Arabidopsis PCaP2 Functions as a Linker Between ABA and SA Signals in Plant Water Deficit Tolerance

Water stress has a major influence on plant growth, development, and productivity. However, the cross-talk networks involved in drought tolerance are not well understood. Arabidopsis PCaP2 is a plasma membrane-associated Ca2+-binding protein. In this study, we employ qRT-PCR and β-glucuronidase (GUS) histochemical staining to demonstrate that PCaP2 expression was strongly induced in roots, cotyledons, true leaves, lateral roots, and whole plants under water deficit conditions. Compared with the wild type (WT) plants, PCaP2-overexpressing (PCaP2-OE) plants displayed enhanced water deficit tolerance in terms of seed germination, seedling growth, and plant survival status. On the contrary, PCaP2 mutation and reduction via PCaP2-RNAi rendered plants more sensitive to water deficit. Furthermore, PCaP2-RNAi and pcap2 seedlings showed shorter root hairs and lower relative water content compared to WT under normal conditions and these phenotypes were exacerbated under water deficit. Additionally, the expression of PCaP2 was strongly induced by exogenous abscisic acid (ABA) and salicylic acid (SA) treatments. PCaP2-OE plants showed insensitive to exogenous ABA and SA treatments, in contrast to the susceptible phenotypes of pcap2 and PCaP2-RNAi. It is well-known that SNF1-related kinase 2s (SnRK2s) and pathogenesis-related (PRs) are major factors that influence plant drought tolerance by ABA- and SA-mediated pathways, respectively. Interestingly, PCaP2 positively regulated the expression of drought-inducible genes (RD29A, KIN1, and KIN2), ABA-mediated drought responsive genes (SnRK2.2, -2.3, -2.6, ABF1, -2, -3, -4), and SA-mediated drought responsive genes (PR1, -2, -5) under water deficit, ABA, or SA treatments. Taken together, our results showed that PCaP2 plays an important and positive role in Arabidopsis water deficit tolerance by involving in response to both ABA and SA signals and regulating root hair growth. This study provides novel insights into the underlying cross-talk mechanisms of plants in response to water deficit stress.


INTRODUCTION
Water deficit is one of the most acute abiotic stresses affecting plant growth and the economic yield of crop plants. It leads to alterations in various cellular processes in plants, for example, gene expression, photosynthesis, protein synthesis, carbon partitioning, lipid metabolism, and osmotic homeostasis (Hua et al., 2012;Jarzyniak and Jasiński, 2014;Fleta-Soriano and Munné-Bosch, 2016). Phytohormones, such as abscisic acid (ABA), salicylic acid (SA), gibberellin (GA), indole-3-acetic acid (IAA), and jasmonic acid (JA) function as central factors that link and reprogram these complex cellular processes (Hua et al., 2012).
Among the 10 SnRK2s in Arabidopsis, SnRK2.2, SnRK2.3, and SnRK2.6 function as central regulators in response to ABA and drought. Genetic analysis has shown that the Arabidopsis triple mutant snrk2.2/snrk2.3/snrk2.6 exhibits greatly reduced water deficit tolerance and is extremely insensitive to ABA (Fujii and Zhu, 2009;Fujita et al., 2009). The triple mutant is strongly impaired in ABA-and drought-responsive genes expression under water stress . The phenotype of the triple mutant indicates that these three SnRK2s are partially redundant, although all of them are crucial for plants' response to water stress and ABA, as well as ABA-mediated seed germination and dormancy. Seed dormancy, germination, and seedling growth of snrk2.2/snrk2.3 mutants are greatly insensitive to exogenous ABA. In contrast, snrk2.6 shows a significant increase in leaf water loss after ABA treatment (Yoshida et al., 2002;Fujii et al., 2007;Nakashima et al., 2009). This could be because SnRK2.6/ OST1 are important for stomatal movements, as they phosphorylate anion (SLAC1) and cation (KAT1) channels, which might be required for ABA-dependent stomatal closing in response to water deficit (Pilot et al., 2001;Geiger et al., 2009Geiger et al., , 2010Lee et al., 2009;Sato et al., 2009). In addition, guard cells (GCs) display transcriptional memory in a daily dehydration stress and watered recovery cycle. SnRK2.2, SnRK2.3, and SnRK2.6 have distinguishable roles in the process: SnRK2.2 and SnRK2.3 are more important for implementing guard cell stress memory, while SnRK2.6 is more important for overall stomatal control in the subsequent dehydration response (Virlouvet and Fromm, 2015). It has been found that SnRK2.2, -2.3, and 2.6 are regulated by nitric oxide (NO), phosphatidic acid (PA), and Ca 2+ changes (Boudsocq et al., 2004;Fujii and Zhu, 2009;Kulik et al., 2011), suggesting SnRK2s may be regulated by complex pathways during water stress; however, such mechanisms remain largely unknown.
Arabidopsis microtubule-associated protein-18/plasma membrane-associated Ca 2+ -binding protein-2 (MAP18/PCaP2) is important for several physiological activities. For example, it is involved in Ca 2+ binding and the organization of cortical microtubules (MTs) and F-actin. It also has a critical role in root hair, pollen tube growth, and directional cell growth Kato et al., 2010Kato et al., , 2013Zhu et al., 2013;Zhang et al., 2015;Kang et al., 2017). For example, the cell polarity and cortical microtubule array in line 2 of MAP18-overexpressing Arabidopsis (OE2) and line 18 of MAP18 RNAi transgenic Arabidopsis (R18) are altered . The T-DNA insertion line map18 (SALK_021652), which is confirmed as a knock-down mutant by qRT-PCR analysis, displays abnormal pollen tube growth and root hair growth (Kato et al., 2013;Zhu et al., 2013;Zhang et al., 2015;Kang et al., 2017). Interestingly, the mRNA expression level of PCaP2 is induced by heat, cold, drought, ABA, SA, osmotic stress, and GA3 (Kato et al., 2010(Kato et al., , 2013. This implies that it may function in response to abiotic stress and phytohormone signals. Root hairs are the main sites of water absorption in plants, which is important for water deficit tolerance (Worrall and Roughley, 1976;Zahran and Sprent, 1986;Schnall and Quatrano, 1992). Thus, we hypothesized that PCaP2 might be an important regulator of plant water deficit tolerance in various pathways, such as phytohormone signals and root hair growth, suggesting that PCaP2 might be a cross-talker between complex mechanisms involved in plant water deficit tolerance. In this study, we found that PCaP2 is vital for plant water deficit tolerance by responding to ABA and SA signals, regulating the expression of the key ABA-and SA-mediated genes, and affecting root hair growth. Collectively, our data provide novel evidence of the underlying complex mechanisms, especially of crosstalk between ABA and SA signaling pathways in plant water deficit tolerance.

The Expression of PCaP2 Is Highly Induced in All Tissues in Water Deficit
To fully understand the expression pattern of PCaP2 under water deficit stress, we examined the expression of PCaP2 in more details by quantitative real-time PCR (qRT-PCR) and β-glucuronidase (GUS) staining. Firstly, the wild type (WT) plants were exposed to dehydration conditions for 1, 3, 6, 9, and 12 h. The qRT-PCR results showed that PCaP2 expression was highly induced by water deficit treatments from 1 to 12 h, with the peak level of 10-fold increased at 6 h treatment ( Figure 1A). Furthermore, GUS staining showed that the promoter activity of PCaP2 was significantly increased after water deficit treatments for 6 h which was consistent with the results of qRT-PCR ( Figure 1B) and PCaP2 expression was induced in primary roots and lateral roots, cotyledons, true leaves, and the whole seedlings after water deficit treatments (Figures 1B-D).

PCaP2-OE Plants Display Increased Tolerance While PCaP2 RNAi and Mutant Seedlings Are Hypersensitive in Response to Water Deficit
To elucidate the function of PCaP2 in plant tolerance to water deficit, the previous identified one PCaP2 overexpression (PCaP2-OE) line , one knockdown of T-DNA insertion (pcap2) line (SALK_021652; Kato et al., 2013;Zhu et al., 2013;Zhang et al., 2015;Kang et al., 2017), and one fully silenced PCaP2 RNA interference (PCaP2-RNAi) line  were used. The PCaP2 expression of these lines was analyzed by qRT-PCR which is consistent with the previous publications (Supplementary Figure S1). We firstly investigated their seed germination rates under normal or drought conditions. Under normal conditions, germination rates of PCaP2-RNAi and pcap2 seeds were lower than those of PCaP2-OE and WT seeds at 1 day, and then these seeds gradually showed similar germination rate. At 4 and 5 days, the germination of these seeds was identical (Figure 2A). Water deficit significantly inhibited the germination of all seeds at 1 day, germination rates of PCaP2-RNAi were lower from 1 to 4 days, then the germination of these seeds was identical germination at 5 days (Figure 2A).
Next, the growth and survival status of the four-day-old seedlings were observed in water deficit treatments. It was found that the grown status showed no significant differences among WT, PCaP2-OE, pcap2, and PCaP2-RNAi lines under normal conditions. The PCaP2-OE seedlings showed longer roots and larger leaves than WT while pcap2 and PCaP2-RNAi showed opposite growth phenotypes under water deficit (Figures 2B,C). The results of survival status after water deficit treatments showed that PCaP2-OE plants were more tolerant in response to water deficit than WT (Figures 2C,D) while pcap2 and PCaP2-RNAi plants exhibited sensitivity to water deficit. To determine recovery after water deficit, we re-watered these seedlings for 7 days. The results showed that PCaP2-OE plants, WT, and some pcap2 seedlings were recovered while all PCaP2-RNAi seedlings exhibited water deficit sensitivity ( Figure 2D). The changes of leaf water loss and root water absorption in WT, PCaP2-OE, pcap2, and PCaP2-RNAi lines under water deficit conditions were observed. The results showed that pcap2 and PCaP2-RNAi lines lost water faster and wilted earlier than WT in dehydration stress while PCaP2-OE seedlings showed opposite phenotypes (Figures 2E,F). Additionally, the pcap2 and PCaP2-RNAi lines displayed shorter root hairs and less relative water content than WT in normal condition, the phenotypes were more significant in water deficit stress (Figures 2G-I). These results indicated that overexpression of PCaP2 enhanced the tolerance of plants to water deficit and down of PCaP2 led to plant water deficit hypersensitivity.
The PCaP2 Expression Is Highly Induced in Response to Exogenous ABA and SA Treatments Abscisic acid and SA are important regulatory signals in water deficit stress, and the higher expression level of PCaP2 has been found in exogenous 100 µM ABA and 100 µM SA treatments (Kato et al., 2010;Trivedi et al., 2016). To determine PCaP2 in ABA and SA signaling pathways, the expression of PCaP2 was examined in exogenous ABA and SA treatments by qRT-PCR and GUS staining assay in more details. Because the expression of many genes in Arabidopsis can be induced by 10-100 µM ABA Kato et al., 2010;Tian et al., 2015) and we found the expression of PCaP2 is higher in 40 µM ABA treatments than that in 100 µM ABA treatment ( Figure 3A; Kato et al., 2010). Thus, 40 µM ABA and 100 µM SA treatments were used in our studies. The results showed that the PCaP2 mRNA level was significantly increased with ABA and SA treatments from 1 to 12 h, the peaks appeared at 6 h ABA and 3 h SA treatments (Figures 3A,B). Fourteen-day-old WT seedlings were treated with water deficit for 0, 1, 3, 6, 9, and 12 h, and the expression of PCaP2 was detected by qRT-PCR. Data represent mean values of three biological replicates ± SE. The significant difference was determined by ANOVA in comparison to 0 h: * * P < 0.01, * * * P < 0.001. (B) Comparison of PCaP2 expression levels treated with water deficit different times by GUS staining. Seven-day-old seedlings of pPCaP2::GUS transgenic plants were treated with water deficit for 0, 1, 3, 6, 9, and 12 h. Scale bar = 2.5 mm. (C, D) Analysis of the expression pattern of PCaP2 by GUS staining of pPCaP2::GUS transgenic seedlings in normal (without water deficit treatment) (C) and water deficit treatments for 6 h (D). At least 15 seedlings from each sample were used for every technical replicate and three biological replicates were conducted. The scale bar is 5 mm in the pictures of whole plant, and 0.25 mm in the pictures of cotyledons, true leaves, hypocotyls, lateral roots and main roots.
The expression level of PCaP2 was higher in 40 µM ABA than 100 µM SA treatments (Figures 3A,B). Further, the PCaP2 expression was induced in the whole seedlings, cotyledons, true leaves, hypocotyls, primary roots, and lateral roots by ABA and SA treatments (Figures 3B-E), which were consistent with the results from water deficit treatments.

PCaP2-OE Plants Show Increased Tolerance, While Its RNAi and Mutant Are More Sensitive in Exogenous ABA and SA Treatments
Then we tested the germination rate of seeds and seedling growth of WT, PCaP2-OE, pcap2, and PCaP2-RNAi lines under normal, ABA, and SA treatments. To choose the most appropriate concentrations for ABA and SA treatments, 0.01-40 µM of ABA and SA were tested in the previous experiments used (data not shown). The results showed WT and PCaP2-OE showed the similar germination rates. The PCaP2-RNAi and pcap2 showed much higher germination rates in 0.8 µM ABA and 0.3 mM SA treatments, compared to WT and PCaP2-OE ( Figure 4A). The germination rates of PCaP2-RNAi and pcap2 seeds were lower than those of PCaP2-OE and WT seeds in 1-to-5-day ABA and in 1-to-3-day SA treatments. The germination rates of these seeds were similar at day 4 of SA treatment and identical at day 5 of SA treatment. The growth status of WT, PCaP2-OE, pcap2, and PCaP2-RNAi lines was significantly different under 0.5 µM. ABA and 0.05 mM SA treatments (Figures 4B,C). The PCaP2-OE seedlings showed larger leaf area and longer primary roots than WT while pcap2 and PCaP2-RNAi lines showed smaller leaves and shorter roots and root hairs, which were consistent with the water deficit inducible phenotypes (Figures 4B,C).

PCaP2 Positively Regulates the Expression of SnRK2 Genes and PR Genes Under Water Deficit Stress
To analyze whether the role of PCaP2 in water deficit tolerance was mediated by ABA and SA signaling pathways, we firstly checked the PCaP2 expression of WT, PCaP2-OE, pcap2, and . About 250 root hairs from 30 roots of each sample were used for every technical replicate and three biological replicates were conducted. (I) The relative water content (RWC) of WT, pcap2 and PCaP2-RNAi lines in 1/2MS or coated with 15% PEG. About 30 seedlings from each sample were used for every technical replicate and three biological replicates were conducted. All data are mean values of three biological replicates ± SE. The significant difference was determined by ANOVA: * P < 0.05, * * P < 0.01, * * * P < 0.001. PCaP2-RNAi in water deficit conditions by qRT-PCR. The analysis showed that the increasing expression of PCaP2 in PCaP2-OE and the decreasing expression in pcap2 and PCaP2-RNAi were compared with that in WT under water deficit treatments (Supplementary Figure S1).
Next, we examined the expression of the following ABA-responsive genes, including SnRK2.2, -2.3, -2.6, ABF2, -3, -4 (Choi et al., 2000;Uno et al., 2000;Boudsocq et al., 2004;Gong et al., 2015;Jia et al., 2015); drought-inducible genes, including RD29A (Yamaguchi-Shinozaki and Shinozaki, 1994); KIN1 and KIN2 (Kurkela and Borg-Franck, 1992); and SA-responsive genes such as PR1, -2, -5 genes (Liu P. et al., 2013;Liu W.X. et al., 2013) in WT, PCaP2-OE, pcap2 and PCaP2-RNAi with dehydration conditions at 0, 6, and 12 h by qRT-PCR FIGURE 4 | Effects of PCaP2 on Arabidopsis seed germination and seedling growth in ABA or SA treatments. (A) Quantification of germination rates of the PCaP2-OE, WT, pcap2, and PCaP2-RNAi seeds in 1/2 MS medium or mixed with 0.8 µM ABA and 0.3 mM SA from 1 to 5 days. A total of 40 seeds of each line were used for every technical replicate and three biological replicates were conducted. (B) Phenotypes of the PCaP2-OE, WT, pcap2, and PCaP2-RNAi seedlings grown in exogenous ABA and SA treatments. The seedlings were sown on 1/2 MS medium for 5 days, then transferred to 0.5 µM ABA or 0.05 mM SA medium and grew for 10 days. Scale bar = 1 cm. (C) The phenotype statistics of the PCaP2-OE, WT, pcap2, and PCaP2-RNAi seedlings grown in exogenous ABA and SA treatments. The main root length and leaf area of these seedlings were calculated after growth on 1/2 MS medium or supplemented with 0.5 µM ABA and 0.05 mM SA. At least 70 roots or 200 leaves from about 70 seedlings from each sample were used for every technical replicate and three biological replicates were conducted. All data are mean values of three biological replicates ± SE. The significant difference was determined by ANOVA: * P < 0.05, * * P < 0.01, * * * P < 0.001.
analysis. Under normal conditions, pcap2 and PCaP2-RNAi inhibited the expression of ABA-and drought-responsive genes, except SA-responsive genes (Figure 5), and PCaP2-OE increased the expression of SA-responsive genes, suggesting that PCaP2 might be the complex mechanisms in regulating gene expression in normal conditions. Under water deficit stress, the expression of all these genes, including ABA-, drought-, and SA-responsive genes, was significantly inhibited in pcap2 and PCaP2-RNAi while increased in PCaP2-OE, compared with that in WT, indicating that PCaP2 was involved in regulating drought-, ABA-and SA-responsive genes under water deficit stress.

PCaP2 Positively Regulates ABA-Mediated SnRK2s Expression and SA-Mediated PRs Expression in Water Deficit
To further analyze whether the effect PCaP2 on gene expression in response to water deficit was mediated by ABA and SA signaling pathways, we firstly checked the PCaP2 expression of WT, PCaP2-OE, pcap2 and PCaP2-RNAi in ABA and SA treatments by qRT-PCR. The analysis showed that the increase expression of PCaP2 in PCaP2-OE and the decreasing expression in pcap2 and PCaP2-RNAi were compared with that in WT FIGURE 5 | Effect of PCaP2 on the expression of ABA-, drought-and SA-responsive genes under water deficit. The relative gene expression of the 14-day-old seedlings of WT, PCaP2-OE, pcap2, and PCaP2-RNAi with water deficit treatments for 0, 6, and 12 h was determined by qRT-PCR analysis. Data represent mean values of three biological replicates ± SE. The significant difference was determined by ANOVA: * P < 0.05, * * P < 0.01, * * * P < 0.001. Figure S1). Next, we examined the expression of SnRK2.2, -2.3, -2.6 and PR1, -2, -5 genes in WT, PCaP2-OE, pcap2, and PCaP2-RNAi with ABA and SA treatments at 0, 6, and 12 h by qRT-PCR analysis. The results showed that the effect of PCaP2-OE, pcap2, and PCaP2-RNAi on affecting the ABA-responsive gene expression in ABA treatments, rather than SA treatments, were similar to that in drought treatments, (Figure 6A). In contrast, the effect of PCaP2-OE, pcap2, and PCaP2-RNAi on affecting the SA-responsive gene expression in SA treatments, rather than ABA treatments, were similar to that in drought treatments ( Figure 6B). These results illustrated that PCaP2 regulated the expression of ABA-mediated SnRK2 genes and SA-mediated PR genes under water deficit stress, indicating that PCaP2 was a crosstalk factor in response to ABA and SA signals in water deficit.

DISCUSSION
Water deficit significantly influences plant growth, development, and productivity. One of the most important regulators in drought is ABA, and some cases of SA-mediated plant water deficit tolerance have been documented. However, details of crosstalk between ABA and SA signaling cascades in response to water deficit remain largely unknown. Previous studies have suggested that PCaP2 responds to some phytohormone signaling pathways and abiotic stress through mRNA expression pattern analysis (Kato et al., 2010). Our results further support a key role for PCaP2 in Arabidopsis water deficit tolerance and suggest that it is connected to the main ABA and SA signaling pathways.
Previous studies have shown that PCaP2 is mainly expressed in roots and flower tissues in normal conditions, and its mRNA level in whole seedlings is induced by drought as well as treatment with ABA or SA Kato et al., 2010). Compared with previous studies, our data provide more detailed findings. High expression of PCaP2 in roots was found under normal and stress conditions. This might be related to its function in root water absorption and root hair growth in both normal and water deficit conditions. Additionally, PCaP2 was not expressed in true leaves under normal conditions, but was highly induced in water deficit and treatment with ABA or SA. This indicated that in true leaves, the increase of PCaP2 level is a key response to stress that is triggered by phytohormone signals, which may be associated with its function in controlling water deficit-induced leaf water loss.
Some ABA-sensitive seedlings show improved drought tolerance in soil, such as SnRK2s-OE and AREBs-OE, while some mutants show the same sensitivity to ABA and drought treatments, such as Atrgga and AtDi19-3 Qin et al., 2014;Ambrosone et al., 2015). Generally, ABAsensitive mutants improve drought tolerance by regulating stomatal movement. However, changes in the drought tolerance of Atrgga and AtDi19-3 are not dependent on stomatal movement Qin et al., 2014;Ambrosone et al., 2015). In the present study, we used the previously identified PCaP2-OE line  pcap2 mutant (Kato et al., 2013;Zhu et al., 2013;Zhang et al., 2015;Kang et al., 2017) and PCaP2-RNAi lines , and our analysis of PCaP2 expression of these lines is the same with the prior findings. Our results showed that pcap2 and PCaP2-RNAi lines were more sensitive to ABA and drought, while PCaP2-OE plants were more resistant to ABA and drought stress. Although our results do not address whether PCaP2 can regulate stomatal movement, it is possible that PCaP2 has a similar mechanism to AtRGGA and AtDi19-3. PCaP2 do not fully depend on the regulation FIGURE 6 | Effect of PCaP2 on the expression of SnRK2 and PR genes with exogenous ABA and SA. (A) The relative expression of SnRK2 genes and PR genes of the 14-day-old plants with 40 µM ABA treatments for 0, 6, and 12 h was determined in WT, PCaP2-OE, pcap2, and PCaP2-RNAi by qRT-PCR assay. (B) The relative expression of PR and SnRK2 genes of the 14-day-old plants with 100 µM SA treatments for 0, 6, and 12 h was determined in WT, PCaP2-OE, pcap2, and PCaP2-RNAi by qRT-PCR assay. Data represent mean values of three biological replicates ± SE. The significant difference was determined by ANOVA: * P < 0.05, * * P < 0.01, * * * P < 0.001. of stomatal movement in response to drought stress; PCaP2 functionality in plant water deficit tolerance can be borne out by regulating root hair growth to benefit water absorption and to induce the expression of numerous key drought-responsive genes (Figure 7). Furthermore, the water deficit-inducible phenotypes in fully silenced PCaP2-RNAi plants were greater in magnitude to those in knockdown pcap2 mutants. This suggests that the role of PCaP2 in water deficit is dependent on its inducible expression level. These results indicate that PCaP2 is a positive regulator of water deficit tolerance in Arabidopsis.
Root hairs are the major water absorbing tissue in plants. Under water-limited conditions, root hair growth is significantly inhibited, which affects root water absorption and plant water deficit tolerance (Worrall and Roughley, 1976;Zahran and Sprent, 1986;Schnall and Quatrano, 1992). Our results showed that reduced expression of PCaP2 leads to shorter root hairs under normal conditions, which is consistent with previous results from Zhang et al. (2015). Moreover, significant inhibition of shorter root hair development was seen under water deficit conditions in pcap2 and PCaP2-RNAi, which could be one main pathway through which PCaP2 induces plant water deficit tolerance.
Compared to the ABA signaling pathway, relatively few cases of SA-mediated drought tolerance have been presented,  and Kang G.Z. et al. (2012) identified 76 proteins as potentially involved in the SA signaling pathway in drought-exposed T. aestivum. Genetic and gene expression analysis of these proteins show that PR1, -2, and -5 are required in SA signaling in drought stress (Liu P. et al., 2013;Liu W.X. et al., 2013). Arabidopsis cpr5 and acd6 mutants exhibit SA accumulation, and their drought tolerance is improved by inducing SA-mediated expression of PR1, -2, and -5 (Liu P. et al., 2013). Additionally, PR1, -2, and -5 overexpression enhances drought tolerance in Arabidopsis (Liu W.X. et al., 2013;Qin et al., 2014).
Transcription factor Di19 regulates the expression of PR1, -2, and -5 in drought conditions (Liu W.X. et al., 2013;Qin et al., 2014). Water loss rates of PR1-, PR2-, and PR5-overexpressing lines are lower than those of WT plants, and the three genes are highly expressed in stomata, suggesting that PR1, PR2, and PR5 may regulate stomatal movement in response to drought tolerance (Liu W.X. et al., 2013;Qin et al., 2014). Additionally, PRs play significant roles in regulating phytohormone-signaling, such as auxin and JA (Thaler et al., 1999;Wang et al., 2001;Wen et al., 2008;Iglesias et al., 2011), suggesting that PR1, PR2, and PR5 may mediate the complex crosstalk between phytohormone signals under drought stress. However, complete mechanisms of PR actions under drought stress are largely unknown. Our results show that PCaP2 positively regulates the expression of PR1, -2, and -5 after SA and water deficit treatments, but not ABA treatment. In addition, PCaP2-RNAi and pcap2 lines lost water faster and wilted earlier than WT lines, which stands in contrast to the phenotypes of PR1-, PR2-, and PR5-overexpressing seedlings (Liu W.X. et al., 2013;Qin et al., 2014). However, the phenotypes of PCaP2-OE plants are consistent with those of PRs-overexpressing seedlings. Thus, these results support the idea that PCaP2 activates the key SA-mediated signaling pathway in response to water deficit, providing a new pathway to regulate PRs under water deficit conditions.
Interactions between SA and ABA signals have been shown to occur during abiotic stress, such as salt and cold stress. In S. lycopersicum, treatment with SA improves plant growth, osmotic adaptation, and ABA accumulation under normal conditions and during salt stress (Szepesi et al., 2009). Under cold conditions, exogenous ABA treatment increases endogenous SA level and oHCA content in Z. mays, suggesting that the ABA signal may combine with SArelated stress responses during cold stress (Szalai et al., 2011). Our results showed that ABA and SA can trigger PCaP2, and that PCaP2 then regulates the ABA-mediated SnRK2 signaling pathway as well as the expression of SA-mediated PR genes during water deficit stress. Thus, PCaP2 mediates crosstalk in response to ABA and SA signals during water deficit.
We have not interrogated whether PCaP2 can directly regulate SnRK2s and/or PRs in normal, water deficit, or exogenous ABA and SA conditions. However, previous studies have shown that PCaP2 can bind PtdInsPs, Ca 2+ and Ca 2+ /CaM complexes. These ligands are important components of intracellular signaling networks involved in plant water deficit tolerance (Zielinski, 1998;Meijer and Munnik, 2003;Carlton and Cullen, 2005;Perera et al., 2008;DeFalco et al., 2009;Kleerekoper and Putkey, 2009;Luan, 2009;Xue et al., 2009;Kato et al., 2010Kato et al., , 2013. Our results also showed that PCaP2 regulated the expression of many key ABA-responsive genes, SA-responsive genes, and drought-inducible genes, including upstream genes such as SnRK2.2, -2.3, and -2.6. Thus, it is well-accepted that PCaP2 mainly functions as a Ca 2+ -binding protein to participate in intracellular signaling transduction under water deficit stress. In conclusion, water deficit triggers ABA and SA accumulation, which, in turn, induce PCaP2 expression. PCaP2 then activates the expression of many key droughtresponsive genes, including ABA-mediated genes (SnRK2.2, -2.3, -2.6, ABF1, -2, -3, -4) and SA-mediated genes (PR1, -2, -5) and drought-inducible genes (RD29A, KIN1 and KIN2), and affects root hair growth to increase water absorption, which improve plant water deficit tolerance (Figure 7). This study highlights the positive and important role that PCaP2 plays in plant water deficit tolerance by mediating crosstalk between ABA and SA signals, providing novel evidence relating to the underlying and complex mechanisms mediated by ABA and SA signals during plant water deficit tolerance.

Plant Materials, Growth Conditions, and Treatments
Arabidopsis ecotype Col-0 was the background for all transgenic and mutant plants in this study. The seeds of PCaP2-OE line , pcap2 mutant (Kato et al., 2013;Zhu et al., 2013;Zhang et al., 2015;Kang et al., 2017), and PCaP2-RNAi line  and PCaP2 promoter::GUS (pPCaP2::GUS)  were provided by Professor Ming Yuan, China Agricultural University, Beijing, China. All these seeds were used in the previous reports.
Surface sterilized Arabidopsis seeds were sown on MS plates with 1/2 MS (Murashige & Skoog) medium (pH adjusted to 5.8-6.0 with 1 M NaOH) with vitamins, 0.6 % phytoagar (PlantMedia), and 1% sucrose, and then in darkness at 4 • C for 3 days. Seedlings were then transferred to a growth chamber at 22 • C, with a 16 h/8 h (light/dark) photoperiod at approximately 120 µmol m −2 s −1 . The 1/2 MS plus 1% sucrose media are usually used in water deficit, ABA, and SA treatments (Jayakannan et al., 2015;Schwalm et al., 2017;Zhu et al., 2017). PCaP2-OE, WT, pcap2, and PCaP2-RNAi were used for all experiments. For germination rate analysis, seeds were sown on 1/2 MS medium with 25 % PEG (dehydration), 0.8 µM ABA and 0.3 mM SA for 1, 2, 3, 4, 5 days, respectively. For seedling growth experiments, 4-day-old seedlings were transferred to soil for 7 days, then with 40 % PEG (dehydration), 0.5 µM ABA and 0.05 mM SA for 10 days. For survival experiments, 4-day-old seedlings were transferred to soil for 7 days, then were treated for 10 days without water, following with water for 7 days. At least 20 seedlings were harvested for observing in three biological replicates in each experiment. The best concentration of ABA and SA in the present paper was chosen by the prepared experiments for 0.01 to 40 µM of ABA and SA (data not shown), because the previous findings show that significantly different ABA and SA concentrations from 1 nM to 80 µM are used in seed germination and seedling growth experiments (Reyes and Chua, 2007;Zhu et al., 2007;Horváth et al., 2015;Huang et al., 2017). For waterloss assays, rosette leaves of comparable size from 4-week-old plants grown under long days were detached, placed on a Petri dish and weighted at a specified time (0, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, and 4 h) after detachment. For root hair growth assays, 3-day-old seedlings were transferred to 1/2 MS medium with or without 15% PEG for 3 days, and the root hair length was assayed by ImageJ software 1 . Three-day-old seedlings 1 http://rsb.info.nih.gov/ij/ transferred to 1/2 MS medium with or without 15% PEG for 7 days were used to test relative water content. Relative water content was measured as previous reports described (Brini et al., 2007).

RNA Isolation and Quantitative RT-PCR (qRT-PCR) Analysis
Total RNA was isolated from 1-day-old seedlings using RNasy plant mini kit (Qiagen), treated with RNase-Free DNase (Qiagen) at 37 • C for 1 h to degrade genomic DNA, and 1 µg total RNA to synthesize cDNA by oligo-(dT) 20-primed reverse transcription by the Omniscript RT Kit (Qiagen). The cDNA was amplified using SuperReal PreMix Plus (SYBR Green, TIANGEN, China) in a 10 µL volume. The expression levels of 18S rRNA was used as an internal control. Analysis was performed using the BioRad Real-Time System CFX96TM C1000 Thermal Cycler (Singapore).
For the expression of PCaP2 under drought, ABA, and SA treatment assays, 14-day-old seedlings from WT plants were exposed 1/2 MS medium with or without drought, 40 µM ABA and 100 µM SA for 0, 1, 3, 6, 9, and 12 h. The previous finding showed the higher expression level of PCaP2 has been found in exogenous 100 µM ABA and 100 µM SA treatments (Kato et al., 2010). The expression of many genes can be increased in 10 to 100 µM ABA treatments in Arabidopsis Kato et al., 2010;Tian et al., 2015;Zhu et al., 2017). We found the expression level of PCaP2 can be induced in 40 µM ABA treatments, and the expression is higher than that in 100 µM ABA treatments (Figures 3A; Kato et al., 2010). Thus 40 µM ABA was used in our experiments. At each time point, all seedlings were immediately frozen by liquid nitrogen and then stored at −80 • C for RNA preparation. To assay the expression of drought-, ABA-and SA-responsive genes under water deficit stress in WT and pcap2, WT and mutant seedlings grown at normal conditions for 14 days were harvested and treated by dehydration stress for 0, 6, 12 h. To assay the expression of SnRK2 and PR genes in WT and pcap2 treatment with ABA and SA, the seedlings of 14-day-old at normal condition were harvested and treated with 40 µM ABA and 100 µM SA for 0, 6, 12 h. Total RNA extraction and reverse transcription were performed as described above. All primer pairs used for qRT-PCR examination are listed in Supplementary Table S1. Each representative experiment was performed with at least three replicates.

Histochemical Staining of GUS Activity
Seven-day-old or 14-day-old pPCaP2::GUS seedlings treated with 40% PEG, 40 µM ABA, or 100 µM SA for 0, 1, 3, 6, and 12 h were collected for observing the changes of GUS activity. The GUS staining procedure was executed by the method mentioned in Wang et al. (2007). Samples were observed on an Olympus microscope equipped with a color CCD camera (Sutter Instrument; LAMBDA 10-2) or by an Epson scanner.

Statistical Analysis
The experiments were conducted at least three times, each of which contained three technical replicates. Data presented as the mean ± SE of three biological replicates. The significant difference was analyzed by SPSS statistical software (ver.16.0, SPSS Inc., Chicago, IL, United States) via one-way or twoway ANOVA ( * P < 0.05, * * P < 0.01, * * * P < 0.001). In addition, we tested the normality of data about the main root length, leaf area, and root hair length under drought stress with Lilliefors corrected K-S a test and Shapiro-Wilk test by SPSS (Supplementary Table S2). The detailed information of the chemicals and kits used in this study are listed in Supplementary  Table S3.

AUTHOR CONTRIBUTIONS
CW and XW designed the study. YW, LW, HL, BZ, and QC performed the experiments and data analysis. XL, SB, YL, and QW provided help in experimental methods. SZ participated in the discussion. MH and ST trained the use of experimental equipment. SY helped to revised the language and grammar. CW wrote the manuscript.