The Pepper CaOSR1 Protein Regulates the Osmotic Stress Response via Abscisic Acid Signaling

Plants are sessile organisms, and their growth and development is detrimentally affected by environmental stresses such as drought and high salinity. Defense mechanisms are tightly regulated and complex processes, which respond to changing environmental conditions; however, the precise mechanisms that function under adverse conditions remain unclear. Here, we report the identification and functional characterization of the CaOSR1 gene, which functions in the adaptive response to abiotic stress. We found that CaOSR1 gene expression in pepper leaves was up-regulated after exposure to abscisic acid (ABA), drought, and high salinity. In addition, we demonstrated that the fusion protein of CaOSR1 with green fluorescent protein (GFP) is localized in the nucleus. We used CaOSR1-silenced pepper plants and CaOSR1-OX-overexpressing (OX) transgenic Arabidopsis plants to show that the CaOSR1 protein regulates the osmotic stress response. CaOSR1-silenced pepper plants showed increased drought susceptibility, and this was accompanied by a high transpiration rate. CaOSR1-OX plants displayed phenotypes that were hypersensitive to ABA and hyposensitive to osmotic stress, during the seed germination and seedling growth stages; furthermore, these plants exhibited enhanced drought tolerance at the adult stage, and this was characterized by higher leaf temperatures and smaller stomatal apertures because of ABA hypersensitivity. Taken together, our data indicate that CaOSR1 positively regulates osmotic stress tolerance via ABA-mediated cell signaling. These findings suggest an involvement of a novel protein in ABA and osmotic stress signalings in plants.


INTRODUCTION
Drought, cold, and high salinity in the soil cause osmotic stress to plants, and limit plant growth and development and agricultural crop productivity (Zhu, 2002;Sengupta and Majumder, 2010). In the natural environment, growth conditions are rarely optimal and plants constantly encounter various osmotic stresses; these alter the water potential in plant cells and cause serious damage (Grondin et al., 2015). During exposure to osmotic stresses, plants exhibit many physiological and molecular changes such as reduction of water content, stomatal closure, alteration of cell growth, and inhibition of photosynthesis, to limit cellular damage and restore homeostasis (Zhu, 2002;Ambrosone et al., 2015). The cellular and physiological defense-related response to osmotic stress has been elucidated; however, the precise mechanisms underlying functional modification remain unclear, because of the complexity and diversity at the cellular level and the whole plant level (Ding et al., 2015;Lim et al., 2015a;Zou et al., 2015).
Abscisic acid (ABA) is a plant hormone that promotes the processes of adaptation to abiotic and biotic stresses (Gonzalez-Guzman et al., 2012;Lee and Luan, 2012;Lim et al., 2015a). ABA plays a primary role in the cellular defense response under osmotic stress conditions. The level of ABA is increased in response to osmotic stress, and this initiates the signal transduction pathway related to defense responses. In comparison with other plant hormones, ABA regulates a large number of genes; more than 10% of Arabidopsis genes are induced by ABA (Goda et al., 2008;Mizuno and Yamashino, 2008). The most well-known response is ABAmediated stomatal closure via the efflux of cations and anions from guard cells; this leads to decreased transpiration, which is vital for plant survival under osmotic stress conditions (Geiger et al., 2009;Lee et al., 2009). Hence, ABA-responsive and ABA-deficient mutants exhibit a phenotype that is tolerant to abiotic stress (Vlad et al., 2009;Umezawa et al., 2013;Zou et al., 2015). In contrast, ABA hypersensitivity is associated with abiotic stress tolerance (Lee et al., 2013;Lim and Lee, 2015). Moreover, a large number of genes associated with the defense response to osmotic stress are regulated by ABA (Shinozaki and Yamaguchi-Shinozaki, 2007;Lim et al., 2015b). Genetic and molecular analysis studies have identified many stress-related genes and transcription factors involved in defense responses. For example, ABA signal transduction components-from ABA receptors to SnRK2-type kinaseregulate ion channels and bZIP transcription factors (Lee and Luan, 2012;Lim et al., 2015a). In addition, transcription factorsincluding dehydration-responsive element binding protein/crepeat binding factor (DREB/CBF) (Lee et al., 2010;Lee and Thomashow, 2012;Ding et al., 2015), ABA binding factor (ABF) (Kim et al., 2004;Yoshida et al., 2015), and ethylene-responsive factor/Apetala2 (ERF/AP2) (Li et al., 2011;Dubois et al., 2015)activate different defense-related genes in response to osmotic stress, to be rebalanced and fine-tuned.
In the present study, we identified a novel pepper gene, CaOSR1 (Capsicum annuum Osmotic Stress Resistance 1). We isolated this gene from a pepper cDNA library, using an ABAinducible cDNA as a probe. We examined the molecular and phenotypic aspects of CaOSR1-silenced pepper and CaOSR1overexpressing (OX) transgenic Arabidopsis plants. We showed that the alteration of CaOSR1 expression influences drought stress tolerance via regulation of the transpiration rate and induction of stress-responsive genes. Our findings indicate that the CaOSR1 protein is a positive regulator of ABA signaling and osmotic stress tolerance.

Sequence Alignment and Phylogenetic Tree Analysis
The encoded amino acid sequences for CaOSR1 and its homologs were obtained using BLAST searches (http://www.ncbi.nlm.nih. gov/BLAST). The amino acid alignment was performed using ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2), and the results were edited using Genedoc software (http://www.nrbsc. org/gfx/genedoc). The amino acid alignments were manually regulated to compare the cDNA clones of CaOSR1 with those of other organisms. Based on the data of multiple sequence alignment, phylogenetic tree was drawn with MEGA software (version 5.2). To investigate sequence identity and similarity between two proteins, pairwise sequence alignment was performed using EMBOSS Needle webtool (http://www.ebi. ac.uk/Tools/psa/emboss_needle) with default parameter.

Virus-Induced Gene Silencing and Overexpression of CaOSR1
We used the tobacco rattle virus (TRV)-based virus-induced gene silencing (VIGS) system to generate CaOSR1 gene knockdown in pepper plants. We used a 1598-1805-bp fragment and the full length CaOSR1 cDNA to generate CaOSR1-silenced pepper plants and CaOSR1-overexpressing (OX) transgenic Arabidopsis plants, respectively, according to the protocol described previously (Park C. et al., 2015).

ABA, Drought, and NaCl Treatments
To examine the expression pattern of the CaOSR1 gene in pepper plants after ABA treatment, six-leaf-stage pepper plants were sprayed with 100 µM ABA or control solution. For the NaCl and drought treatments, pepper plants were irrigated with 200 mM NaCl solution and were then carefully removed from the soil to avoid injury. The plants were placed onto 3-mm filter paper (Whatman, Clifton, UK). Leaves were harvested at 0-24 h after each treatment and were subjected to RNA isolation and reverse transcription-polymerase chain reaction (RT-PCR) analysis.
To measure the rate of germination, root elongation and seedling establishment, 36 seeds each of wild-type and CaOSR1-OX transgenic Arabidopsis plants were stratified at 4 • C for 2 days and were then plated on 0.5× MS agar medium supplemented with various concentrations of ABA. The plates were incubated at 24 • C under white fluorescent light (130 µmol photons·m −2 ·s −1 ) with a 16-h light/8 h-dark cycle.
Three-week-old seedlings from wild-type and CaOSR1-OX transgenic Arabidopsis lines were randomly planted and were then subjected to drought stress treatment by withholding watering for 9 days and rewatering for 2 days. Survival rates were measured in each individual sample, and each experiment was performed three times with 20 plants. For pepper plants, drought stress was imposed on four-leaf-stage plants by withholding watering for 12 days. Plants were rewatered for 2 days to allow recovery, and the survival rate of the plants was then calculated. Survival rates were measured in each individual sample, and each experiment was performed three times with 20 plants. The drought resistance was determined in a quantitative manner by measuring the transpirational water loss. Fifty leaves were detached from four-leaf-stage pepper plants and 3-week old Arabidopsis plants and placed in Petri dishes. The dishes were maintained in a growth chamber at 40% relative humidity, and the loss of fresh weight was determined at the indicated time points. All the experiments were performed at least in triplicate.

Thermal Imaging
For thermal imaging analysis, 4-week-old pepper plants having full expanded 1st and 2nd leaves and 3-week-old Arabidopsis plants and were treated with 50 µM ABA. Thermal images were obtained using an infrared camera (FLIR systems; T420) and leaf temperature was measured by FLIR Tools + ver 5.2 software.

Stomatal Aperture Bioassay
To measure the stomatal aperture, epidermal peels were stripped from rosette leaves of 3-week-old plants and floated in a stomatal opening solution (SOS: 50 mM KCl and 10 mM MES-KOH, pH 6.15, 10 µM CaCl 2 ) in the light. After incubation for 3 h, the buffer was replaced with fresh SOS containing 20 µM ABA. After additional 2 h incubation, stomatal apertures were measured in each individual sample, and each experiment was performed three times with 20 leaves.

RNA Isolation and Semi-Quantitative and Quantitative Reverse Transcription-Polymerase Chain Reaction
Total RNA was isolated from the Arabidopsis leaf tissues, which were dehydrated or infected with the bacterial pathogen using an RNeasy Mini kit (Qiagen, Valencia, CA, USA). To remove genomic DNA, all RNA samples were digested with RNA-free DNase. After quantification using a spectrophotometer, 1 µg of total RNA was used to synthesize cDNA using a Transcript First Strand cDNA Synthesis kit (Roche, Indianapolis, IN, USA) according to the manufacturer's instructions. Concomitantly, cDNAs were synthesized without reverse transcriptase and were subjected to semi-quantitative RT-PCR to rule out the possibility of contamination by genomic DNA in the cDNA samples. For quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis, the synthesized cDNA was amplified in a CFX96 Touch TM Real-Time PCR detection system (Bio-Rad) with iQ TM SYBR Green Supermix and specific primers (Supplementary Table 1). Every reaction was performed in triplicate. The PCR was programmed as follows: 95 • C for 5 min; 45 cycles each at 95 • C for 20 s and 60 • C for 20 s; and 72 • C for 20 s. The relative expression of each gene was calculated using the Ct method, as previously described (Livak and Schmittgen, 2001). The Arabidopsis actin8 gene (AtACT8) was used for normalization.

Statistical Analyses
To determine significant differences between genotypes in response to treatments, statistical analyses were performed using one way analysis of variance (ANOVA) or student's t-test. A P < 0.05 was considered significant difference.

Isolation and Sequence Analysis of the Pepper CaOSR1 Gene
We used differential hybridization analysis to isolate the pepper CaOSR1 (Capsicum annuum Osmotic Stress Resistance 1) gene from a cDNA library constructed from ABA-treated pepper leaf tissues (Lim et al., 2014). Among the gene clones, we selected genes that were upregulated by ABA (data not shown). The putative CaOSR1 consists of a 2643-bp open reading frame, and the predicted CaOSR1 encodes 880 amino acid residues (Figure 1) The mature protein has a molecular weight of 93,865 Da and an isoelectric point of 5.65. The results of multiple sequence alignment analysis and the phylogenetic tree showed that CaOSR1 (accession no. KT693385) is clustered into the same clade with five low-temperature-induced 65 kDa (LTI65) proteins from the family Solanaceae ( Figure 1A) of which LTI65 protein of Solanum tuberosum (accession no. XP_006353392.1) shares highest identity (60.9%) and similarity (64.3%) withCaOSR1. Although sequence homology is <35%, CaOSR1 shares sequence homology with three proteins from Arabidopsis such as RD29B (LTI65; At5g52300), CAP160 (At4g25580), and RD29A (COR78; At5g52310). In particular, RD29B shares 26.4% identity and 34.3% similarity with CaOSR1. Domain analysis revealed that CaOSR1 have two conserved regions: acidic region (72-88 aa) and CAP160 domain (653-679 aa), which are shown in common in most of CaOSR1-homologous proteins ( Figure 1B and Supplementary Figure 1). CAP160 domain is first reported in spinach cold acclimation protein (CAP) 160 protein (Kaye et al., 1998). Although the precise function of CAP160 is still unknown, CAP160 is induced by drought stress as well as low temperature exposure (Kaye et al., 1998). These stresses involve induction of dehydration and plants respond to them very similarly at molecular level . Based on the data, we proposed that CaOSR1 may function in plant response to dehydration-involved stresses.
Induction of the CaOSR1 Gene by ABA, Drought, and High Salt Stress Treatments and Subcellular Localization of the CaOSR1 Protein The CaOSR1 gene was isolated from ABA-treated pepper leaves, using the differential hybridization assay (Lim et al., 2014). To Phylogenetic tree analysis of CaOSR1 protein. Blast search was performed by using deduced amino acid sequences of CaOSR1 gene and top ranked sequences from each plant species were gathered. For Arabidopsis, as CaOSR1 homolog, a well-known gene RD29B (At5g52300; accession number BAA02375) and it's homologous gene CAP160 (At4g25580; accession number NP_194288) were also added. Multiple alignment of those amino acid sequences was performed by using ClustalW2 and the phylogenetic tree was drawn with MEGA software (version 5.2). (B) Multiple sequence alignment analysis of Cold Acclimation Protein (CAP) 160 domain. Amino acid residues of CAP160 domains are shaded according to percent identity in ClustalW.
Frontiers in Plant Science | www.frontiersin.org investigate the potential involvement of CaOSR1 in dehydration stress responses, we examined the expression levels of this gene after ABA, drought, and high salinity treatments (Figure 2). We found that the accumulation of CaOSR1 transcripts was first detected at 6 h after ABA treatment and reached a maximum level at 24 h (Figure 2A). ABA and abiotic stress signals, including drought and high salt, seem to share common elements in their respective signaling pathways; however, these stress signals are not solely dependent on ABA signaling (Jakab et al., 2005). In the present study, we found that after drought stress treatment, the accumulation of CaOSR1 transcripts was first detected at 12 h and reached a maximum level at 24 h ( Figure 2B). On the other hand, after treatment with NaCl, CaOSR1 transcripts were weakly expressed at 24 h ( Figure 2C). Our data suggest that CaOSR1 functions in osmotic stress responses.
To investigate the subcellular localization of the CaOSR1 protein in plant cells, we fused the green fluorescent protein (GFP) reporter gene to the C-terminal region of CaOSR1 under the control of the 35S promoter. We found that expression of the 35S:CaOSR1-GFP fusion protein in Nicotiana benthamiana epidermal cells generated GFP signals in the nucleus ( Figure 2D). We used DAPI staining as a nucleus marker, and observed that the blue signals localized to nucleus overlapped with the GFP signals.

Increased Susceptibility of CaOSR1-Silenced Pepper Plants to Drought Stress
The CaOSR1 gene was induced by abiotic stresses, and therefore we postulated that CaOSR1 is involved in osmotic stress responses. To test this hypothesis, we performed VIGS-based gene function analysis in pepper plants and an overexpression assay in Arabidopsis plants. We examined the level of VIGS using reverse transcription-polymerase chain reaction (RT-PCR) analysis of control (TRV:00) and CaOSR1-silenced pepper (TRV:CaOSR1-RNAi) leaves. We found that CaOSR1 expression in the leaves of CaOSR1-silenced peppers was compromised but remained marginally detectable ( Figure 3A). The expression of CaOSR1 was induced by drought stress, implying that this gene functions in the drought stress response. Hence, we compared the phenotypes displayed by CaOSR1-silenced pepper plants and control plants after drought stress treatment ( Figure 3B). Under well-watered conditions, we observed no phenotypic differences between control and CaAIP1-silenced pepper plants ( Figure 3B, upper panel). However, when we subjected plants to drought stress by withholding watering for 12 days, CaOSR1-silenced pepper plants displayed a more wilted phenotype than control plants ( Figure 3B, middle panel). After recovery by rewatering, CaOSR1-silenced pepper plants resumed growth more slowly than control plants ( Figure 3B, lower panel). Moreover, the  survival rates of CaOSR1-silenced plants and control plants were 42.8 and 85.7%, respectively ( Figure 3C). Based on the wilted phenotype displayed by CaOSR1-silenced pepper plants after drought stress treatment, we postulated that transpirational water loss in detached CaOSR1-silenced pepper leaves is increased because of reduced water retention. To test this hypothesis, we analyzed the transpiration rate in the leaves of control plants and CaOSR1-silenced pepper plants. We found that the rate of water loss was higher in the leaves of CaOSR1silenced pepper plants than in the leaves of control plants ( Figure 3D). The observed phenotypic discrepancy between CaOSR1-silenced pepper plants and control plants under drought stress conditions prompted us to investigate the ABA sensitivity by measuring the leaf temperatures and stomatal apertures. The leaf temperature decreases when the stomata open, because of evaporative cooling. We found that before ABA treatment, the leaf temperatures are not significantly different between both plants (Supplementary Figure 2A). However, the leaf temperatures of CaOSR1-silenced pepper plants were significantly lower than those of control plants after ABA treatment ( Figure 3E). The ABA sensitivity can further be determined by measuring the change in stomatal pore size in response to ABA treatment. In the absence of ABA, we determined no significant differences in stomatal aperture between the leaves of control plants and CaOSR1-silenced pepper plants ( Figure 3F). However, consistent with our leaf temperature data, we found that after ABA treatment, the stomatal apertures were significantly larger in the leaves of CaOSR1-silenced pepper plants than in the leaves of control plants. Taken together, our results imply that the high rate of transpiration, and hence the increased drought susceptibility of CaOSR1-silenced pepper plants, is mainly derived from reduced ABA sensitivity.

Enhanced Sensitivity of CaOSR1-OX Transgenic Plants to ABA
We further studied the function of CaOSR1 in osmotic stress responses by generating transgenic Arabidopsis plants that overexpressed the CaOSR1 gene under the control of the cauliflower mosaic virus (CaMV) 35S promoter. We obtained two independent T 3 transgenic lines (CaOSR1-OX #1 and CaOSR1-OX #2) showing relatively high expression of CaOSR1 (Supplementary Figure 3), and we used these two lines in our phenotypic analyses. Under normal growth conditions, we observed no significant differences in the growth of wild-type and CaOSR1-OX plants (Figures 4, 5). The primary function of ABA is defense response to abiotic stress; moreover, ABA signaling overlaps with the defense signaling response to abiotic stress (Zhu, 2002). To examine ABA sensitivity during germination, we sowed seeds on MS medium supplemented with various concentrations of ABA (0.0, 0.5, and 1.0 µM). In the absence of ABA, we determined no significant difference in germination rates between wildtype and CaOSR1-OX seeds. However, in the presence of ABA, the germination rate of CaOSR1-OX seeds was lower than that of wild-type seeds ( Figure 4A). Next, we determined the rates of cotyledon greening 5 days after sowing and the root lengths 8 days after sowing. We found that treatment with increasing concentrations of ABA resulted in decreased rates of cotyledon greening and reduced root lengths; these (B,C) Seedling development of CaOSR1-OX mutants and wild-type plants exposed to ABA. The numbers of seedlings in each line with expanded cotyledons were counted (C) and representative photographs were taken 5 days after plating (B). Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. (D,E) Root elongation of wild-type and transgenic lines exposed to ABA. The root lengths of each plant were measured 8 days after sowing (E) and representative images were taken (D). Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. Different letters indicate significant differences between wild-type and transgenic lines (P < 0.05; ANOVA followed by Fisher's LSD test). effects were observed with ABA concentrations as low as 0.5 µM (Figures 4B-E). As predicted, CaOSR1-OX plants exhibited an ABA-hypersensitive phenotype. In the presence of ABA, the rate of cotyledon greening was significantly lower in CaOSR1-OX seedlings than in wild-type seedlings (Figures 4B,C); moreover, root elongation of CaOSR1-OX seedlings was significantly inhibited (Figures 4D,E). Our results indicate that ectopic expression of CaOSR1 confers ABA hypersensitivity in Arabidopsis. Leaf peels were harvested from 3-week-old plants and stomatal apertures were measured under a microscope. Data represent the mean ± standard error of three independent experiments, each evaluating 20 plants. Different letters indicate significant differences between wild-type and transgenic lines (P < 0.05; ANOVA followed by Fisher's LSD test).

Reduced Sensitivity of CaOSR1-OX Transgenic Plants to High Salinity and Osmotic Stress
The expression of CaOSR1 in pepper leaves was induced by NaCl treatment (Figure 2C), and therefore we predicted that overexpression of this gene alters the response to high salt stress. In the absence of NaCl, we determined no significant differences in germination rates between wild-type and transgenic seeds. However, in the presence of 100 and 150 mM NaCl, the germination rates of CaOSR1-OX seeds were significantly higher than those of wild-type seeds (Supplementary Figure 4A). Next, we assessed the seedling growth and development of wild-type and CaOSR1-OX plants in the presence of NaCl. We found that CaOSR1-OX transgenic plants showed reduced sensitivity to high salinity stress at the seedling stage (Supplementary Figures 4B-E).
CaOSR1-OX plants displayed an ABA-hypersensitive phenotype during the seed germination and seedling growth stages (Figure 4), and therefore we postulated that the CaOSR1-OX gene functions in the defense response to osmotic stress (Supplementary Figure 5). To test this hypothesis, we first assessed the germination rates in wild-type and transgenic plants after treatment with various concentrations of mannitol. After exposure to 300 or 400 mM mannitol, the germination rate of CaOSR1-OX seeds was significantly higher than that of wild-type seeds (Supplementary Figure 5A). Next, we determined the rates of cotyledon greening 5 days after sowing and the root lengths 8 days after sowing (Supplementary Figures 5B-E). We found that after exposure to mannitol, the rate of cotyledon greening and root elongation was significantly higher in CaOSR1-OX seedlings than in wild-type seedlings (Supplementary Figures 5B-E). As predicted, our results indicate that CaOSR1 functions as a regulator of the osmotic stress response.

Enhanced Tolerance of CaOSR1-OX Transgenic Plants to Drought Stress
CaOSR1-OX plants showed less sensitive phenotypes to mannitol-and salt-induced osmotic stresses during the seed germination and seedling growth stages, and therefore we investigated the drought tolerance of these transgenic plants (Figure 5). When grown under well-watered conditions, we observed no phenotypic differences between wild-type and CaOSR1-OX plants (Figure 5A, upper panel). However, when we subjected plants to drought stress by withholding watering for 9 days, transgenic plants displayed a less wilted phenotype than wild-type plants ( Figure 5A, middle panel). After recovery by rewatering for 2 days, the survival rates of CaOSR1-OX lines #1 and #2 were 75 and 56%, respectively, whereas that of wildtype plants was only 38% (Figure 5A, lower panel). Next, we compared the fresh weight of detached rosette leaves to monitor the transpirational water loss and thus determine whether the drought-tolerant phenotype displayed by CaOSR1-OX plants was derived from an altered transpiration rate ( Figure 5B). We found that the transpiration rate was lower in the leaves of CaOSR1-OX plants than in the leaves of wild-type plants, implying that the drought-tolerant phenotype was derived from increased water retention.
Generally, drought tolerance is determined by at least two cellular or molecular parameters. Previous studies have used measurements of leaf temperature and stomatal aperture to establish that ABA hypersensitivity leads to enhanced drought tolerance (Cheong et al., 2007;Lim et al., 2015b;Park S. Y. et al., 2015). Other studies have revealed correlations between drought tolerance and low or high levels of stress-related gene expression; these correlations may lead to increased or decreased drought tolerance, respectively (Gonzalez-Guzman et al., 2012;Park C. et al., 2015). To determine whether the enhanced drought tolerance of CaOSR1-OX plants is associated with increased ABA sensitivity, we measured the leaf temperature FIGURE 6 | Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis of drought-inducible genes in the CaOSR1-OX mutant exposed to drought stress at 6 h after detachment. The relative expression levels (∆∆CT) of each gene were normalized to the geometric mean of Actin8 as an internal control gene. Data represent the mean ± standard error of three independent experiments. Different letters indicate significant differences between wild-type and transgenic lines (P < 0.05; ANOVA followed by Fisher's LSD test). and stomatal aperture. We found that after ABA treatment, the leaf temperatures were significantly higher in CaOSR1-OX plants than in wild-type plants (Figure 5C), implying that the CaOSR1 protein plays a key role in ABA-mediated stomatal closure. In the absence of ABA, we determined no significant differences in leaf temperature and stomatal aperture between the leaves of wild-type and transgenic plants (Supplementary Figure 2B and Figure 5D). However, after exposure to 20 µM ABA, the stomatal apertures in the leaves of CaOSR1-OX plants were significantly smaller than those in the leaves of wild-type plants ( Figure 5D).
Finally, we performed quantitative reverse transcriptionpolymerase chain reaction (qRT-PCR) analysis in wild-type and CaOSR1-OX plants to investigate the function of CaOSR1 in controlling the expression of stress-responsive genes under drought conditions (Figure 6). In general, the levels of ABA in plant tissues increase under drought stress conditions, and this induces the expression of many stress-related genes, including NCED3, KIN2, COR15A, and RD29B, After 6 h of drought stress treatment, we determined significantly higher expression levels of these genes in CaOSR1-OX plants than in wild-type plants, implying that conferred expression of CaOSR1 affects the expression levels of stress-related genes. Taken together, our results confirm our hypothesis that the CaOSR1 protein functions as a positive regulator of multiple osmotic stresses in plants.

DISCUSSION
In the present study, we identified and functionally characterized CaOSR1, which functions as a positive regulator in the osmotic stress response via ABA-mediated signaling. Under water-deficit conditions, expression of CaOSR1 in pepper and Arabidopsis plants resulted in altered phenotypes.
The function of specific proteins in plant cells can be predicted by homology search analysis with other known proteins. Using the deduced amino acid sequence of CaOSR1 gene as query, a BLASTP search at NCBI revealed that CaOSR1 is highly close to low-temperature induced (LTI) protein 65, but uncharacterized, from several plant species. Most of these proteins have distinctly two conserved domains: acidic domain and CAP160 domain. Of them, we initially focused on Arabidopsis RD29B, alternately named LTI65, even though CaOSR1 has low homology to RD29B. It is well-known that RD29B and its homologous gene RD29A are induced by cold, drought, NaCl, and ABA treatments (Yamaguchi-Shinozaki and Shinozaki, 1993;Msanne et al., 2011). Promoter sequences of RD29A and RD29B genes have different number of Dehydration-Responsive Element (DRE) and ABA-Responsive Element (ABRE): several DRE and one ABRE in RD29A and several ABRE and one DRE in RD29B (Msanne et al., 2011). Similarly to RD29B promoter, 2kb-upstream sequence of CaOSR1 gene contains one core motif of dehydration-responsive element/C-repeat (DRE/CRT; GCCGAC) and 7 ABREs (CACGT). This feature may contribute to induction of CaOSR1 in response to ABA, drought, and salt treatments as shown in Figure 2. Considered primary structure of RD29B and RD29A, we found that CAP160 domain is found in RD29B, not RD29A. Although internal sequence repeats that are shown in CaOSR1 are found only in RD29A, we predicted that CaOSR1 can be orthologous protein of Arabidopsis RD29B and contribute to mitigating abiotic stress.
Intriguingly, CaOSR1 showed different pattern of subcellular localization compared with RD29B: CaOSR1-GFP fusion proteins are localized in the nucleus (Figure 2D), while RD29B targets to the cytoplasm (Msanne et al., 2011). Information on protein subcellular localization is very crucial because it provides clue to understand biological function of protein. Considered this difference and low sequence homology between CaOSR1 and RD29B, we cannot rule out the possibility that CaOSR1 can have alternative biological functions in pepper plants, compared with RD29B. Next, we focused on CAP160 domain which is first reported in Spinach CAP160 protein (Kaye et al., 1998). Similarly to RD29B, CAP160 gene is induced by drought stress as well as low temperature exposure and its encoded protein localized to the cytoplasm. However, sequence homology between the two proteins is low (25.9% identity/36.7% similarity) and this pattern is also shown between CaOSR1 and CAP160 (22.9% identity/33.3% similarity). Functional role of RD29B and CAP160 remains unclear and shared sequence homology between the proteins is very limited, which imply that CaOSR1 may be a novel stress-responsive protein in pepper plant.
To examine the role of CaOSR1, we conducted VIGS and overexpression-based genetic analysis for loss-of function and gain-of function in pepper and Arabidopsis, respectively. We found that CaOSR1-silenced pepper plants showed increased drought susceptibility; this was characterized by decreased water retention capacity, indicating that compromised expression of CaOSR1 impedes stomatal closure. In contrast to CaOSR1-silenced pepper plants, CaOSR1-OX Arabidopsis plants exhibited enhanced tolerance to drought stress via ABA-mediated signaling. In addition, we demonstrated that CaOSR1-OX plants were hypersensitive to ABA, implying that CaOSR1 regulates osmotic stress tolerance via ABA-mediated cell signaling. We further observed no phenotypic differences between CaOSR1-silenced pepper plants and control plants after exposure to NaCl (data not shown). On the other hand, CaOSR1-OX plants showed hyposensitivity to high salt stress during the seed germination and seedling growth stages, implying that CaOSR1 acts as a positive regulator of the high salinity stress response. The observed phenotypic discrepancy between CaOSR1-silenced pepper plants and CaOSR1-OX transgenic Arabidopsis plants is presumably derived either from the use of different species or from the functions of CaOSR1 that distinguish it from Arabidopsis genes in terms of response to high salinity.
Based on the effect of the expression level of CaOSR1 on the sensitivity to ABA, we examined whether CaOSR1 affects the expression of the NCED3 gene, which encodes the rate-limiting ABA biosynthesis enzyme in Arabidopsis (Iuchi et al., 2001;Tan et al., 2003). Under optimal plant growth conditions, we determined no significant difference in the expression level of NCED3 between CaOSR1-OX and wildtype plants. However, after 6 h of drought stress treatment, we determined significantly higher expression levels of this gene in CaOSR1-OX plants than in wild-type plants (Figure 6). Furthermore, CaOSR1-OX plants showed higher accumulation levels of the stress-responsive marker genes RD29B, KIN2, and COR15A, which are associated with abiotic tolerance via ABA-dependent and ABA-independent pathways, respectively (Artus et al., 1996;Uno et al., 2000;Hoth et al., 2002). Our results imply that CaOSR1 acts upstream of these genes in the drought stress response. In addition, the enhanced expression levels of abiotic stress marker genes may reflect the increased ABA sensitivity of CaOSR1-OX plants. Nevertheless, the increased or decreased expression levels of stress-related genes do not fully explain the altered phenotypes displayed by CaOSR1-OX plants in response to ABA and osmotic stress treatments.
In conclusion, we have shown that CaOSR1 functions as a positive regulator of the osmotic stress response in plants. CaOSR1-OX plants exhibited an ABA-hypersensitive phenotype, and this was characterized by increased ABA-induced stomatal closure and enhanced expression levels of ABA-mediated stressresponsive genes. Our findings provide a valuable insight into the osmotic stress response in plants. Nevertheless, the precise mechanism whereby CaOSR1 serves as a positive component of osmotic stress responses remains unclear. Further studies based on genetic and molecular analysis of the upstream and downstream regions of the CaOSR1 gene will help to clarify the role of CaOSR1 in the osmotic stress response. Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. (B,C) Seedling development of CaOSR1-OX mutants and wild-type plants exposed to NaCl. The numbers of seedlings in each line with expanded cotyledons were counted (C) and representative photographs were taken 5 days after plating (B). Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. (D,E) Root elongation of wild-type and transgenic lines exposed to NaCl. The root lengths of each plant were measured 8 days after sowing (E) and representative images were taken (D). Data represent the mean ± standard error of three independent experiments. Different letters indicate significant differences between wild-type and transgenic lines (P < 0.05; ANOVA followed by Fisher's LSD test).

AUTHOR CONTRIBUTIONS
Supplementary Figure 5 | Reduced sensitivity of CaOSR1-OX transgenic Arabidopsis lines to osmotic stress during germination and seedling development. (A) Germination rates of CaOSR1-OX mutants and wild-type (WT) plants on 0.5× MS medium supplemented with various concentrations of mannitol. Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. (B,C) Seedling development of CaOSR1-OX mutants and wild-type plants exposed to mannitol. The numbers of seedlings in each line with expanded cotyledons were counted (C) and representative photographs were taken 5 days after plating (B). Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. (D,E) Root elongation of wild-type and transgenic lines exposed to mannitol. The root lengths of each plant were measured 8 days after sowing (E) and representative images were taken (D). Data represent the mean ± standard error of three independent experiments, each evaluating 36 seeds. Different letters indicate significant differences between wild-type and transgenic lines (P < 0.05; ANOVA followed by Fisher's LSD test).