Glehnia littoralis Fr. Schmidt ex Miq. is a medicinal and edible plant in the Umbelliferae family. Glehnia littoralis is rich in coumarins, coumarin glycosides, phospholipids, and polysaccharides (Yuan et al., 2002). The peeled and dried roots and rhizomes of G. littoralis are commonly used as a traditional Chinese herbal medicine for moistening the lungs, removing phlegm, relieving cough, curing gastrointestinal disorders, recovering from surgery, and immunoregulation, as well as for their anti-inflammatory properties (Huang et al., 2012; Li et al., 2020). Moreover, the tender leaves of G. littoralis are edible as a vegetable. Therefore, the demand for G. littoralis as a clinical medication and health care product is always high. Glehnia littoralis is a perennial halophyte that grows on sandy beaches in Northern Pacific countries and regions, such as eastern China, Japan, the Korean Peninsula, Russia, and the United States (Li et al., 2018). A previous study elucidated how G. littoralis adapts to high-salinity environments by analyzing its anatomical and morphological characteristics, such as the secretory trichomes and thick cuticle cover on leaves (Voronková et al., 2011), but the molecular mechanism of salt adaptation in G. littoralis remains largely unknown.

The phospholipid bilayer of the plasma membrane is the first barrier to the external environment. Phospholipases, including phospholipase A (PLA₁, sPLA₂, and pPLA), phospholipase C (PLC), and phospholipase D (PLD) hydrolyze membrane phospholipids, and affect the structure and stability of the cellular membrane and signaling responses to environmental stimuli (Wang, 2005; Scherer et al., 2010; Hong et al., 2016). The products of phospholipases, such as phosphatidic acid (PA) and inositol 1,4,5-trisphosphate, are cellular messengers that regulate various biological processes (Testerink and Munnik, 2005; Wang et al., 2006; Tang et al., 2007). PLD and PLC/diacylglycerol kinases (DGK) play a major role in the stress-induced generation of PA (Munnik and Testerink, 2009). The activities of the phospholipases are induced rapidly, coupled with a highly dynamic level of PA, in response to hormones, as well as abiotic and biotic stimuli (Bargmann and Munnik, 2006; Li et al., 2006; Peters et al., 2010). PA acts as a cellular mediator by binding with some regulatory proteins, anchoring these proteins to the membrane, modulating their catalytic activities, and altering membrane structure (Wang et al., 2006; Testerink and Munnik, 2011). PA-interacting proteins include various kinases, transcription factors, phosphatases, and other target proteins during cellular processes (Zhang et al., 2004, 2012; Yu et al., 2010; Kim et al., 2019; Shen et al., 2019, 2020).

Non-specific phospholipase C (NPC) is a subtype of PLC that hydrolyzes common membrane phospholipids, such as phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylserine (PS), to produce sn-1,2-diacylglycerol (DAG) and a corresponding phosphorylated headgroup. DAG is phosphorylated by DGK to generate PA. Unlike PI-PLC (another PLC subtype), which is found widely in animals, plants, and bacteria, NPC is found only in bacteria and plants, and has distinct evolutionary features (Nakamura, 2014; Nakamura and Ngo, 2020). In the model plant Arabidopsis thaliana, six NPCs have been identified, most of which have been reported to be involved in multiple biological processes. NPC4 is associated with the plasma membrane, and the recombinant protein of NPC4 shows enzymatic activity toward PC and PE (Nakamura et al., 2005). NPC4 is involved in the plant response to abscisic acid (ABA), auxin, phosphate deficiency, hyperosmotic conditions, and salt and Al stresses (Nakamura et al., 2005; Gaude et al., 2008; Peters et al., 2010; Wimalasekera et al., 2010; Kocourková et al., 2011; Yang et al., 2021). NPC4 and NPC3 are important in brassinolide-mediated signaling in root development (Wimalasekera et al., 2010). NPC4 is also involved in phosphosphingolipid hydrolysis and remodeling in the presence of a phosphate deficiency during root growth (Yang et al., 2021). NPC1 plays a vital role in plant heat tolerance (Krčková et al., 2015). NPC1 is localized at secretory pathway compartments, such as the endoplasmic reticulum or Golgi apparatus, so is NPC2. NPC2 expression is suppressed after Pseudomonas syringae attack, and NPC2 is involved in plant immune responses, such as PTI, ETI, and SA (Krčková et al., 2018). The double-mutant npc2npc6 displays a lethal homozygous phenotype and a defect in the heterozygous gametophyte (Ngo et al., 2018). NPC5 is a cytosolic protein. NPC5 and its derived DAG mediate lateral root development under salt stress, and NPC5 is involved in galactolipid accumulation during phosphate limitation (Gaude et al., 2008; Peters et al., 2014). The localization of NPC6 is present in both chloroplast and microsomal fractions, but not in cytosolic or nuclear fractions (Cai et al., 2020). NPC6 promotes seed oil content and enhances PC and galactolipid turnover to TAG (Cai et al., 2020).

The important roles of NPCs in other plant species have been gradually revealed following the NPC findings in A. thaliana. For example, Cai et al. (2020) reported that BnNPCs affect seed mass and yield, and that BnNPC6.C01 is positively associated with seed oil content in oilseed rape. In rice, five OsNPCs are identified. The cytosolic and membrane-associated OsNPC1 modulates silicon distribution and secondary cell wall deposition in nodes and grains, affecting mechanical strength and seed shattering (Cao et al., 2016). OsNPC6 is involved in mesocotyl elongation in rice. OsNPC6-overexpressing plants exhibit a shorter mesocotyl than the wild-type (WT) and npc6 mutant (Yang et al., 2021). Yang et al. (2021) also indicated that all five OsNPCs showed plasma membrane localization. These studies suggest that plant NPCs are involved in numerous biological processes, although the functional properties of most plant NPCs await exploration. In the current study, we cloned and identified GlNPC3 from G. littoralis, which played a positive role in the salt stress response.

Glehnia littoralis is a medicinal halophyte that grows in coastal habitats. There is an inevitable association between saline adaptability and the accumulation of active medicinal ingredients in G. littoralis. Shu et al. (2019) indicated that the accumulation of furocoumarins was increased by NaCl stress in G. littoralis. Few studies have been reported on the salt tolerance mechanism of G. littoralis. In our previous study, we performed a comprehensive transcriptome analysis of the response of G. littoralis to salt stress, and obtained a large number of differentially expressed genes involved in basic metabolism, secondary metabolism, transportation, and signal transduction (Li et al., 2018). We cloned and identified a differentially expressed NPC gene named GlNPC3 (Figure 1). RT-qPCR showed that the expression level of GlNPC3 increased continuously upon salt treatment (Figure 2). The characterization of A. thaliana GlNPC3-overexpressing plants showed that overexpression of GlNPC3 enhanced the salt tolerance of A. thaliana. Therefore, we performed a series of detailed experiments to elucidate the involvement of GlNPC3 in the salt stress response in G. littoralis.

Similar to most NPCs, GlNPC3 contained a conserved phosphoesterase domain necessary for esterase activity and NPC catalytic activity (Figure 1 and Supplementary Figure S1). In A. thaliana, AtNPC1, AtNPC2, and AtNPC6 contain a signal peptide at the C-terminus, but AtNPC3, AtNPC4, and AtNPC5 lack it (Pokotylo et al., 2013). GlNPC3 contains no signal peptide and was clustered with AtNPC3, AtNPC4, and AtNPC5 in the phylogenetic analysis. The 50–100 amino acids at the C-terminus of NPCs are the most variable regions among NPC subfamilies. These regions may be responsible for various molecular functions and subcellular localization of different NPCs (Pokotylo et al., 2013). The gene expression changes under the different treatments showed that GlNPC3 responds to various stresses (Figure 2). Our analysis of the GlNPC3 promoter sequence also showed the variety of stress-related regulatory elements in the GlNPC3 promoter sequence, suggesting that GlNPC3 might be involved in a variety of transcriptionally regulatory processes. Moreover, we found that the expression of the GUS reporter driven by the GlNPC3 promoter in the roots of A. thaliana was upregulated after NaCl, mannitol or the hormone treatments during 2h (Supplementary Figure S4). However, the expression of GlNPC3 did not increase at 6 and 24h after ABA treatment in the root of G. littoralis (Figure 1B). We supposed that the response of GlNPC3 to ABA might be earlier or later than the time points we analyzed in G. littoralis.

In previous studies, AtNPCs were expressed at different levels in the roots (Peters et al., 2010), and some AtNPCs were involved in the regulation of root development. For example, AtNPC2 and AtNPC6 are required for root growth in Arabidopsis (Ngo et al., 2019). AtNPC4 is involved in the root response to salt stress (Kocourková et al., 2011), and AtNPC5 mediates lateral root development during salt stress (Peters et al., 2014). Our results showed that GlNPC3 was expressed at the highest levels in the roots of G. littoralis, followed by flowers (Figure 4). Expression of the GUS reporter driven by the native GlNPC3 promoter in A. thaliana exhibited a similar pattern with that of G. littoralis. However, there were some differences compared to G. littoralis. For example, Promoter_GlNPC3::GUS was highly expressed in the stem of A. thaliana, while G. littoralis had extremely short rhizomes and the expression level in rhizomes was similar to that in the petiole. Moreover, the roots of A. thaliana GlNPC3-overexpressing plants were longer than that of the WT under the salt treatment (Figure 3). Based on the growth characteristics of G. littoralis, the roots of G. littoralis, which are deeply embedded in the beach, are not only the main medicinal parts but also play a positive role in salt tolerance. Therefore, GlNPC3 may be related to the response of roots to salt stress.

The subcellular localization of GlNPC3 was mainly at the plasma membrane and partially at the tonoplast, different from other NPCs (Figure 5). Both AtNPC1 and AtNPC2 are localized at the endoplasmic reticulum and Golgi apparatus (Krčková et al., 2015; Ngo et al., 2018), and AtNPC2 is localized at the plastid (Ngo et al., 2018). AtNPC4 is associated with the plasma membrane, whereas AtNPC5 localizes to the cytosol (Gaude et al., 2008; Peters et al., 2014). The subcellular localization of the NPC isoforms is generally related to enzyme activity and biological function (Nakamura and Ngo, 2020). The localization of GlNPC3 suggests that it may hydrolyze phospholipids at the plasma membrane and tonoplast, and may be involved in the lipid-mediated signaling pathway or in the lipid metabolism.

Previous studies have reported that NPCs hydrolyze a variety of common lipids, such as PC, PE, and PS (Pokotylo et al., 2013). Changes in NPC substrates and products affect intracellular lipid signaling messengers, and the composition and remodeling of membrane phospholipids, and are involved in a series of cellular responses (Nakamura et al., 2005; Gaude et al., 2008; Krčková et al., 2015; Ngo et al., 2018). We analyzed GlNPC3 catalytic activity using fluorescent head group-labeled lipids as substrates. Compared to previously reported NPCs, the hydrolytic activity of GlNPC3 in the presence of PS was higher than that in the presence of PC in vitro (Figure 6A); however, the PS content in the plants was significantly lower than that of PC and PE (Figures 7A,B,D). Because of the limitation associated with the use of a single molecular species substrate (18:1) in the in vitro assay, detailed analysis of phospholipids was needed in the plants. In this study, NPC enzyme activity increased in GlNPC3-overexpressing plants after the in vitro salt treatment (Figure 6B). Then, the changes in phospholipid content of the WT and OE1 were determined under salt stress (Figure 7). The common phospholipids, such as PC, PS, and PG, decreased significantly in OE1 compared to the WT under NaCl treatment, suggesting that these lipids may be hydrolyzed by GlNPC3 in plant. Direct measurement of the DAG product showed that the DAG contents increased significantly in the WT at 45min after the salt treatment. However, no significant difference was observed between WT and OE1. Previous studies showed that DAG was phosphorylated to PA under hyperosmotic stress conditions (Arisz et al., 2009; Munnik and Testerink, 2009; Nakamura, 2014). We then analyzed PA content; however, the increase in PA did not occur at 45min after the salt treatment. Why OE plants had less substrates during NaCl treatment, and no significant increase in DAG and PA levels? On one hand, the lipid molecules involved in hydrolysis and synthesis are not limited to these classes that we analyzed. For example, PC also contributes to DAG-mediated triacylglycerol (TAG) synthesis (Bates et al., 2009; Lu et al., 2009). Moreover, Cai et al. (2020) found that AtNPC6 promoted polar glycerolipid and galactolipid [monogalactosyldiacylglycerol (MGDG) and digalactosyldiacylglycerol (DGDG)] conversion to TAG production. In contrast, MGDG and DGDG were converted to DAG by monogalactosyldiacylglycerol synthase (MGD) and digalactosyldiacylglycerol synthase (DGD). TAG and galactolipid were not measured in this study. So, we speculated that OE plants may promote polar glycerolipid turnover to DAG to form TAG during salt treatment, and affect DAG and PA levels. Under normal condition, although no significant difference was observed in DAG content between WT and OE1, OE1 still displayed more DAG content than WT (average 5.47 vs. 3.78nmol/mg DM), and less total polar lipids than WT (average 10.49 vs. 11.82nmol/mg DM; Figures 7F,G). This data suggested that GlNPC3 overexpression contributed the generation of DAG under the normal condition. Together, lipid profiles suggested over-accumulation of GlNPC3 affected the proportion of different membrane lipids and the spatial and electronic properties of membrane, conferring salt tolerance to plants.

On the other hand, the “time limitation” might affect the determination of PA and DAG content. A significantly salt-induced increase in PA and DAG between WT and OE line may occur in the other time points during salt treatment. PA is metabolized rapidly in plants, requiring appropriate time points and rigorous operation, and the extraction and determination of PA are complicated and costly (Welti et al., 2002; Shadyro et al., 2004; Nguyen et al., 2017). Because of the spatiotemporal complexity of monitoring lipid in living cells and tissues of plants, a real-time quantitative and visualized method is required. For example, Potocky et al. (2014) constructed a PA biosensor with a Spo20p-PA binding domain fusing YFP protein, which reflected PA dynamics in transiently transformed tobacco pollen tubes as shown by confocal laser scanning microscopy. Subsequently, Li et al. (2019) developed a PA-specific biosensor, PAleon, which monitors and visualizes the concentration and dynamics of PA in plant cells under abiotic stress based on Förster resonance energy transfer (FRET) technology. PA biosensors have revealed the spatiotemporal dynamics of PA in plants. Similarly, DAG, PS, PtdIns(4,5)P₂ (phosphatidylinositol-4,5-biphosphates), PtdIns4P (phosphatidylinositol-4-phosphate), and PtdIns3P (phosphatidylinositol-3-phosphate) biosensors have been successfully used in plant cells (Vermeer et al., 2006, 2009; van Leeuwen et al., 2007; Thole et al., 2008; Yamaoka et al., 2011; Simon et al., 2014; Vermeer et al., 2017). These biosensors could be used for real-time monitoring of lipid dynamics and provide new insight into lipid metabolism in G. littoralis.

Finally, we analyzed the expression of some stress-related genes in WT and VIGS-silenced G. littoralis plants to investigate whether the GlNPC3-mediated lipid changes were associated with gene expression. The expression of SOS1 (salt overly sensitive 1), TPC1 (two-pore channel 1), SnRK2 (Ser/Thr protein kinase), P5CS (∆1-pyrroline-5-carboxylate synthetase), CIPK (CBL-interacting protein kinase), and the WRKY transcription factor genes in G. littoralis was significantly upregulated under the salt treatment, but little changed after VIGS-induced GlNPC3 silencing of the plants (Figure 8). These genes play pivotal roles in responses to multiple stresses in model plants. Among them, the vacuolar ion channel TPC1 is responsible for a reactive oxygen species-assisted Ca²⁺ wave during the salt stress response (Evans et al., 2016). The SOS signaling pathway gene SOS1 plays a key role in Na⁺ export (Zhu, 2001, 2016). Therefore, combined with the membrane localization of GlNPC3, these findings suggest that a GlNPC3-mediated lipid change may be accompanied by the response of stress-related genes affecting G. littoralis salt tolerance.

In conclusion, we cloned and identified a NPC (GlNPC3) from the medicinal halophyte G. littoralis, and uncovered a potentially functional connection between GlNPC3-mediated phospholipid signaling and the salt-stress response in G. littoralis. Further investigation is needed to determine the role of GlNPC3 in lipid remodeling, and how it is regulated by upstream mediators during the stress response. These results will provide insights into studies on plants living in special environments.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

LL and YZ designed the experiments. LL performed the experiments, analyzed the data, and wrote the manuscript. YB provided support for protein expression and enzyme activity assays. NL and QC provided support for plant growth. XQ helped in bioinformatics analysis. HF, XY, and DL provided the materials and technical support. CL and YZ revised the manuscript. All authors contributed to the article and approved the submitted version.

This study was supported by grants from the National Natural Science Foundation of China (No. 31800272).

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.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.