The Examination of the Role of Rice Lysophosphatidic Acid Acyltransferase 2 in Response to Salt and Drought Stresses

Phosphatidic acid (PA) is an important signal molecule in various biological processes including osmotic stress. Lysophosphatidic acid acyltransferase (LPAT) acylates the sn-2 position of the glycerol backbone of lysophosphatidic acid (LPA) to produce PA. The role of LPAT2 and its PA in osmotic stress response remains elusive in plants. Here we showed that LPAT2-derived PA is important for salt and drought stress tolerance in rice. Rice LPAT2 was localized to the endoplasmic reticulum (ER) to catalyze the PA synthesis. The LPAT2 transcript was induced by osmotic stress such as high salinity and water deficit. To reveal its role in osmotic stress response, an LPAT2 knockdown mutant, designated lpat2, was isolated from rice, which contained a reduced PA level relative to wild type (WT) plants under salt stress and water deficit. The lpat2 mutant was more susceptible to osmotic stress and less sensitive to abscisic acid (ABA) than that of WT, which was recovered by either PA supplementation or genetic LPAT2 complementation. Moreover, suppressed LPAT2 also led to a large number of differentially expressed genes (DEGs) involved in diverse processes, particularly, in ABA response, kinase signaling, and ion homeostasis in response to salt stress. Together, LPAT2-produced PA plays a positive role in osmotic tolerance through mediating ABA response, which leads to transcriptional alteration of genes related to ABA response, protein kinase signaling, and ion homeostasis.


Introduction
Osmotic stress, mostly imposed by drought and high salinity, is the most adverse factor that inhibits plant growth, development, and crop yield, and hence reduces agricultural productivity globally [1][2][3][4]. Plants have developed different mechanisms to adapt to environmental stresses [4][5][6][7][8]. To protect cells from damage by osmotic stress, plants must sense environmental stimuli and trigger the response to proceed the signal transduction, resulting in the changes of protein activity, subcellular localization, gene expression, which thus causes physiological, metabolic, and morphological alterations to deal with adverse environment [4][5][6][7][8][9][10]. Of these mechanisms, sensing environmental stimuli is an early and important event. However, our knowledge about molecular mechanisms in the early event of signaling involved in osmotic stress is still limited.
Abscisic acid (ABA) is a key phytohormone involved in stomatal closure to reduce water loss in response to osmotic stress including salt and drought stresses [2,10]. It also plays important roles in seed maturation, seed germination, and flowering [11]. A higher physiological level of ABA under osmotic stress has been shown to regulate signal cascades in which protein kinases are involved [12]. For example, high saline conditions guide the cells to produce more endogenous ABA, leading to stomatal closure to reduce water loss [2]. The ABA receptors PYR (pyrabactin resistance) and PYL (pyrabactin resistance-like) have been identified in Arabidopsis [13,14]. Two components, sucrose nonfermenting (SNF1)related protein kinase 2 (SnRK2) and protein phosphatase (PP2C, also named as ABI1), are with yeast LPAT (ScLPAT0231) (Supplementary Figure S2), suggesting that rice OsLPAT1 and OsLPAT3 are more closely related to Arabidopsis LPAT1. Rice OsLPAT4 (Os05g0502200), OsLPAT5 (Os01g0782500), Arabidopsis AtLPAT4 (At1g75020), and AtLPAT5 (At4g18850) are classified into the same group containing algal LPAT (NoLPATU6U0). Rice OsLPAT2 (Os11g0637800) is grouped into the same clade with Arabidopsis AtLPAT2 (At3g57650) and AtLPAT3 (At1g51260) with more distantly related to other LPATs within the clade (Supplementary Figure S2).
Spatial and temporal regulation is important to all signaling events, which is particularly critical to lipid messengers because of their limited mobility. To investigate the tissue distributions of LPAT2 transcript, the total RNA was extracted from various tissues for the expression analysis by RT-qPCR. The LPAT2 mRNA was expressed in different tissues including leaf, root, stem, and sheath with the highest level in the leaf ( Figure 1A). The LPAT2 expression was dramatically induced by ABA treatment and salt stress, with the highest 12 h after treatment with ABA and NaCl, as compared to control plants without treatment ( Figure 1B,C). The LPAT2 transcription level was also highly induced by water deficit being three-fold higher than that of control conditions ( Figure 1D). These results suggest that LPAT2 is induced by osmotic stress and may be involved in osmotic stress response. . Two-week-old plants were treated with 10 μM ABA for 12 h. The LPAT2 transcript level was determined by RT-qPCR by using β-Actin as an internal standard. ** indicates significant difference at p < 0.01 compared to WT plants based on Student's t test.

LPAT2 Enhances Plant Tolerance to Osmotic Stress
To determine whether LPAT2 is involved in osmotic stress, four-leaf-stage seedlings of the lpat2 mutant and WT were transferred to soil and treated with NaCl stress and water deficit after one week of recovery. Under control conditions grown in soil with regular watering, the lpat2 mutant exhibited a similar phenotype with WT ( Figure 2A). However, when plants were treated with 150 mM NaCl, the shoot fresh weight of lpat2 Total RNA was extracted from leaves of two-week-old plants at different time points (0, 3, 6, 12, and 24 h) for 100 mM NaCl and 10-µM ABA treatments. For water deficit, the leaves were sampled from plants grown under well-watered conditions (control) or under water deficit (25 to 30% of soil water capacity) when watering were withheld for five days. Values are means ± SD (n = 3). (E) The T-DNA insertion site in rice LPAT2. The arrows indicate the locations of primers for T-DNA insertion by PCR. (F) The lpat2 mutant isolation by PCR. (G,H) The LPAT2 expression level in the lpat2 mutant, wild type (WT) and LPAT2 complementation (COM) plants under the control conditions (G) and ABA treatment (H). Data are means ± SD (n = 3). Two-week-old plants were treated with 10 µM ABA for 12 h. The LPAT2 transcript level was determined by RT-qPCR by using β-Actin as an internal standard. ** indicates significant difference at p < 0.01 compared to WT plants based on Student's t test.

of 19
The role of LPAT2 and its PA in osmotic stress response remains elusive due to the LPAT2 knockout mutant being lethal at the female gametophytic developmental stage [26]. To further characterize the role of LPAT2 in post-embryonic stages, a knockdown mutant, designated as lpat2, was isolated in rice in this study ( Figure 1E-G). The lpat2 mutant was caused by a T-DNA insertion in the LPAT2 promoter region located 358 bp upstream of the start codon ( Figure 1E,F). The LPAT2 transcript level in the lpat2 mutant was substantially lower than that of WT, with~40% of WT ( Figure 1G). The results suggest that lpat2 is a knockdown mutant with a significantly lower level of LPAT2 mRNA relative to WT. Moreover, the LPAT2 mRNA level of the lpat2 mutant was 25% of WT in response to ABA treatment due to less induction in the lpat2 mutant relative to WT ( Figure 1H).

LPAT2 Enhances Plant Tolerance to Osmotic Stress
To determine whether LPAT2 is involved in osmotic stress, four-leaf-stage seedlings of the lpat2 mutant and WT were transferred to soil and treated with NaCl stress and water deficit after one week of recovery. Under control conditions grown in soil with regular watering, the lpat2 mutant exhibited a similar phenotype with WT ( Figure 2A). However, when plants were treated with 150 mM NaCl, the shoot fresh weight of lpat2 mutant plants was decreased by~20%, as compared to WT ( Figure 2B). The lpat2 mutant plants were more damaged in response to salt stress, exhibiting higher malondialdehyde (MDA) content and increased ionic leakage relative to that of WT ( Figure 2C,D). Likewise, the lpat2 mutant was also more susceptible to water deficit with decreased shoot fresh weight and increased cellular damage (Figure 2). During the course of water deficit, the lpat2 mutant seedlings wilted faster than WT, only 32% of the lpat2 mutant seedlings were recovered after 7 days of water deficit treatment followed by re-watering, whereas 59% of WT seedlings were recovered.
To confirm the role of LPAT2 in osmotic stress, the full length of the LPAT2 codon sequence under the control of its native promoter was introduced into the lpat2 mutant to generate genetic LPAT2 complementation (COM) plants. The LPAT2 transcript level in the COM plants was significantly higher than that of the lpat2 mutant and comparable to that of WT plants as confirmed by RT-qPCR ( Figure 1G). The sensitivity to salt and drought stresses in the lpat2 mutant was rescued by genetic LPAT2 complementation to the WT phenotype. The shoot fresh weight, ionic leakage, and MDA content in COM plants were almost similar to those in WT ( Figure 2). These results suggest that LPAT2 plays a positive role in osmotic stress tolerance.

LPAT2 Enhances ABA Response
Abscisic acid is an important phytohormone playing a positive role in osmotic stress tolerance via reduced water transpiration under osmotic stress [2,10]. To determine whether the involvement of LPAT2 in osmotic stress is mediated by ABA response, the seed germination was compared among WT, lpat2, and COM in response to ABA treatment. The germination rate of the lpat2 mutant was not significantly different from that of WT and COM under control conditions without ABA treatment ( Figure 3A,B). In the presence of 10 µM ABA, the seed germination rate in the lpat2 mutant was significantly higher than that of WT. The germination rate of the lpat2 mutant was~70% on the sixth day of the ABA treatment, whereas that of WT seeds was~37% only ( Figure 3C). The results suggest that the lpat2 mutant seeds were less sensitive to ABA compared to WT. Upon exposure to salt stress, an opposite phenotype of the lpat2 mutant was found with a lower germination rate than that of WT ( Figure 3D). A similar tendency was found when seeds were treated with 10% PEG ( Figure 3E). Similarly, the lpat2 mutant seedlings were less sensitive to ABA treatment than that of WT, exhibiting less inhibited growth in the lpat2 mutant with longer shoots and greater shoot fresh weight than that of WT in the presence of ABA ( Figure 3F,G). The less sensitivity to ABA in the lpat2 mutant was rescued to WT phenotype by LPAT2 complementation (Figure 3). By contrast, the lpat2 mutant seedlings were more sensitive to osmotic stress, with substantial reductions of shoot length and fresh weight, as compared to WT and COM under salt stress and polyethylene glycol (PEG) treatments ( Figure 3F,G). Moreover, the lpat2 mutant leaves increased water loss compared to WT and COM under dehydration condition when plants were removed from soil in various time points ( Figure 3H). The results suggest that LPAT2 enhances ABA sensitivity and plays a positive role in osmotic stress tolerance. To confirm the role of LPAT2 in osmotic stress, the full length of the LPAT2 codon sequence under the control of its native promoter was introduced into the lpat2 mutant to generate genetic LPAT2 complementation (COM) plants. The LPAT2 transcript level in the COM plants was significantly higher than that of the lpat2 mutant and comparable to that of WT plants as confirmed by RT-qPCR ( Figure 1G). The sensitivity to salt and drought stresses in the lpat2 mutant was rescued by genetic LPAT2 complementation to the WT phenotype. The shoot fresh weight, ionic leakage, and MDA content in COM plants were almost similar to those in WT ( Figure 2). These results suggest that LPAT2 plays a positive role in osmotic stress tolerance.

LPAT2 Enhances ABA Response
Abscisic acid is an important phytohormone playing a positive role in osmotic stress tolerance via reduced water transpiration under osmotic stress [2,10]. To determine whether the involvement of LPAT2 in osmotic stress is mediated by ABA response, the seed germination was compared among WT, lpat2, and COM in response to ABA treatment. The germination rate of the lpat2 mutant was not significantly different from that of WT and COM under control conditions without ABA treatment ( Figure 3A,B). In the presence of 10 μM ABA, the seed germination rate in the lpat2 mutant was significantly higher than that of WT. The germination rate of the lpat2 mutant was ~70% on the sixth day of the ABA treatment, whereas that of WT seeds was ~37% only ( Figure 3C). The

LPAT2 Is Localized to the ER and Catalyzes LysoPA Acylation to PA
In plants, PA is de novo synthesized via two pathways including the eukaryotic pathway in the ER and the prokaryotic pathway in chloroplasts [23]. To investigate whether rice LPAT2 is involved in the prokaryotic or the eukaryotic pathway, the LPAT2 subcellular localization was assessed by fusing LPAT2 with green fluorescent protein (GFP) at the C-terminus transiently expressed in tobacco leaves. Confocal microscopy observation revealed that the GFP fluorescent signal of the LPAT2-GFP overlapped with the ER marker protein ( Figure 4A), suggesting that LPAT2 is localized to the ER. Furthermore, to test whether rice LPAT2 encoded protein has acylation activity, the recombinant LPAT2 with fused His-tagged at the N-terminus was expressed in E. coli cells ( Figure 4C). The enzymatic assay showed that LPAT2 is capable of catalyzing the LPA acylation to produce PA by using different acyl-CoA species such as 18:1-CoA and 18:3-CoA. A higher activity was found when using 18:3-CoA as an acyl donor ( Figure 4B). By contrast, the reaction with proteins extracted from E. coli cells containing the empty vector only had relatively low PA production ( Figure 4B). The results suggest that LPAT2 catalyzes LPA to produce PA in the ER, which is involved in the PA do novo synthesis via eukaryotic pathway. mutant seedlings were more sensitive to osmotic stress, with substantial reductions of shoot length and fresh weight, as compared to WT and COM under salt stress and polyethylene glycol (PEG) treatments ( Figure 3F,G). Moreover, the lpat2 mutant leaves increased water loss compared to WT and COM under dehydration condition when plants were removed from soil in various time points ( Figure 3H). The results suggest that LPAT2 enhances ABA sensitivity and plays a positive role in osmotic stress tolerance. Figure 3. Suppressed LPAT2 led to less sensitivity to ABA and more sensitivity to osmotic stress in seed germination and seedling growth. (A-E) The seed germination rate in response to ABA and osmotic stress. The seeds were germinated in various conditions including the control, 10 μM ABA, 100 mM NaCl, and 10% polyethylene glycol (PEG) treatments. One hundred seeds per genotype were measured in each experiment. Values are means ± SD (n = 3). (F,G) Shoot length and shoot fresh weight of seedlings in response to ABA and osmotic stress. Four-day-old seedlings were transferred to small plastic cups without supplement (control), or with 10 μM ABA, 100 mM NaCl, and 10% PEG. Values are means ± SD (n = 15) from one representative of three independent experiments. (H) Water loss in leaves of rice plants. Three-week-old plants were removed from soils and left in ambient conditions. The water loss in leaves was expressed as the percentage of decrease in leaf fresh weight. Values are means ± SD (n = 3) from one of three independent experiments. * and ** indicate significant differences at p < 0.05, and p < 0.01, respectively, compared to the wild type (WT) based on Student's t test.

LPAT2 Is Localized to the ER and Catalyzes LysoPA Acylation to PA
In plants, PA is de novo synthesized via two pathways including the eukaryotic pathway in the ER and the prokaryotic pathway in chloroplasts [23]. To investigate whether rice LPAT2 is involved in the prokaryotic or the eukaryotic pathway, the LPAT2 subcellular localization was assessed by fusing LPAT2 with green fluorescent protein (GFP) at the C-terminus transiently expressed in tobacco leaves. Confocal mi- Figure 3. Suppressed LPAT2 led to less sensitivity to ABA and more sensitivity to osmotic stress in seed germination and seedling growth. (A-E) The seed germination rate in response to ABA and osmotic stress. The seeds were germinated in various conditions including the control, 10 µM ABA, 100 mM NaCl, and 10% polyethylene glycol (PEG) treatments. One hundred seeds per genotype were measured in each experiment. Values are means ± SD (n = 3). (F,G) Shoot length and shoot fresh weight of seedlings in response to ABA and osmotic stress. Four-day-old seedlings were transferred to small plastic cups without supplement (control), or with 10 µM ABA, 100 mM NaCl, and 10% PEG. Values are means ± SD (n = 15) from one representative of three independent experiments. (H) Water loss in leaves of rice plants. Three-week-old plants were removed from soils and left in ambient conditions. The water loss in leaves was expressed as the percentage of decrease in leaf fresh weight. Values are means ± SD (n = 3) from one of three independent experiments. * and ** indicate significant differences at p < 0.05, and p < 0.01, respectively, compared to the wild type (WT) based on Student's t test.

Suppressed LPAT2 Leads to a Substantial Reduction of PA in Plants
The enzymatic activity assay showed that LPAT2 catalyzed LPA to PA. To further test whether suppressed LPAT2 affects PA content in rice plants, lipids were extracted from leaves of lpat2, WT, and COM plants treated without (control) and with salt stress or water deficit for lipid quantitative analysis. Under control conditions without salt and water deficit treatments, the PA level in the lpat2 mutant was slightly lower but not significantly different from that of WT and COM plants ( Figure 4D). However, the PA level in the lpat2 mutant was decreased by~50%, relative to that of WT when plants were treated with 150 mM NaCl for 12 h. The reduced PA in the lpat2 mutant was recovered to WT by LPAT2 complementation ( Figure 4D). Likewise, the PA level in the lpat2 mutant was also significantly lower than that of WT and COM under water deficit condition ( Figure 4D). Together, these results suggest LPAT2 contributes PA production in response to salt and water deficit stresses, and that LPAT2 and its product PA may be important for salt and drought stress tolerance in plants.
ER. Furthermore, to test whether rice LPAT2 encoded protein has acylation activity, the recombinant LPAT2 with fused His-tagged at the N-terminus was expressed in E. coli cells ( Figure 4C). The enzymatic assay showed that LPAT2 is capable of catalyzing the LPA acylation to produce PA by using different acyl-CoA species such as 18:1-CoA and 18:3-CoA. A higher activity was found when using 18:3-CoA as an acyl donor ( Figure  4B). By contrast, the reaction with proteins extracted from E. coli cells containing the empty vector only had relatively low PA production ( Figure 4B). The results suggest that LPAT2 catalyzes LPA to produce PA in the ER, which is involved in the PA do novo synthesis via eukaryotic pathway.  To test whether suppressed LPAT2 affects other glycerolipids, phospholipids such as phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and galactolipids were also quantified. Under the control condition, the contents of PC, PE, and PG in the lpat2 mutant were not significantly different from that of WT, whereas PI content in the lpat2 mutant was higher than that of WT ( Figure 5). However, the PC content in the lpat2 mutant was significantly lower than that of WT, whereas the PE content in the lpat2 mutant was lower, but not significantly different from that of WT when plants were treated with 150 mM NaCl for 12 h ( Figure 5). The contents of galactolipids such as monogalactosyldiacyglycerol (MGDG) and digalactosyldiacylglycerol (DGDG) were not obviously different between the lpat2 mutant and WT under both the control and salt stress conditions ( Figure 5). The results suggest that suppressed LPAT2 led to reduced PC in some extent in response to salt stress. sponse to salt and water deficit stresses, and that LPAT2 and its product PA may be important for salt and drought stress tolerance in plants.
To test whether suppressed LPAT2 affects other glycerolipids, phospholipids such as phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and galactolipids were also quantified. Under the control condition, the contents of PC, PE, and PG in the lpat2 mutant were not significantly different from that of WT, whereas PI content in the lpat2 mutant was higher than that of WT ( Figure 5). However, the PC content in the lpat2 mutant was significantly lower than that of WT, whereas the PE content in the lpat2 mutant was lower, but not significantly different from that of WT when plants were treated with 150 mM NaCl for 12 h ( Figure 5). The contents of galactolipids such as monogalactosyldiacyglycerol (MGDG) and digalactosyldiacylglycerol (DGDG) were not obviously different between the lpat2 mutant and WT under both the control and salt stress conditions ( Figure 5). The results suggest that suppressed LPAT2 led to reduced PC in some extent in response to salt stress.

Exogenous PA Application in the lpat2 Mutant Rescues Its ABA Sensitivity and Osmotic Stress Tolerance
The lpat2 mutant had a lower PA level and less tolerance to osmotic stress as compared to WT plants. To investigate whether LPAT2 itself or LPAT2-derived PA is responsible for ABA signaling and osmotic stress tolerance, the seedlings were transferred to growth medium without or with exogenous PA supplementation under ABA and NaCl treatments. The results showed that the lpat2 mutant was not significantly different from WT under the control condition without supplementations of ABA and NaCl. However, the lpat2 mutant seedlings were less sensitive to ABA treatment with longer shoots and greater shoot fresh weight than that of WT ( Figure 6A-C). When exogenous PA was applied to the growth medium, the lpat2 mutant seedlings restored the sensitivity to ABA, and the shoot length and fresh weight of lpat2 seedlings were rescued to that of WT ( Figure 6A-C). By contrast, the lpat2 mutant had less tolerance to salt stress with shorter shoots and smaller fresh weight than that of WT under salt stress, and the growth inhibition in the lpat2 mutant was relieved by exogenous PA application ( Figure 6D-F). These results suggest that LPAT2-derived PA is required for ABA response and osmotic tolerance.
PA was applied to the growth medium, the lpat2 mutant seedlings restored the sensitivity to ABA, and the shoot length and fresh weight of lpat2 seedlings were rescued to that of WT ( Figure 6A-C). By contrast, the lpat2 mutant had less tolerance to salt stress with shorter shoots and smaller fresh weight than that of WT under salt stress, and the growth inhibition in the lpat2 mutant was relieved by exogenous PA application ( Figure 6D-F). These results suggest that LPAT2-derived PA is required for ABA response and osmotic tolerance. Figure 6. Exogenous PA application rescued the lpat2 mutant phenotype to WT in response to ABA and salt stress. (A-C) Exogenous PA application rescued the lpat2 mutant phenotype to WT in ABA response. Five-day-old seedlings were transferred to liquid medium without supplement (control), with 10 µM ABA, and 10 µM ABA + 20 µg PA for one week. Values are means ± SD (n = 10) from one of three independent experiments. (D-F) Exogenous PA application rescued the lpat2 mutant phenotype to WT in salt stress response. Five-day-old seedlings were transferred to liquid medium without supplement (control), with 100 mM NaCl, and 100 mM NaCl + 20 µg PA for one week. Values are means ± SD (n = 10) from one of three independent experiments. ** indicates significant difference at p < 0.01 compared with wild type (WT) based on Student's t test.

Suppressed LPAT2 Leads to Altered Expression of Genes Involved in Diverse Processes
Phosphatidic acid in a specific membrane region has been shown to bind to various effectors including transcription factors, protein kinases involved in diverse physiological processes such as cell proliferation, cell growth, cell death, vesicular trafficking, hormone response, and abiotic stress responses [19,20,30]. To investigate the effect of LPAT2-derived PA on transcriptome in response to osmotic stress, total RNA was extracted from the leaves of WT and the lpat2 mutant plants treated without (control) and with 100 mM NaCl for 12 h, for RNA-sequencing. After purification and quality control of mRNA, the differentially expressed genes (DEGs) were identified by comparing the control versus (vs.) salt stress conditions. The number of DEGs in WT was greater than that of the lpat2 mutant in response to salt stress. A total of 3776 DEGs in WT was found, with 1881 DEGs upregulated and 1895 DEGs downregulated in response to salt stress, whereas that of the lpat2 mutant was only 2849 with 1172 DEGs upregulated and 1677 DEGs downregulated in response to salt stress ( Figure 7A). The results suggest that suppressed LPAT2 leads to decreased DEG number in response to salt relative to control conditions. However, the number of DEGs between lpat2 vs. WT was 2658 under control conditions, whereas the number of DEGs between lpat2 vs. WT was 665 under salt stress ( Figure 7B-D). The results may indicate that LPAT2 has high impact on overall transcriptional regulation under control conditions without salt stress but the more specific genes were transcriptionally changed in response to salt stress.

Suppressed LPAT2 Leads to Altered Expression of Genes Involved in Diverse Processes
Phosphatidic acid in a specific membrane region has been shown to bind to various effectors including transcription factors, protein kinases involved in diverse physiological processes such as cell proliferation, cell growth, cell death, vesicular trafficking, hormone response, and abiotic stress responses [19,20,30]. To investigate the effect of LPAT2-derived PA on transcriptome in response to osmotic stress, total RNA was extracted from the leaves of WT and the lpat2 mutant plants treated without (control) and with 100 mM NaCl for 12 h, for RNA-sequencing. After purification and quality control of mRNA, the differentially expressed genes (DEGs) were identified by comparing the control versus (vs.) salt stress conditions. The number of DEGs in WT was greater than that of the lpat2 mutant in response to salt stress. A total of 3776 DEGs in WT was found, with 1881 DEGs upregulated and 1895 DEGs downregulated in response to salt stress, whereas that of the lpat2 mutant was only 2849 with 1172 DEGs upregulated and 1677 DEGs downregulated in response to salt stress ( Figure 7A). The results suggest that suppressed LPAT2 leads to decreased DEG number in response to salt relative to control conditions. However, the number of DEGs between lpat2 vs. WT was 2658 under control conditions, whereas the number of DEGs between lpat2 vs. WT was 665 under salt stress ( Figure 7B-D). The results may indicate that LPAT2 has high impact on overall transcriptional regulation under control conditions without salt stress but the more specific genes were transcriptionally changed in response to salt stress.  In response to salt stress, upregulated DEGs in the lpat2 mutant were predominantly involved in plant hormone signaling transduction, MAPK signaling pathway (Figure 8), whereas downregulated DEGs were enriched ribosome and ribosome biogenesis ( Figure  8). These results suggest that LPAT2 is important for hormone response and MAPK sig- In response to salt stress, upregulated DEGs in the lpat2 mutant were predominantly involved in plant hormone signaling transduction, MAPK signaling pathway (Figure 8), whereas downregulated DEGs were enriched ribosome and ribosome biogenesis (Figure 8). These results suggest that LPAT2 is important for hormone response and MAPK signaling in osmotic stress and that PA produced by LPAT2 may be a major mediator in these processes. under control and salt (100 mM NaCl) stress conditions. (C) The numbers of upregulated and downregulated genes of lpat2 vs. WT under control condition. (D) The numbers of upregulated and downregulated genes of lpat2 vs. WT under NaCl stress condition. Red dots represent upregulated genes and green dots represent downregulated genes. The total RNAs were extracted from leaves of three-week-old plants treated without (control) or with 100 mM NaCl for 12 h for RNA-sequencing.
In response to salt stress, upregulated DEGs in the lpat2 mutant were predominantly involved in plant hormone signaling transduction, MAPK signaling pathway (Figure 8), whereas downregulated DEGs were enriched ribosome and ribosome biogenesis ( Figure  8). These results suggest that LPAT2 is important for hormone response and MAPK signaling in osmotic stress and that PA produced by LPAT2 may be a major mediator in these processes. To investigate whether altered LPAT2 and its PA affect the expression levels of genes related to osmotic stress, some genes present in DEGs were selected to compare their average values and shown in heat map (Figure 9). The transcript levels of genes involved in ABA response such as ABA-induced proteins (Os04g0423400, Os01g0963600) and SnRK1A protein kinase (Os08g0516900) in the lpat2 mutant were downregulated under salt stress conditions ( Figure 9A). The expression levels of genes related to guard cell anion channel protein (Os04g0574700), phosphate transport (Os04g0185600) and ATP/ADP carrier (Os05g0302700) in the lpat2 mutant were also downregulated as compared to WT under salt stress conditions ( Figure 9A). The expression levels of genes encoding transcription factors in osmotic stress, such as Myb-like (Os02g0174000), NAC (Os12g0477400), and SRF (Os04g0387300), in the lpat2 mutant were down-regulated relative to WT under salt stress conditions ( Figure 9A). However, the expression levels of genes involved in syntheses of ABA (Os07g0164900), jasmonic acid (JA) (Os05g0102000), lipid oxidation (Os12g0559200), and glycolysis (Os06g0151900, Os01g0225400) were upregulated in the lpat2 mutant relative to WT under salt stress conditions ( Figure 9A). Twelve genes encoding zinc finger proteins were altered with eight of them downregulated and four of them upregulated in the lpat2 mutant relative to WT under salt stress condition ( Figure 9B). Moreover, the expression levels of 28 genes related to kinases in the lpat2 mutants were altered with more promising under salt stress conditions ( Figure 9C). In addition, suppressed LPAT2 also led to increased expression levels of genes involved in lipid hydrolysis processes in which include PLD (Os06g0604200, Os06g0604300, Os03g0391400), PLC (Os05g0127200, Os07g0694000, Os01g0955000), phospholipase A (PLA, Os11g0614400, Os01g0652300, Os12g0611300), and GDSL esterases/lipases under control conditions. In particular, the transcript levels of 24 GDSL genes were increased in the lpat2 mutant relative to WT ( Figure 9D). In accordance with induced GDSL genes, the expression levels of 21 peroxidase genes involved in cellular oxidation were also upregulated in the lpat2 mutant under control conditions ( Figure 9E). By contrast, the expression levels of genes encoding NAC transcription factors, Myb transcription factors and ion transports were downregulated in the lpat2 mutant under control conditions ( Figure 9F). Taken together, the results suggest that LPAT2 and its PA play important roles in transcriptional alternations of genes related to ABA response, ion homeostasis, protein kinases in under osmotic stress. In addition, suppressed LPAT2 may lead to increased lipid hydrolysis processes under control conditions.

Discussion
Of three major sources of PA, PLD-derived PA plays an important role in drought and salt stresses [31,32,34,36]. PLDα1-derived PA enhances stomatal closure under water deficit and ABA treatment in Arabidopsis [31][32][33]. PLDα3 knockout mutant contains a lower PA level and exhibits susceptibility to salt stress compared to WT plants [34]. The pldα1/δ double mutant plants with reduced PA level are more sensitive to salt stress than that of WT and the single mutant [35,36]. Phosphatidic acid produced from PLC/DGK also plays a role in osmotic stress response [38,39]. However, whether the de novo synthesized PA is involved in osmotic stress remains unknown. LPAT catalyzes a committed step in PA de novo synthetic pathway [25,26]. In Arabidopsis, LPAT1 is involved in the prokaryotic pathway of PA synthesis and is essential for embryo development [25,28]. Of the remaining four LPATs involved in the eukaryotic pathway, LPAT2 is most abundant and the LPAT2 knockout mutant is lethal at the female gametophyte developmental stage [26]. By comparison, the single mutant for LPAT3, LPAT4, or LAPT5 exhibits no overt phenotype [27]. These results suggest that LPAT2 is a primary LPAT in the eukaryotic pathway. Recent studies showed that LPAT2 is also involved in phosphate starvation [29], whereas LPAT4 and LPAT5 function redundantly in nitrogen starvation response [27]. Phosphatidic acid is an important signal molecule in osmotic stress [18,20]. However, the role of LPAT-derived PA in salt and drought stress is still unclear. In this study, we isolated a knockdown mutant for LPAT2 in rice that exhibited less performance in osmotic tolerance, accompanied by a reduced PA level under salt and water deficit conditions. Supplemented exogenous PA in growth medium recovered the osmotic sensitive

Discussion
Of three major sources of PA, PLD-derived PA plays an important role in drought and salt stresses [31,32,34,36]. PLDα1-derived PA enhances stomatal closure under water deficit and ABA treatment in Arabidopsis [31][32][33]. PLDα3 knockout mutant contains a lower PA level and exhibits susceptibility to salt stress compared to WT plants [34]. The pldα1/δ double mutant plants with reduced PA level are more sensitive to salt stress than that of WT and the single mutant [35,36]. Phosphatidic acid produced from PLC/DGK also plays a role in osmotic stress response [38,39]. However, whether the de novo synthesized PA is involved in osmotic stress remains unknown. LPAT catalyzes a committed step in PA de novo synthetic pathway [25,26]. In Arabidopsis, LPAT1 is involved in the prokaryotic pathway of PA synthesis and is essential for embryo development [25,28]. Of the remaining four LPATs involved in the eukaryotic pathway, LPAT2 is most abundant and the LPAT2 knockout mutant is lethal at the female gametophyte developmental stage [26]. By comparison, the single mutant for LPAT3, LPAT4, or LAPT5 exhibits no overt phenotype [27]. These results suggest that LPAT2 is a primary LPAT in the eukaryotic pathway. Recent studies showed that LPAT2 is also involved in phosphate starvation [29], whereas LPAT4 and LPAT5 function redundantly in nitrogen starvation response [27]. Phosphatidic acid is an important signal molecule in osmotic stress [18,20]. However, the role of LPAT-derived PA in salt and drought stress is still unclear. In this study, we isolated a knockdown mutant for LPAT2 in rice that exhibited less performance in osmotic tolerance, accompanied by a reduced PA level under salt and water deficit conditions. Supplemented exogenous PA in growth medium recovered the osmotic sensitive phenotype of the lpat2 mutant to WT. Moreover, genetic complementation by LPAT2 was able to restore PA level of the lpat2 mutant to WT plants, and hence recovered the lpat2 mutant phenotype to WT. Thus, the results suggest that PA produced by LPAT2 activity plays an important role in salt and drought stress response in rice.
Abscisic acid is a key phytohormone in promoting stomatal closure to prevent water loss under salt stress and water deficit [2,40]. Two key players, snRK2 and ABI1 (PP2C), exhibit opposite actions in ABA signaling. In the absence of ABA, SnRK2 functions are restricted by ABI1, whereas upon osmotic stress, ABA is produced and binds to ABA receptor PYR1/PYL to relieve ABI1 inhibition. Subsequently SnRK2 was activated to phosphorylate the downstream substrates, thereby promoting stomatal closure to reduce water transpiration [15]. PLD-derived PA is known to play an important role in ABA signaling. PLDα1-derived PA is required for the effects of ABA on stomatal movement through interaction with ABI1 (PP2C), a negative regulator of ABA responses [31,32]. Moreover, PLDα1-derived PA binds to NADPH oxidase to mediate ABA-induced reactive oxygen species (ROS) and nitric oxide (NO) in guard cells in Arabidopsis [41]. Meanwhile, PA is accumulated and increases ABA sensitivity to regulate seed germination [42]. PLD-derived PA can mimic the effects of ABA, including the inhibition of α-amylase production in aleurone cells and suppression of seed germination [43]. A reduced PA resulted from double knockout mutation of PLDα1 and PLDδ renders seeds less sensitive to ABA inhibition of seed germination [44]. Our results showed that LPAT2-derived PA in osmotic stress tolerance is mediated by ABA response, as reduced ABA sensitivity of the lpat2 mutant was restored to WT by either LPAT2 genetic complementation or exogenous PA application. The results suggest that PA derived from different sources may exhibit similar function in osmotic stress response. Further studies are needed to investigate how LPAT2-derived PA mediates ABA response.
PA functions as a signal at upstream, a small change of PA level may lead to a significant impact on physiological effects [20]. Transcriptomic analysis revealed that suppressed LPAT2 led to a large number of DEGs under salt stress. In response to salt stress, downregulated DEGs in the lpat2 mutant included genes related to ABA response, Myb and NAC transcript factors, protein kinases, zinc finger proteins, transports of ions and ATP. By contrast, the expression levels of genes involved in the syntheses of ABA, JA, and glycolysis, were upregulated in the lpat2 mutant relative to WT under salt stress. The results suggest that LPAT2 affects the transcript levels of genes responsible for ABA signaling, hormone synthesis, ion transports, and protein kinases. It was shown that protein kinases play central roles in abiotic stress in plants [2]. PLD-derived PA interacts with mitogen-activated protein kinase 6 (MPK6) to regulate salt stress tolerance [45]. Our results showed that many genes encoding protein kinases were transcriptionally altered with most downregulated in LPAT2-suppressed mutant. The results suggest that LPAT2-derived PA has an effect on the transcript levels of protein kinases under salt stress. It was shown that stress-activated protein kinases such as SnRK2s play important roles in transcriptional regulation through the phosphorylation of the transcription factor [46]. PA is an important mediator in ABA signaling. Most DEGs are probably caused by the indirect effects of LPAT2derived PA. Nevertheless, PA was found to interact with several transcription factors in different cellular processes. PLDζ-derived PA binds to werewolf (WER), a R2R3 MYB transcription factor (TF), to module root hair formation and elongation [47]. PA interacts with two MYB TFs, late elongated hypocotyl (LHY) and circadian clock associated 1 (CCA1), to regulate lipid metabolism and seed oil accumulation [48]. PA also binds to AT-hook motif-containing nuclear localized (AHL) transcription factor to regulate triacylglycerol (TAG) degradation and seedling growth [49]. It is possible that LPAT2-derived PA is also involved in transcriptional regulation through binding to a new transcription factor in osmotic stress response. It will be of interest to identify new target(s) of LPAT2-derived PA that is directly responsible for transcriptional regulation in response to salt and water deficit stress in future studies.

Plant Growth and Treatments
The rice mutant and the genetic complementation (COM) were derived from the rice Zhonghua 11 (Oryza sativa L.). Wild type (WT) plants with same genetic background were used as a control. Rice plants were grown in the experimental field of Huazhong Agricultural University (Wuhan, China) under natural conditions during the seasons of summer and fall for genotype identification and seed collection. For stress treatments, seven-day-old seedlings were transferred to the pots containing soil (paddy soil:histosols = 1:1, v/v) and watered regularly until the plants were ready for treatments. For salt treatments, threeweek-old plants were treated with various concentrations (0, 100, 150 mM) of NaCl solution for various time points in growth room under the conditions of 14 h light (30 • C)/10 h dark (26 • C), photosynthetic photon flux density of 450-600 µmol·m −2 ·s −1 , and 50% relative humidity or under natural conditions during summer/fall seasons. For water deficit treatment, watering was stopped for three-week-old plants until leaves of WT plants began wilting when the soil moisture content was approximately 11% in the growth room under the conditions shown above. The soil moisture content was measured according to the method described by Mishra et al. [50]. Briefly, fresh soil was sampled from the pots under the control and water deficit conditions. The soil was weighed (W1) and dried at 60 • C in an oven. After drying, the soil was reweighed (W2) and moisture content was calculated by the following formula: % soil moisture = 100 × ((W1 − W2)/W2). For PA and ABA treatments, the seeds were germinated in small plastic cups supplemented with 10 µM ABA, 100 mM NaCl, 10% PEG without or with 20 µg PA based on previous methods [21,34,38]. Seed germination rate was scored by taking radical emergence as a criterion for germination.

RNA Extraction and Quantitative Real Time PCR
Total RNA was extracted from various tissues by using Trizol reagent (TransGene Biotech, Beijing, China). The contaminated DNA was removed by RNase free DnaseI. The first strand cDNAs were synthesized from mRNA templates by using TransScript cDNA Synthesis SuperMix kit (Trans-Gene Biotech, Beijing, China). The cDNA concentration was adjusted to a similar level based on the Ct value for β-actin (LOC_Os01g16414) by quantitative real-time PCR. The resultant cDNAs were used for quantitative real-time PCR monitored by a MyIQ real time PCR system (Bio-Rad, Hercules, CA, USA) by using Top Green qPCR SuperMix (TransGene Biotech, Beijing, China). Real-time PCR was run under the following conditions: 95 • C for 1 min; 40 cycles of 9 5 • C for 30 s, 55 • C for 30 s, 72 • C for 30 s. The primers for quantitative real-time PCR are listed in Supplementary Table S1.

Mutant Isolation and Genetic Complementation
A T-DNA insertion mutant in LPAT2, designated lpat2, was identified from the T-DNA insertion collections, and the mutant seeds were obtained from the Huazhong agricultural University (RMD; http://rmd.ncpgr.cn/, accessed on 3 March 2010). The homozygous mutant lpat2 was isolated by using the T-DNA left border primer coupled with LPAT2specific primers. For genetic complementation of the lpat2 mutant, the native LPAT2 promoter (2000 bp) was amplified from rice genomic DNA by using primers COMF and COMR with cutting sites of EcoR I and KpnI. The LPAT2 CDS was cloned from cDNA synthesized from rice leaf mRNA by using LPAT2 FK and LPAT2 Rb primers containing cutting sites of KpnI and BamHI. The resulting fragments were ligated to the pCAMBIA2301 vector. The construct was introduced into Agrobacterium tumefaciens EHA105, which was used to infect the lpat2 mutant calli to generate LPAT2 genetic complementation plants. The primers for real-time PCR are listed in Supplementary Table S1.

Measurements of Ion Leakage and Malonaldehyde
For ion leakage, the leaves of the same size, position, and age were excised from plants. After washing briefly with deionized water, the leaf samples were immersed in 15 mL distilled water in glass tubes. Gas was removed from the leaf surface in a vacuum for 30 min subsequently and incubated with gentle agitation for 3 h. The initial conductivity was measured by using a conductivity meter, and then the samples were boiled in a water bath for 20 min. Total conductivity was measured again after cooling to room temperature. Ion leakage was calculated based on a percentage of the initial conductivity versus the total conductivity.
The malonaldehyde (MDA) content of rice leaves was measured based on Zhao et al. [52]. Briefly, leaves (0.5 g per sample) were homogenized and incubated with 10% (w/v) trichloroacetic acid containing 0.6% (w/v) thiobarbituric acid at 95 • C for 30 min. The solution was cooled to room temperature and centrifuged at 12,000× g for 5 min. The absorbance of the supernatant was measured at 450, 532, and 600 nm, respectively. The MDA content was calculated according to the following equation: 6.45 × (OD 532 − OD 600 ) − 0.559 × OD 450 .

Subcellular Localization
To investigate the subcellular localization of LPAT2, the coding sequence of LPAT2 was cloned into the pMDC83 vector in-frame with the C-terminal fusion to GFP. The construct was transformed into Agrobacterium tumefaciens (EHA105) and the transformed Agrobacterium was grown in liquid LB media until OD 600 reached 0.6, and centrifuged at 4000 rpm for 10 min. The cells were resuspended in the solution (50 mM MES pH 5.7, 10 mM MgCl 2 , 150 mg/mL acetosyringone) and infiltrated the leaves of four-week-old tobacco (Nicotiana benthamiana) [53]. Five days after infection, GFP fluorescence was observed with a Lecia TCS SP2 confocal microscope with GFP excited at 488 nm and imaged at 500 to 530 nm, and RFP excited at 540 nm and imaged at 575 to 625 nm.

LPAT2 Expression and Activity Assay
The full-length of LPAT2 CDS was amplified by PCR from cDNA that was synthesized from mRNA of rice leaves by using LPAT2-ProF and LPAT2-ProR primers containing EcoRI and XhoI cutting sites (Supplementary Table S1). The CDS was ligated in frame into the pET28a expression vector. The resultant construct containing LPAT2 was transformed into Escherichia coli BL21 cells for LPAT2 expression. The LPAT2 activity was assayed in a reaction mixture (50 mM Tris HCl, pH 7.5, 0.5 M MgCl 2 , lysoPA, acyl-CoA) in a final volume of 100 µL with 100 µg proteins. After incubation at 37 • C for 1 h, the reaction was stopped by addition of 1 mL of chloroform/methanol (v/v, 2:1) and 250 µL of 1 N HCl. The organic phase was dried under a stream of nitrogen and dissolved in 20 µL of chloroform. The resultant lipids of reaction were separated on a thin-layer chromatography (TLC) plate (Merck, TLC silica gel 60) by developing a solvent of petroleum ether: diethyl ether: acetic acid (50:50:1, v/v). Finally, TLC plate was exposed to iodine gas to visualize PA. The spots corresponding to PA were quantified by ImageJ software (1.48, National Institutes of Health, Bethesda, MD, USA).

Lipid Extraction and Analysis
Lipids were extracted as a described method [54]. Briefly, leaves were sampled and immediately immersed in 5 mL of isopropanol containing 0.01% butylated hydroxytoluene (BHT, Sigma-Aldrich, St. Louis, MO, USA) at 75 • C for 15 min. After cooling to room temperature, the solvent (3.5 mL) of chloroform/water (2.5:1, v/v) was added to the sample and incubated for 1 h at room temperature by shaking. The extracts were transferred to clean glass tubes, and then the remaining samples were re-extracted with 5 mL chloroform/methanol (2:1, v/v) containing 0.01% BHT at room temperature for several times until leaf tissues became bleached. The extracts from the same sample were combined and washed twice with 1 M KCl and once with water, and dried under a stream of nitrogen gas, and then dissolved in a defined volume of chloroform for analysis by TLC combined with GC measurement. For glycerolipid quantification, the lipids were separated on TLC by using the developing solvent composed of acetone, toluene, and water (91:30:7.5, v/v) [55]. The spots in the TLC plate visualized with iodine were scraped out and methylated in methanol containing 1% H 2 SO 4 and 0.05% BHT at 80 • C for 30 min, and then quantitatively analyzed by GC measurement by using glyceryl triheptadecanoate (Sigma-Aldrich, St. Louis, MO, USA) as the internal standard. The GC was run under the following conditions. The temperature for the injection port was set at 180 • C with the pressure at 16.06 psi and the split ratio of 20:1. The oven temperature was initiated at 180 • C for 2 min and then was increased by 10 • C/min up to 220 • C for 5 min. The temperature of the flame ionization detector is 280 • C with the flow rate of 30, 300, and 25 mL/min for hydrogen, air, and helium, respectively. PA was detected based on previous methods [56]. Briefly, the TLC plates were dried in an oven at 110 • C for 1 h and then cooled to room temperature. The lipids were separated on TLC by using developing solvent (chloroform: ethanol: triethylamine: water = 10:11.3:11.7:2.7, v/v). The spots corresponding PA were visualized with 0.4% bromothymol blue solution containing 10 mM NaOH and quantified by ImageJ software (1.48, National Institutes of Health, USA).

RNA Sequencing and Analysis
The quality and integrity of RNA extracted from leaves were assessed by using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). A total RNA (1 µg RNA/sample) was used as input material for the RNA sample preparations. Briefly, mRNA was purified from total RNA by using poly-T oligoattached magnetic beads and was fragmentated by using divalent cations under elevated temperature. First strand cDNA was synthesized by using random hexamer primer and M-MuLV reverse transcriptase (RNase H-) and the second strand cDNA was synthesized by using DNA polymerase I and RNase H. The library fragments with preferential length of 370~420 bp were selected by AMPure XP system (Beckman Coulter, Beverly, MA, USA). The library quality was assessed on the Agilent Bioanalyzer 2100 system. Sequencing of cDNA was performed on an Illumina Novaseq platform. The read number of a gene was counted by using FeatureCounts v1.5.0-p3. The fragments per kilobase of transcript per million mapped reads (FPKM) of a gene was calculated based on the length of this gene. Differential expressed gene (DEG) analysis was performed by using the DESeq2 R package (1.20.0). The resulting p-values were adjusted by using the Benjamini and Hochberg approach for controlling the false discovery rate. Genes with an adjusted p-value < 0.05 found by DESeq2 were assigned as differentially expressed.

Conflicts of Interest:
The authors declare no conflict of interest.