In cold-shock Arabidopsis stimulated cells (0°C), during the first 10 min up to 80% of PA is produced via the coordinated PLC/DGK pathway, whereas afterward the PLD pathway dominates (Ruelland et al., 2002). In Arabidopsis, the induction of programmed cell death due to the accumulation of H₂O₂ caused by UV irradiation or drought-induced dehydration is drastically reduced as a consequence of an overproduction of PA via PLD (Zhang et al., 2003). Interestingly, oxidative stress and wounding boost PA synthesis via the PLD pathway too (Munnik, 2001; Testerink and Munnik, 2005). This pathway (PLD) is also responsible for the total production of PA in Arabidopsis cells exposed to pathogens (de Torres Zabela et al., 2002). Nevertheless, in tomato cells exposed to the elicitors CH4 and flg22, PA biosynthesis is stimulated by PLC/DGK, whereas PLD does not show activity (van der Luit et al., 2000).

Based on significant contributions by a number of plant biologists (Munnik et al., 1998; Munnik, 2001; Gómez-Merino et al., 2004; Cai et al., 2009; Kirik and Mudgett, 2009; Xue et al., 2009; Gonorazky et al., 2010; Arisz and Munnik, 2011; Dubots et al., 2011; Shin and Loewen, 2011; Boss and Im, 2012; Arisz et al., 2013; Villasuso et al., 2013; Logothetis et al., 2015; Saucedo-García et al., 2015; Singh et al., 2015; Hou et al., 2016), Figure 1 shows the main metabolic pathways involving PLD, PLC, and DGK in the biosynthesis and turnover of PA in plant cells.

Eukaryotic DGKs are a widespread family of enzymes, well conserved among most multicellular organisms. Plant DGK isoforms have a conserved catalytic domain with an ATP-binding site (consensus GXGXXG/A) required for kinase activity (Gómez-Merino et al., 2004). DGK activity is likely to be tightly regulated to control DAG and PA levels coordinately, enabling the cell to perform its physiological functions appropriately (Gómez-Merino et al., 2004). In higher plants, DGK enzymes are present in different isoforms and their activity has been proved in tomato (Solanum lycopersicum; SlDGK; Snedden and Blumwald, 2000), Arabidopsis (AtDGK; Gómez-Merino et al., 2004, 2005a), maize (Zea mays; ZmDGK; Sui et al., 2008), wheat (Triticum aestivum; TaDGK; Zhang et al., 2008), rice (Oryza sativa; OsDGK; Ge et al., 2012), and apple (Malus prunifolia; MpDGK; Li et al., 2015), among others. To date, cumulative evidence points to a crucial role of DGK enzymes in vital processes like growth, development and responses to environmental stimuli and stresses. Phylogenetic analyses have been performed for some plant species, including Arabidopsis and rice DGKs, though a more robust evolution analysis of DGKs in higher plants is still lacking, and their implications in response to nutrients, including potential and well-known beneficial elements, await further studies. Herewith we describe the physiological roles of characterized DGK from higher plants isolated so far and present a comprehensive phylogenetic analysis of DGK isoforms identified in the proteomes of important model and cultivated plants. Furthermore, based on the Genevestigator platform (Zimmermann et al., 2014) and experimental data, we also report a detailed expression profiling analysis of DGK genes in response to known beneficial elements such as aluminum (Al) and sodium (Na), as well as other metal and metalloid ions, including silver (Ag), arsenic (As), cadmium (Cd), chromium (Cr), and mercury (Hg). Beneficial elements are not required by all plants, but may be essential for certain taxa depending on environmental conditions, element concentration and plant species. When applied at low concentrations, they may elicit positive responses in growth, yield, and responses to environmental stresses (Pilon-Smits et al., 2009; Poschenrieder et al., 2013). Indeed, beneficial elements can prompt hormesis, a stimulatory effect of a low dose of a non-essential element (Calabrese, 2014). Herewith we provide evidence for the first time that the expression of some DGK genes may be modified when plants are exposed to some of those hormesis-inducing elements. Furthermore, we explored the expression profiling data of such genes in different plant tissues.

Plants have evolved effective response strategies to react against environmental stimuli and protect themselves from various stress factors. One of the most important signaling pathways that mediate these responses is represented by the PPI, in which one of the emerging components is PA (Arisz et al., 2013; Saucedo-García et al., 2015). In the plasma membrane, PA can be produced principally by the PLD and PLC/DGK pathways. Interestingly, most of the PA synthetized in plants stimulated over a prolonged period is through the PLD pathway (Testerink and Munnik, 2011). Instead, a rapid PA accumulation in response to low temperature stress is generated through the PLC/DGK pathway (Arisz et al., 2013).

DGK activity has been reported in several plant species, including Catharanthus roseus, tobacco (Nicotiana tabacum), wheat, tomato, Arabidopsis, rice, and apple (Gómez-Merino et al., 2005a; Zhang et al., 2008; Li et al., 2015), and the molecular data bases reveal that they are present in a number of other crop plants such as maize, grape, sweet orange, and cotton. Nonetheless, detailed analysis of transcriptional activation of DGK genes and functional analysis of their corresponding protein products are still fragmentary. Two DGK cDNAs, LeDGK1 and LeCBDGK, have been cloned from tomato and found to be derived from the same gene via alternative splicing. LeCBDGK, the protein product of LeCBDGK, contains a calmodulin (CaM)-binding domain (CBD). The two proteins lack the cysteine-rich domain (CRD) present in other eukaryotic DGKs, but are active in vitro. LeCBDGK (SlDGK1b in Table 1) is found both in associations with membranes and in soluble cell extracts. By contrast, LeDGK1 (SlDGK1a in Table 1) only associates with the membrane fraction, via a Ca²⁺/CaM-independent mechanism, which might represent a means of encoding specificity in cellular responses by alternative splicing (Snedden and Blumwald, 2000). In Arabidopsis, the AtDGK1 cDNA has been isolated and reported to be mainly expressed in roots, shoots, and leaves, but its enzyme product was not active in vitro (Katagiri et al., 2001). However, Vaultier et al. (2008) found DGK activity in most membrane compartments, and speculated that AtDGK1 and AtDGK2 contribute to this activity. Two Arabidopsis DGK cDNAs (AtDGK2 and AtDGK7) were cloned and their encoded enzymes were catalytically active in in vitro assays. AtDGK2 transcripts are found in the whole plant except in stems and are induced by exposure to cold (4°C), pointing to a role in cold signal transduction (Gómez-Merino et al., 2004), whereas the AtDGK7 gene is mainly found in flowers, young seedlings, and cauline leaves (Gómez-Merino et al., 2005a). In rice, it has been reported that the overexpression of the DGK gene OsBIDK1 enhances disease resistance in transgenic tobacco (Zhang et al., 2008). Importantly, the inhibition of DGK activity drastically reduces root elongation and plant growth (Gómez-Merino et al., 2005a), indicating that DGKs may play a pivotal role not only in stress responses but also in developmental processes in plants. Whether this inhibition impact on plant development is related to an effect on the nutrient status of the plants remains to be elucidated. Nonetheless, with the highest DGK inhibitor dose used, plants exhibited a general chlorosis and died earlier as the inhibitor concentrations increased in the growth medium, which indeed points to a role of DGK enzymatic activity on nutrient status of the plant (Gómez-Merino et al., 2005a).

In terms of plant nutrition, essential elements are classified into macronutrients and micronutrients, according to the concentrations in which they are found in plant tissues. Apart from C, N, and O, the first group encompasses N, P, K, Ca, Mg, and S, which are found in plant tissues in concentrations of approximately 1000 mg kg^-1 dry biomass weight. Instead, plant micronutrients, which are represented by Cl, Cu, Fe, Mg, Mo, Ni, and Zn, are found in concentrations close to 500 mg kg^-1 of biomass in dry bases (Alcántar-González et al., 2016). Beneficial elements are not essential for plants, but when applied at low dosages, they may improve yield and quality parameters of some plant species of agricultural importance (Pilon-Smits et al., 2009; Trejo-Téllez et al., 2016). To date, the proposed beneficial elements include Al, Ce, Co, Fl, La, Na, Se, Si, Ti, V, and W (Pilon-Smits et al., 2009; Poschenrieder et al., 2013; Trejo-Téllez et al., 2016). Furthermore, other chemical elements such as Ag, As, Cd, and Hg may trigger hormetic effects in plants. Herewith we demonstrated that some DGK genes are induced in response to Ag, Al, Cd, Cr, Hg, and Na, suggesting a possible role of these genes and their protein products on plant nutrition.

We implemented an evolutive method in order to construct a consensus phylogenetic tree of DGKs from higher plants. In doing so, we took into consideration all DGK sequences in which enzymatic activity had been previously demonstrated, or that had been considered in other phylogenetic analyses. We found that plant DGK enzymes were distributed into three major clusters, as previously reported for Arabidopsis, rice, maize, wheat, tomato, and apple (Snedden and Blumwald, 2000; Gómez-Merino et al., 2004; Sui et al., 2008; Zhang et al., 2008; Li et al., 2015).

So far, all plant DGKs exhibit a bipartite catalytic region composed of a catalytic domain (DGKc; Pfam accession number PF00781) followed by an accessory domain (DGKa; Pfam accession number PF00609) at the C-terminus. Since DGKa is associated with DGKc, the former may contribute to the functionality of the catalytic domain. Whereas mammalian DGK enzymes can be classified into five groups according to sequence homology (Topham and Prescott, 2010), plant DGKs only form three phylogenetic clusters, and exhibit a simpler domain organization. Though less varied and complex, plant DGKs included in Cluster I resemble the closest sequence orthologous of metazoan enzymes with proteins that display one or two DAG-binding domains (C1), an upstream basic region and an extended CRD in their N-terminus (Gómez-Merino et al., 2004; Vaultier et al., 2008; Cacas et al., 2016). Conversely, Cluster II DGKs just harbor the DGKa and DGKc domains. Interestingly, Cluster III DGKs may exhibit a C-terminal CBD generated by alternative splicing (Arisz et al., 2009; Cacas et al., 2016).

The yeast S. cerevisiae was initially thought to lack a DGK, though Dgk1p, a novel type of DGK, utilizes CTP rather than the common ATP was discovered (Han et al., 2008a). This enzyme does not exhibit sequence similarity to DAG kinases from other species (Han et al., 2008b), and contains a short motif identified in a family of CTP-dependent phytol and dolichol kinases (Shridas and Waechter, 2006). Together with PA phosphatase, this enzyme controls the levels of PA and DAG for phospholipid synthesis, membrane growth, and lipid droplet formation (Qiu et al., 2013). The fact that the yeast DGK enzyme utilizes CTP, instead of ATP, as the phosphate donor in the reaction, explains why an ATP-dependent DAG kinase activity or a putative gene encoding a DAG kinase enzyme had not been identified in S. cerevisiae before (Han et al., 2008a).

Both experimental data previously reported and microarray analyses revealed that some DGK genes are expressed more prominently in young tissues, whereas their expression declines in older tissues. Importantly, most DGK transcripts are detectable in all tissues throughout all developmental stages, but their abundance decreases in tissues when plants become older.

Expression of DGK genes in response to both biotic and abiotic stimuli and stress factors has been proved (Gómez-Merino et al., 2004; Arisz et al., 2009; Pleskot et al., 2012; Ruelland et al., 2015). Accordingly, our survey revealed that some DGK genes are indeed regulated by Ag, Al, As, Cd, Cr, Hg, and Na, which may display hormetic dose–response curves.

According to Pilon-Smits et al. (2009), Ag can be considered as a potentially beneficial element, while information on the cellular basis for the positive effects of this element in plants is still fragmentary. Currently, silver nanoparticles (AgNPs) have remarkable uses in agricultural and environmental systems. Sharma et al. (2009) have pointed to a crucial role of AgNPs as antimicrobial agents. In plants, AgNPs impact oxidative stress-related gene expression, seed germination, and root elongation, leading to both positive and negative effects on plant growth (Cox et al., 2016). The final effects of AgNPs depend on nanoparticle (NP) size, shape, surface coating, and concentration used. Likewise, plant genotypes differ in their responses to NP exposure. In our survey, when Ag was applied to Arabidopsis plants, inducible gene expression was strongly evident for AtDGK5b and AtDGK7. Indeed, AgNPs have been implicated in signal transduction pathways involving kinases within the PI signaling pathway in eukaryotic cells (Kang et al., 2012), which suggests that Ag might influence other components of this pathway, including PLC and DGK in plants. This topic will be an interesting area for future study.

In the case of Al, Gómez-Merino et al. (2005b) reported induction of AtDGK2 gene expression in hydroponically growing Arabidopsis plants exposed to 200 μM Al in the nutrient solution. In this study, Al induced the expression of the DGK gene rapidly, within 30 min after exposure, and it returned to the pretreated level 20 h after treatment (Gómez-Merino et al., 2005b). Likewise, Martínez-Estévez et al. (2001b) reported that Al entered coffee cells 30 min after treatment, and the signal was retained for up to 2 h. Furthermore, it was found that Al increases phosphorylation of particular proteins in cellular suspension cultures of coffee (Martínez-Estévez et al., 2001a). Subsequently, it was demonstrated that Al quickly (in 1 min) increased enzymatic activity of PLC as well as PI4K, PIP5K, and DGK (Martínez-Estévez et al., 2003), supporting the hypothesis that DGK genes might be implicated in the production of PA in response to Al. Nevertheless, this issue awaits further research at the molecular level. Just recently, Bojórquez-Quintal et al. (2014) reported that Al effects on primary root growth in coffee plants were dose-dependent: 100–300 μM Al stimulated primary root growth, while 500 μM induced damage to the root tips and inhibited primary root growth. Furthermore, 100 μM Al also increased the K and Ca contents by around 33 and 35% in roots, as compared to the control, which demonstrates the beneficial effects of Al when applied at low concentrations. Indeed, Al may improve nutrient status in the Al-resistant wheat cultivar Yecora, since this genotype retained larger concentrations of Ca²⁺ and Mg²⁺ in the leaves (Moustaka et al., 2016). Nonetheless, we have to keep in mind that Al exhibits hormetic effects in plants, which means that low dosages may induce beneficial responses in some genotypes, and higher concentrations are likely to cause toxicity in most plant species, interfering with cytoskeleton structure and function, disrupting Ca homeostasis, hampering P metabolism, and inducing oxidative stress (Blamey et al., 2015). In our survey, only the soybean GmDGK1 gene was tested, showing no significant changes upon Al exposure. Whether other DGK genes are indeed induced by Al is still an open question that remains to be answered in future plant signaling studies.

Arsenic (As) is a metalloid occurring in natural environments in some abundance (1.2–1.4 ppm) in the Earth’s crust and in small quantities in rocks, soils, water bodies, and air. Hence, plants have evolved in the presence of this element, and it has been postulated that As might induce hormetic effects on plant growth. Although the mechanism of action is still unknown, it has been suggested that the growth benefit triggered by As arises from As stimulation of Pi uptake (Tu and Ma, 2003; Finnegan and Chen, 2012). However, because of its general toxic effects in living organisms, As is of considerable concern, and special care has to be taken in order to take advantage of its potential use in agriculture. In our study, As (in the form NaHAsO₄) induced the expression of rice OsDGK2, but slightly downregulated that of OsDGK3a and OsDGK8. This fact, as well, points to a possible role of As in the PI signaling pathway.

In some plant species, Cd has been proved to trigger beneficial protective effects at low doses, in an hormetic dose–response manner (Liu et al., 2015). In fact, hormesis was found during the growth of switchgrass (Panicum virgatum) under hydroponic conditions, since plants were capable of developing better in the nutrient solution containing 100–175 μM Cd at pH 4.1–5.9, showing enlargement of the root length and surface, enhanced absorption of essential nutrients and increased biomass (Wang et al., 2015). Similarly, Liu et al. (2015) reported higher concentrations of chlorophylls (a, b, and total) and carotenoids, as well as relative water contents in leaves of Lonicera japonica plants receiving 0.5–5.0 mg L^-1 Cd in the nutrient solution, as compared to the control. In our analysis, Cd repressed the transcriptional activity of AtDGK1, AtDGK2, AtDGK3, AtDGK4b, AtDGK5b, and AtDGK7. In rice, Cd slightly enhanced the expression of OsDGK8. It has been previously demonstrated that Cd affects the content of PPI in mammalian cells (Borikov and Kaliman, 1999). Just recently, Rajakumar et al. (2016) reported that Cd disrupts lipid metabolism in S. cerevisiae. As a hormetic factor, Cd is also a hazardous metal for plants when certain threshold concentrations in their cells are exceeded (Trejo-Téllez et al., 2014), as a result of its great toxicity inducing oxidative stress, genotoxicity, disruption of the photosynthetic apparatus, and inhibition of root metabolism (López-Millán et al., 2009; Han et al., 2012; Andresen and Küpper, 2013; Liu et al., 2015). Whether Cd interferes with the PI signaling pathway, and especially with the activity of PLC and DGK, is a question worthy of further study.

Cr has been generally referred to as an essential element for animals (Thorvaldsson and Jónsdóttir, 2005). In fact, Ma and Hooda (2010) established that trivalent chromium Cr(III) is an essential/beneficial nutrient that in trace amounts regulates the sugar and cholesterol metabolism in human and animal cells, though its hexavalent form Cr(VI) is a potent carcinogen and extremely toxic for those biological systems. Importantly, a recent report by the European Food Safety Association’s Panel on Dietetic Products, Nutrition and Allergies (EFSA NDA Panel, 2014) determined that there is no evidence of beneficial effects associated with Cr intake in human health, and setting of an adequate Cr intake level is also not appropriate. Concerning the plant system, though its abundance in the Earth’s crust ranges from 100 to 350 ppm (which is higher than that of essential elements like Ni, Zn, and Cu), Cr displays low solubility in soils and plants absorb just small amounts of this element. Hence, normal concentrations of Cr in plant tissues are between 0.02 and 1.0 mg kg^-1, which depends on soil concentrations of this element and plant species (Ma and Hooda, 2010). Similar to humans and animals, Cr(VI) is much more toxic to plants than Cr(III). Just recently, Kabir (2016) reported that Cr stress tolerance in rice cv. Pokkali is not related to metal sequestration but is associated with reduced Fe transport and increased antioxidant defense. As an hormetic element, González et al. (2015) reported that low dosages of Cr (i.e., 2 and 4 mg L^-1) enhanced defense response and maintained photosynthetic activity in Eichhornia crassipes, implying that the antioxidant defense system enzymes attempted to ensure the redox homeostasis. In our analysis, Cr slightly induced the expression of most OsDGK isoforms. Bovykin et al. (1999) found that Cr(III) ions are adsorbed on the bilayer lipid membrane surface, changing the intramembrane potential difference. As pH increases, the adsorption of those ions decreases. Furthermore, McCarty (2006) reported that bioactive Cr can induce the plasmalemmal Ca²⁺-ATPase in mammals, which is connected with the regulation of phospholipid levels (Govindaraju et al., 1989). In fact, it has been well documented that Cr alters phospholipid metabolism in animals (Tandon, 1982), but its implications on this pathway in plants is still at issue.

Low levels of Hg salts induce stress-dependent increase in glucose uptake in mammalian cells (Stearns, 2007). Furthermore, the antimicrobial properties of Hg are well documented (Lemire et al., 2013), though its use for agricultural proposes has been restricted since the 1990s (EPA, 2002). Hormetic effects of Hg have been proved in different organisms, including mammals (Schmidt et al., 2004), Caenorhabditis elegans (Helmcke and Aschner, 2010), Anas platyrhynchos (Heinz et al., 2012), and plants. In Lemna minor and Allium cepa, Subhadra et al. (1991) reported that low levels of aquatic Hg (0.001–1.0 mg L^-1) accelerated catalase and peroxidase activities. In our survey, Hg application in barley resulted in a strong induction of all three HvDGK3 genes (a, b, and c). Nevertheless, additional experimental evidence demonstrating the connection between Hg and the PI signaling pathway remains to be elucidated.

Na has been proved to be an essential element for halophytes (i.e., salt-tolerant plants), including C4 or CAM species that use phosphoenolpyruvate (PEP) to fix CO₂ during photosynthesis, since Na mediates the regeneration of PEP from pyruvate. Additionally, Na has been widely reported to trigger hormetic responses on most plant taxa, including C3 plants (Maathuis, 2013). Importantly, in hydrated form, Na⁺ and K⁺ are chemically and structurally very similar. Consequently, in environments with low K⁺ conditions, Na⁺ can be useful for plants. Hence, several functions carried out by K⁺ plants, including some of the physiological and metabolic ones (i.e., osmotic regulation, guard cell movement, and cell expansion), can be fulfilled by Na⁺. Nevertheless, plant species significantly vary in their capacity to replace K⁺ with Na⁺ (Pilon-Smits et al., 2009). Na is also an excellent accompanying cation for long-distance transport (Subbarao et al., 2003). In our survey, we could observe that Na slightly induced the expression of AtDGK1 and AtDGK2 in shoots. On the other hand, the activity of both genes was slightly repressed in roots. In wheat, Na hardly induced the expression of TaDGK in shoots, while it differentially repressed its expression in roots, depending on the time of exposure and the dosage used. In barley, Na induced the expression of HvDGK3b, but repressed that of HvDGK3c. In tomato, Na strongly induced transcriptional activity of SlDGK1. Interestingly, Peters et al. (2002) reported that the non-specific phospholipase C5 (NPC5) and its derived lipid mediator DAG regulate lateral root development induced by sodium chloride (NaCl) in Arabidopsis thaliana. Moreover, another non-specific phospholipase C, NPC4, promotes responses to abscisic acid and tolerance to hyperosmotic stress induced by NaCl in Arabidopsis (Peters et al., 2010). These findings suggest a connection between Na and the PLC/DGK pathway in plants, which has to be further investigated.

Beneficial elements are emerging as potential biostimulators for agricultural proposes, since they can induce desirable plant responses in a hormetic manner. In light of the growing need for sustainable agriculture, the general global challenges related to climate change and an increasing population demanding more food, beneficial elements open up new possibilities to attain food security. Though cumulative evidence points to beneficial elements like Al and Na having a role in the regulation of DGK gene expression, the possible impact of dosages, chemical forms, time of application, and genotype responses remains to be further investigated. Additionally, understanding the roles of beneficial elements and other hormesis-inducing ions in different plant species under stress conditions may be a possible area of further study. Herewith we provide evidence for the first time that DGK genes encoding enzymes implicated in the synthesis of PA, an emerging second messenger, are responsive to Ag, Al, As, Cd, Cr, Hg, and Na. Importantly, we found most DGK genes ubiquitously expressed in plant tissues, and encoded proteins widely distributed in higher plants. PA, the protein product of DGK activity, has been proved to be a crucial second messenger. Its potential role as signaling mediator in response to beneficial elements and other ions could be of crucial significance for plants, providing the cell with spatial and transient information to perform better under limiting environment conditions, or simply to improve plant production and productivity. The discovery and description of the molecular bases of such relationships would be a noteworthy contribution to the field of agriculture.

HE-S and FG-M developed and designed the experiments and the survey strategies. LT-T supervised the physiological and plant nutrition experiments. PP-R and JH-C supervised the bioinformatics analyses. HE-S and FG-M wrote the manuscript. LT-T, PP-R, and JH-C revised the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.