Rice OsGL1-6 Is Involved in Leaf Cuticular Wax Accumulation and Drought Resistance

Cuticular wax is a class of organic compounds that comprises the outermost layer of plant surfaces. Plant cuticular wax, the last barrier of self-defense, plays an important role in plant growth and development. The OsGL1-6 gene, a member of the fatty aldehyde decarbonylase gene family, is highly homologous to Arabidopsis CER1, which is involved in cuticular wax biosynthesis. However, whether OsGL1-6 participates in cuticular wax biosynthesis remains unknown. In this study, an OsGL1-6 antisense-RNA vector driven by its own promoter was constructed and introduced into the rice variety Zhonghua11 by Agrobacterium-mediated transformation to obtain several independent transgenic plants with decreased OsGL1-6 expression. These OsGL1-6 antisense-RNA transgenic plants showed droopy leaves at the booting stage, significantly decreased leaf cuticular wax deposition, thinner cuticle membrane, increased chlorophyll leaching and water loss rates, and enhanced drought sensitivity. The OsGL1-6 gene was constitutively expressed in all examined organs and was very highly expressed in leaf epidermal cells and vascular bundles. The transient expression of OsGL1-6-GFP fusion indicated that OsGL1-6 is localized in the endoplasmic reticulum. Qualitative and quantitative analysis of the wax composition using gas chromatography-mass spectrometry revealed a significantly reduced total cuticular wax load on the leaf blades of the OsGL1-6 antisense-RNA transgenic plants as well as markedly decreased alkane and aldehyde contents. Their primary alcohol contents increased significantly compared with those in the wild type plants, suggesting that OsGL1-6 is associated with the decarbonylation pathways in wax biosynthesis. We propose that OsGL1-6 is involved in the accumulation of leaf cuticular wax and directly impacts drought resistance in rice.


Introduction
The cuticle is a continuous hydrophobic lipid layer structure that covers the exposed ground parts of terrestrial plants and forms a protective barrier against the external environment. The cuticle is synthesized by the epidermal cells and is composed of cutin polymer matrix and waxes [1,2]. Cuticular waxes comprise the primary structure of the cuticle and play the following important roles: limiting non-stomatal water loss [3]; repelling bacterial and fungal pathogens and herbivorous insects [4]; mediating the interaction with other organisms, such as bacteria, fungi, and insects [5][6][7]; and preventing UV radiation and frost damage [2,8].
The cuticular wax of all plants consists of derivatives of very long-chain fatty acid (VLCFAs) including alkanes, aldehydes, ketones, primary and secondary alcohols, and esters [9,10]. The biosynthesis of cuticular wax is accomplished by two steps: the fatty acid elongase-mediated extension of the C16 and C18 fatty acids to VLCFA chains and the conversion of each VLCFA to wax components by the decarbonylation and acyl reduction pathways in the endoplasmic reticulum (ER). The acyl reduction pathway mediates the production of primary alcohols and wax esters, whereas the decarbonylation process produces aldehydes, secondary alcohols, alkanes, and ketones [9,10].
The genes associated with wax synthesis have been successfully isolated and identified from maize and Arabidopsis using genetic analysis. Among these genes, FAE1, FDH, KCS1, PAS2, CER6, CER10, GL8A and GL8B are involved in the synthesis of very longchain fatty acid wax precursors [11][12][13][14][15][16][17][18][19]. Genes including CER4, WSD1 and MAH1 are involved in the synthesis of wax components [20][21][22], of which, CER4 and WSD1 participate in the acyl reduction pathway to catalyze the production of primary alcohol and wax ester, respectively [20,21]. MAH1 participates in the decarbonylation pathway to catalyze the conversion of alkanes into secondary alcohols and ketones [22]. CER5 and LTPG are involved in wax secretion [23][24][25]. Genes such as WIN1/SHN encode regulatory proteins [26,27]. All known VLCFA elongases and wax biosynthesis enzymes are located in the ER, and the cuticular lipids synthesized in the ER must be transported to the cuticle, where they apparently self-assemble into the cuticle proper [10]. CER5 in Arabidopsis is the first characterized gene that encodes the plasma membrane-localized ABC transporter required for the transport of wax components from the epidermal cells to the cuticle [23].
LTPG in Arabidopsis encodes a glycosylphosphatidylinositolanchored lipid transfer protein that is localized in the plasma membrane of stem epidermal cells. LTPG has the capacity to bind to the lipid, which is required for the export of lipids to the plant surface, suggesting that LTPG may function as a component of the cuticular lipid export machinery [24,25]. WIN1/SHN encodes an APETALA2/EREBP-type transcription factor [26,27]. Overexpression of WIN1/SHN1 increases wax production and enhances the drought tolerance in Arabidopsis [28]. WXP1 is a putative AP2 domain-containing transcription factor gene from the model legume Medicago truncatula and activates wax production in the acyl reduction pathway. Overexpression of WXP1 also leads to increased cuticular wax loading on the leaf surfaces, reduced water loss, and enhanced drought tolerance in transgenic alfalfa [29].
Many wax-defective mutants were obtained in Arabidopsis, maize and other plants that either showed a reduced wax accumulation or modified wax composition and were often characterized by their smooth or light green phenotype. CER1 and CER3/WAX2/ YRE/FLP in Arabidopsis and GL1 in maize have been isolated and characterized using wax-defective mutants. The cer1 mutants in Arabidopsis display glossy green stems and fruits. Biochemical studies have shown that the cer1 mutants exhibited a significant reduction of alkanes, secondary alcohols, and ketones as well as an increased aldehyde content. CER1 was predicted to encode an aldehyde decarboxylase, a key wax biosynthetic enzyme that catalyzes the conversion of aldehyde to alkane [9,[30][31][32][33][34]. The cer1 mutants were conditionally male sterile, suggesting that the CER1 gene has an essential function in pollen development [31]. Overexpression of CER1 dramatically increased alkane production and resulted in an increased wax load, reduced cuticle permeability, and enhanced plant stress resistance [33]. A screen for CER1 physical interaction partners revealed CER1 interacts with the wax-associated protein CER3 and ER-localized cytochrome b5 isoforms (CYTB5s) [34].
The cer3 (cer3/wax2/yre/flp1) mutants showed glossy stems as well as altered cuticle membrane and cuticular waxes. Biochemical studies have shown that the aldehyde, hydrocarbon, ketone, and secondary alcohol contents were significantly reduced in the mutants [35][36][37], suggesting that CER3/WAX2/YRE/FLP might encode an aldehyde-producing enzyme that catalyzes the conversion of acyl-CoA to an intermediate aldehyde [36]. The gl1 mutants showed waxless seedling leaves and altered cuticle membrane and cuticular waxes similar to those observed in the wax2 mutant. Levels of the aldehydes and alcohols in the gl1 mutant were markedly reduced, indicating that GL1 is essential for the elongation process during the synthesis of cuticular wax [38].
Until now, three numbers of the fatty aldehyde decarbonylase gene family were characterized in rice. Wda1, the first characterized gene, is strongly expressed in the epidermal cells of the anthers. The wda1 mutant demonstrated absent epicuticular wax crystals in the outer layer of the anther and severely retarded microspore development. Contents of the fatty acids, alkanes, alkenes, and primary alcohols in the wda1 mutant were severely reduced, indicating that Wda1 may be involved in the general processes of VLCFA biosynthesis [39]. The osgl1-1 (also known as wsl2) mutant showed a decreased cuticular wax deposition and enhanced drought sensitivity [40,41]. A wax composition analysis The closed black boxes indicate exons, connecting lines indicate introns, and closed white boxes indicate the 59 and 39 untranslated region. The ATG start codon and TGA stop codon are also indicated. p1 and p2 are the primers used in the semi-quantitative reverse transcription polymerase chain reaction analysis of the OsGL1-6 transcript. p3 and p4 are the primers used to amplify the antisense fragment. p5 and p6 are the primers used to amplify the intact ORF. (B) Phylogenic analysis of proteins homologous to OsGL1-6. The coding region sequences were aligned using Clustal W [51] and subjected to neighbor-joining analysis as implemented in MEGA5.0 [52]. Bar = 0.5 kb. doi:10.1371/journal.pone.0065139.g001 revealed a substantially reduced quantity of C22-C32 fatty acids in the wsl2 mutant, suggesting that WSL2 may be involved in the VLCFA elongation [41]. The osgl1-2 mutant showed reduced cuticular wax synthesis and increased sensitivity to drought stress, while OsGL1-2 overexpression resulted in increased amounts of cuticular waxes and enhanced drought tolerance [42]. Hence, members of the fatty aldehyde decarbonylase gene family play important roles in cuticular wax synthesis and drought resistance.
A systematic sequence analysis revealed a total of 11 GL1 homologous genes in rice, designated OsGL1-1 to OsGL1-11 [42]. However, only three genes, i.e. Wda1 (also called OsGL1-5), OsGL1-1 (WSL2), and OsGL1-2, were characterized. The mutations in the three genes confer different phenotypic changes on the cuticular wax compositions, implying that they may perform different roles in cuticular wax biosynthesis. To better understand the molecular mechanism of cuticular wax biosynthesis in rice,  1, subcultured Hgy-resistant calli; 2, differentiated Hgy-resistant calli; 3, spikelets at the meiosis stage; 4, spikelets at the uninucleate stage; 5, spikelets at the binucleate stage; 6, local amplification of spikelets at the binucleate stage; 7, root; 8, stem; 9, young leaf; 10, local amplification of the young leaf; 11, mature leaf; 12, cross-section of the mature leaf; 13, enlarged cross-section of the stem; 14, enlarged cross-section of the mature leaf. The different stages of spikelet development were collected according to Feng et al. [53]. Red arrows indicate the epidermis cell layer and vascular bundle. Scale bars = 1mm in (1) to (12) and 50 mm in (13) to (14). doi:10.1371/journal.pone.0065139.g002 more putative genes that are deduced to participate in the process need to be characterized.
In this study, we conducted a functional analysis of OsGL1-6 in rice growth and development by generating OsGL1-6 antisense-RNA transgenic plants. OsGL1-6 is homologous to CER1 in Arabidopsis and Wda1 in rice, universally expressed in vegetative and reproductive organs, and especially highly expressed in leaf epidermal cells and vascular bundles. The results of a phenotypic characterization and drought sensitivity experiment of OsGL1-6 antisense-RNA transgenic plants indicated that OsGL1-6 is involved in cuticular wax accumulation and drought resistance.

Plants and Other Experimental Materials
The rice variety Zhonghua11 (Oryza sativa L. ssp. japonica cv. Zhonghua11) was used for all of the experiments in this study. The wild type (WT) and transgenic plants used in the drought stress experiments were hydroponically cultured as reported previously [43], and all rice plants were grown in a greenhouse under normal growth conditions. Escherichia coli strain DH10B and Agrobacterium tumefaciens strain EHA105 were used for the cloning and transformation experiments. pCAMBIA1380 was used as the binary vector for Agrobacterium-mediated transformation.
Generating OsGL1-6 Antisense-RNA Transgenic Plants The OsGL1-6 antisense transgene driven by the cognate promoters of the OsGL1-6 was constructed following the protocol described by Li et al. [44]. A 624-bp fragment of the OsGL1-6 cDNA was amplified using reverse transcription polymerase chain reaction (RT-PCR) from Zhonghua11 with the gene-specific OsGL1-6Af and OsGL1-6Ar primers (Table S1), and the PCR product was inversely inserted into the multiple cloning site of pCAMBIA1380. A 1946-bp genomic fragment located upstream of the annotated ATG start codon of the OsGL1-6 was isolated from Zhonghua11 by using OsGL1-6Pf and OsGL1-6Pr primers (Table S1) and was then inserted upstream of the antisense fragment. The construct was then transformed into A. tumefaciens strain EHA105, which was used to transform rice calli generated from mature seeds of Zhonghua11 according to a previously described protocol [45].

RT-PCR Analysis
Total RNA from different Zhonghua11 tissues and transgenic plant seedlings was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions.

Histochemical b-glucuronidase (GUS) Analysis
A modified pCAMBIA1300 vector containing the CaMV35S promoter, GUS gene, and NOS terminator inserted into the multiple cloning sites was used for the GUS fusion construction [46]. A 1,946-bp genomic fragment located upstream of the annotated ATG start codon of OsGL1-6 was isolated from Zhonghua11 using the primer pair OsGL1-6Gf and OsGL1-6Gr (Table S1) and then substituted for the CaMV35S promoter in the above construct. The GUS fusion construct was then transformed into Zhonghua11 using Agrobacterium-mediated transformation [45]. For the GUS staining, the tissues were fixed for 45 min in a 50 mM sodium phosphate buffer, pH 7.0, containing 1% (v/v) formaldehyde and 0.5% (v/v) Triton X-100 after being vacuumed for 10 min. The fixed tissues were rinsed and vacuum-infiltrated three times (5 min each) in 50 mM sodium phosphate buffer, pH 7.0, containing 0.5% (v/v) Triton X-100, 1 mM potassium ferricyanide and 1 mM potassium ferrocyanide. The above tissues were subsequently stained in a 1 mM sodium phosphate buffer, pH 7.0, containing 0.5% (v/v) Triton X-100, 0.25 mg/mL X-Gluc, and 1 mM potassium ferrocyanide at 37uC overnight after being vacuumed for 10 min. The samples were then rinsed and stored in 70% (v/v) ethanol and observed under a microscope (Nikon).

Subcellular Localization
The CaMV35S promoter, eGFP/CFP fragments, and NOS terminator were successively cloned into pUC18 to construct the transient expression vectors. The intact open reading frame (ORF) of OsGL1-6 was amplified using the primer pair OsGL1-6-eGFPf, OsGL1-6-eGFPr (Table S1) and then subcloned into the transient expression vector between the CaMV35S promoter and the eGFP gene, generating in-frame with the N-terminus of eGFP being driven by the CaMV35S promoter. The CFP-KDEL fusion transient expression vector was constructed by subcloning of the ER retention signal KDEL to the 39 end of the CFP gene [47], and sequences of the used primers were listed in Table S1. Protoplasts from the leaf sheaths of the 14-day rice seedlings were prepared and transient expression was analyzed following a described protocol [48,49].

Western Blot Analysis
The polyclonal antibody of OsGL1-6 for the Western blot analysis was made in our laboratory. The intact ORF of OsGL1-6 was amplified using primer pair OsGL1-6-pETf, OsGL1-6-pETr   (Table S1) and then cloned into pET-23d. The resulting recombinant vector was introduced into E. coli strain BL21 (DE3) and induced to express the OsGL1-6 fusion protein using isopropylthio-b-galactoside. The polyclonal antibodies were obtained using OsGL1-6 fusion protein as the antigen in immune rabbits. Seedlings (0.3 g) from the WT and transgenic plants were ground into a homogenate in an extraction buffer, pH 7.5, containing 0.1 M Tris-HCl and 25 mM ethylenediaminetetraacetic acid. The homogenate was centrifuged at 12,0006g for 10 min at 4uC, and the upper aqueous solution containing the proteins was collected. The immunoblot analysis was performed as described previously by Li et al. [44]. Heat shock protein 90 (BGI, Shenzhen, China) was used for reference [50].

Drought Sensitivity and Cuticular Permeability Experiment
The seeds of the WT and transgenic plants were germinated and cultured on half-strength Murashige and Skoog's medium under a 12 h light/12 h dark schedule at 26uC for 4 d. The seedlings were then transferred out and hydroponically cultured for 10 d in a greenhouse. Drought treatment was subsequently performed in the above seedlings in the air for 10 h. The plants were re-watered for 7 d, and plant survival rates were calculated.
The chlorophyll leaching experiment was performed as previously described [40]. Briefly, the second leaves from the top were sampled from each tiller at the plant booting stage, and the leaves were then cut into segments and immersed in 30 mL of 80% (v/v) ethanol with gentle agitation in the dark. At 10, 20, 30, 40, 50, 60, 90, 120, and 150 min, 3-mL aliquots were taken for chlorophyll quantification. These samples were poured back to the same tube after the measurements. The chlorophyll concentration was quantified using a spectrophotometer at wavelengths of 664 and 647 nm.
For the water-loss measurements, the second leaves from the top were detached from each tiller at the plant booting stage and soaked in deionized water for .2 h in the dark. The detached leaves were subsequently removed from soaking, blotted gently with a napkin to remove excess water, and weighed at 0.5, 1, 1.5, 2, 2.5 and 3 h as described by Chen et al. [35]. Measurements were performed in a very low-light environment. The percentage of fresh weight loss was calculated based on the initial sample weight.

Scanning Electron Microscopic Analysis
The tips of the booting-stage flag leaves were fixed at 4uC in a 0.1 M sodium phosphate buffer, pH 6.8, containing 2.5% (v/v) glutaraldehyde for 24 h. The samples were rinsed three times (10 min each) in a 0.1 M sodium phosphate buffer, pH 6.8, dehydrated first with an ethanol series (10 min each) of 50% to 100% (v/v) and then twice (15 min each) with 100% (v/v) acetone, and then exchanged three times (15 min each) with isoamyl acetate. The samples were processed for critical point drying in liquid CO 2 (Bal-Tec), gold-coated (10-nm-thick), and examined in an XL-30-ESEM (FEI) with an accelerating voltage of 20 kV.

Transmission Electron Microscopic Analysis
The tips of the booting-phase flag leaves were fixed for 24 h in sodium phosphate buffer, pH 7.2, containing 4% (v/v) glutaraldehyde and 3% (w/v) paraformaldehyde. The samples were rinsed three times (20 min each) in sodium phosphate buffer, pH 7.2 and

Wax Extraction and Composition Analysis
The total wax content was determined using the weight method. In brief, 2 g of the booting-stage flag leaves was immersed in 30 mL of chloroform at 60uC for 30 s, and the wax content was then determined by analytical balance after complete chloroform evaporation.
The leaf wax composition was analyzed as previously described [39]. Briefly, five 10-cm booting-stage flag leaf blades were immersed in 30 mL of chloroform at 60uC for 30 s; 20 mg of tetracosane was used as an internal standard. The solvent was evaporated under a nitrogen stream. The samples were transferred to gas chromatography (GC) vials and converted to the trimethylsilyl derivatives in 10 mL of pyridine and 10 mL of bis-(N, N-trimethylsilyl)-tri-fluoroacetamide (BSTFA) for 40 min at 70uC. After cooling, the BSTFA and pyridine were removed under a nitrogen stream and their derivatives were dissolved in 100 mL of chloroform for gas chromatography-mass spectrometry (GC-MS) analysis. The wax composition was determined using a GC (HP1-MS) column and mass spectrometric detector (5973N, Agilent). The wax loads were estimated by quantifying the major peak areas compared with the internal standard area. The GC was conducted with an on-column injection at an initial oven temperature of 50uC for 2 min, followed by an increase of 40uC/min to 200uC, 2 min at 200uC, an increase of 3uC/min to 320uC, 30 min at 320uC, and helium carrier gas at 2 mL/min.

Prediction of OsGL1-6 Function
According to the DNA sequence in the NCBI database, OsGL1-6 is present on chromosome 2 of the rice genome and contains nine exons and eight introns ( Figure 1A). The full length of OsGL1-6 cDNA (NM_001055025) is 2,218 bp. The ORF of OsGL1-6 is 1,908 bp, and it encodes a protein with 635 amino acids that has a molecular mass of 71.6 kD and an isoelectric point of 8.64 and that belongs to the fatty aldehyde decarbonylase superfamily.  Conserved domain analysis using the Conserved Domain Database program (http://www.ncbi.nlm.nih.gov/Structure/cdd/ wrpsb.cgi) showed that the N-terminus of OsGL1-6 contains a conserved FA_hydroxylase domain of the fatty acid hydroxylase superfamily. The members of this family are membrane proteins and are involved in the biosynthesis of plant cuticular wax [31]. The prediction of a signal peptide and subcellular localization using SignalP3.0 (http://www.cbs.dtu.dk/services/SignalP/) and PSORT (http://psort.nibb.ac.jp/form.html) showed no signal peptide at the N-terminus of OsGL1-6. However, OsGL1-6 has an N-terminal transmembrane domain and is potentially localized in the ER, peroxisome, and plasma membrane. The OsGL1-6 protein contains one His-rich motif, HX 3 HH, starting at the 147 th amino acid as well as two HX 2 HH motifs beginning at 160 th and 249 th amino acids. This domain forms a bivalent iron ion-binding site that is necessary for the catalytic activities of integrated membrane proteins such as sterol desaturase, acyl desaturase, and alkyl-hydroxylase [31].
Phylogenic analysis of the predicted protein sequences for the six characterized genes with high similarities to OsGL1-6 showed that they can be grouped into two clades ( Figure 1B). OsGL1-6 groups with WDA1 and CER1, whereas OsGL1-1, OsGL1-2, maize GL1, and WAX2 form the other group. Wda1 is involved in wax production in rice anther walls [39]. CER1 functions in the biosynthesis of stem wax and, more specifically, in very-long-chain alkane biosynthesis [31][32][33][34]. The phylogenic analysis results suggested that OsGL1-6 is likely involved in wax biosynthesis.

OsGL1-6 Expression within Epidermal Cells and Vascular Bundles
To investigate the expression pattern of the OsGL1-6 gene, a semi-quantitative RT-PCR analysis was performed. As shown in Figure 2A, OsGL1-6 was expressed in all examined tissues, and the highest expression was observed in the spikelets at the pollen mother cell stage. To confirm the RT-PCR results and further determine the precise expression pattern of OsGL1-6, transgenic rice lines expressing the GUS reporter gene under the control of the OsGL1-6 promoter were generated. A total of 10 independent transformed plants were obtained, and the heterozygotes and homozygotes of three transgenic lines were used for GUS analysis with Zhonghua11 as a control.
Analysis of the GUS activity showed that it was detected in the calli, roots, stems, leaves and spikelets at different developmental stages ( Figure 2B). Notably, GUS activity was much weaker in the subcultured calli than in the differentiated calli which feature vascular bundle differentiation ( Figures 2B-1, 2B-2). During the meiotic stage, GUS was highly expressed in the node of the tassel, rachis, and pedicel but was not expressed in the glumous flowers ( Figure 2B-3). During the uninucleate stage, GUS was expressed in the rachis of the spikelets, pedicel, glumes, and glume veins ( Figure 2B-4). During the binucleate stage, GUS staining was detected in the junction of the primary and secondary rachis, pedicels, and glume veins but was not expressed in the rachis of the spikelets ( Figure 2B-5, 2B-6). GUS activity remained at high levels in the roots, stems and leaves ( Figures 2B-7-14), and its expression was concentrated in the vascular bundle area ( Figures 2B-6-14). Strong GUS activity was also detected in the epidermal cells of the stems and leaves ( Figures 2B-13 and 2B-14). These results indicate that although the expression of OsGL1-6 had tissue and developmental stage specificity, most of it was localized in the epidermal cells and vascular bundle area.

Subcellular Localization of OsGL1-6
Each VLCFA and all of the known wax synthesis-related enzymes are located in the ER [10,21,22,41]. To verify the subcellular localization of OsGL1-6, a 35S:GL1-6-eGFP transient expression vector was constructed and co-transfected with CFP-KDEL into rice leaf sheath protoplasts. OsGL1-6 was fused with eGFP, which expresses a green fluorescent protein. The ER retention signal KDEL was fused with CFP, which expresses a deep blue fluorescent protein. Overlap of the GFP and CFP signals results in a light blue fluorescent signal. Fluorescence microscopy revealed that the ER marker fluorescence and the OsGL1-6-eGFP fluorescence overlapped completely, indicating that OsGL1-6 was located in the ER (Figure 3). Characterization of OsGL1-6 Antisense-RNA Transgenic Plants To investigate the role that OsGL1-6 plays in wax accumulation, an OsGL1-6 antisense-RNA vector was constructed and transformed into rice variety Zhonghua11. A total of nine independent transgenic lines were obtained. Phenotype observation showed that most of the transgenic plants exhibited drooping leaves at the booting stage ( Figure 4A). However, the pollen fertility ( Figure S1) and the seed setting ( Figure 4B) did not change compared with those in WT plants. Three independent transgenic lines, i.e., 2-1, 3-4, and 21-1, each of which has a single insertion, were used in the further studies.
Semi-quantitative RT-PCR and Western blot results showed that the OsGL1-6 expression was significantly decreased in all three transgenic lines (Figures 4C, 4D). We also detected the relative expression levels of Wda1, OsGL1-1 (WSL2), and OsGL1-2 in the transgenic plants, and all three genes showed steady expression compared with those of the WT plants ( Figure S2), implying that the expression of three homologous genes was not affected in the transgenic plants.

Altered Cuticular Wax in OsGL1-6 Antisense-RNA Transgenic Plants
To investigate the effect of OsGL1-6 on wax accumulation, scanning electron microscopy (SEM) was used to perform a detailed observation of the leaf surfaces of the WT and OsGL1-6 antisense-RNA transgenic plants. In the WT plants, the adaxial and abaxial leaf surfaces were both covered with a dense layer of wax crystals, whereas fewer wax crystals were observed on both leaf surfaces in the three transgenic lines (Figures 5A, 5B, and S3). These results suggest that the reduced OsGL1-6 expression affected the wax accumulation on the leaf surfaces of the transgenic plants.
To more accurately determine the differences in the wax accumulation of the leaf surfaces between the WT and transgenic plants, the leaf cuticular wax was extracted using hot chloroform, and the total wax content was then determined by analytical balancing. The treated leaves were subsequently analyzed using SEM to examine the degree of cuticular extraction, and the results showed that the leaf cuticular wax layer was nearly invisible ( Figure S4), indicating that the extraction was thorough. Compared to the WT, the leaf wax contents of the three OsGL1-6 antisense-RNA transgenic lines were reduced by 36-58% (Table 1), indicating that the decreased expression of OsGL1-6 in the transgenic plants directly affected the leaf wax accumulation.
To determine the exact change in leaf wax accumulation between the WT and OsGL1-6 antisense RNA transgenic plants, GC-MS was used to analyze the leaf wax composition both qualitatively and quantitatively. The total cuticular wax contents were reduced by 18% in line 21-1 (Table S2). Compared to the WT plants, the alkane and aldehyde contents in the transgenic plants were decreased by 23% and 21%, respectively. However, the primary alcohol content was increased by 30%, and C30 in particular was increased by 41.2% ( Figure 6; Table S2), which suggests that OsGL1-6 is associated with the decarbonylation pathways in wax biosynthesis.
Altered Cuticle Structure in OsGL1-6 Antisense-RNA Transgenic Plants To investigate the effect of OsGL1-6 on cuticle structure, transmission electron microscopy was used to observe the ultrastructural changes of cross-sections of the WT and transgenic plant leaves. The results showed that cuticle membrane thickness was obviously reduced in the OsGL1-6 antisense-RNA transgenic plants (30-40 nm), compared to the WT plants (80-100 nm) (Figure 7). The cuticular layers on the WT plants were thicker than those on the transgenic plants. However, unlike the compact or indistinct cuticle proper seen in the WT plants, the cuticle proper of the OsGL1-6 antisense-RNA transgenic plants showed a fluffy appearance and disorganized bulges (Figure 7), suggesting that the reduced OsGL1-6 expression in the transgenic plants affected the cuticle structures.
Altered Drought Sensitivity and Cuticular Permeability in OsGL1-6 Antisense-RNA Transgenic Plants Studies have shown that changes in wax accumulation generally lead to changes in plant drought sensitivity and cuticular permeability. As such, drought-sensitivity, water loss, and chlorophyll leaching assays were conducted in both WT and transgenic plants. When the plants were drought stressed, only 4, 4, and 3 of 24 seedlings of the transgenic lines 21-1, 2-1, and 3-4 recovered, respectively, whereas 23 of 24 WT seedlings survived (Figure 8). The drought resistance capacity was also evaluated by measuring water loss from the detached leaves. Compared to the WT plants, the detached leaves from the OsGL1-6 antisense-RNA transgenic plants lost water more rapidly at all examined time points ( Figure 9A). We also checked the leaf cuticular permeability using the chlorophyll leaching method. As shown in Figure 9B, the chlorophyll leached more quickly from the OsGL1-6 antisense-RNA transgenic plant leaves than from the WT plant leaves. These results indicate that the OsGL1-6 antisense-RNA transgenic plants are more sensitive to drought stress than the WT plants, implying that the OsGL1-6 gene may be involved in rice drought protection.

Discussion
In this study, we constructed an OsGL1-6 antisense transgene driven by its cognate promoters and obtained several independent transgenic plants with decreased OsGL1-6 expression. The OsGL1-6 antisense-RNA transgenic rice plants showed drooping leaves at the booting stage as well as a smooth leaf surface, decreased leaf cuticular wax deposition and cuticle thickness, and enhanced drought sensitivity. OsGL1-6 shared high similarity to CER1, WAX2 in Arabidopsis and Wda1, OsGL1-1(WSL2), and OsGL1-2 in rice. Our study showed that the OsGL1-6 antisense-RNA transgenic plants exhibited reduced amounts of leaf total cuticular wax, alkane, and aldehyde but increased amounts of primary alcohol, especially C30, indicating that OsGL1-6 was primarily associated with leaf cuticular wax accumulation. However, the cer1 mutants exhibited a significant reduction in alkanes, secondary alcohols, and ketones as well as an increased aldehyde content [54].
CER1 encodes an aldehyde decarboxylase and functions in both stem wax and pollen development [31][32][33][34]. The leaves and stems of the wax2 mutants showed proportional deficiencies in aldehyde, alkane, secondary alcohol, and ketone contents as well as increased primary alcohol contents, especially the C30 primary alcohols on the stem surface [35]. Compared with the WT plants, the fatty acid, alkane, aldehyde, and primary alcohol contents in the wda1, wsl2, and osgl1-2 mutants were all reduced [39,41,42]. Wda1 was expressed in the epidermis cells of anthers and involved in microspore exine development in the tapetum [39]. OsGL1-1 (WSL2) and OsGL1-2 were involved in leaf wax accumulation and drought resistance [40][41][42]. The different phenotype and effect on wax composition caused by the disruption of OsGL1-6, CER1, WAX2, Wda1, OsGL1-1 (WSL2), and OsGL1-2 function reminded us that they played at least partially distinct roles in plant growth and development as well as in the biosynthesis of cuticular wax, although these genes shared high sequence similarity (identity .30%).
Many wax biosynthesis-related genes -such as CER6, CER4, WAX2, MAH1, Wda1, and OsGL1-1have been reported to be highly expressed in epidermal cells [12,20,22,36,39,40]. GUS staining analysis showed that OsGL1-6 was also highly expressed in epidermal cells ( Figures 2B-13, 2B-14), indicating that OsGL1-6 may be involved in wax biosynthesis. Although GUS staining was also observed in the calli, roots, stems, and different floral organs during different developmental stages ( Figure 2B), the reduced expression of OsGL1-6 caused phenotypic alterations in the leaves but not in the other organs in which it is expressed. OsGL1-6 homologs in the rice genome might compensate for the depletion of the OsGL1-6 gene in these organs.
In addition, GUS activity was stronger in the differentiated calli which have vascular bundle differentiation than that in the subcultured calli ( Figure 2B-1, 2B-2). The expressions of OsGL1-6 in the roots, stems, and leaves were also concentrated within the vascular bundle regions ( Figures 2B-6, 2B-7, 2B-12-14). Some genes involved in wax biosynthesis are reported to be highly expressed in the vascular bundles [21,40,57]. However, it remains unknown why this is the case. Early studies in maize and barley reported that the nsLTPs gene was expressed in the vascular bundles [58,59]. The transcripts of several nsLTPs in common wheat were also observed in the vascular bundles of leaves, roots, and florets of transgenic rice plants detected by the GUS reporter gene [60]. Although in vitro experiments have shown that nonspecific lipid transfer proteins have the capacity to transfer lipids between lipid bilayers [61], their exact biological function remains unclear, especially the roles that they play within the vascular tissues [60]. Our study showed that OsGL1-6 was highly expressed in the vascular bundles, a finding that implies that OsGL1-6 may have functions in addition to contributing to leaf wax accumulation.
One of the important functions of cuticular wax is to prevent non-stomatal water loss from the aerial parts of terrestrial plants [62]; thus, it is closely correlated to plant drought resistance. In our study, the OsGL1-6 antisense-RNA transgenic plants showed increased chlorophyll leaching and water loss rates as well as enhanced drought sensitivity. The wax gene mutants generally exhibited reduced wax accumulation, increased chlorophyll leaching and water loss rates, and enhanced drought sensitivity [35,42,63], whereas the overexpression of wax biosynthesis-related genes generally resulted in increased wax accumulation and enhanced plant drought tolerance [28,29,33,42]. The drought susceptibility of the OsGL1-6 antisense-RNA transgenic plants was in agreement with their deficient cuticles and positively correlated with the reduced accumulation of the leaf cuticular wax, implying its role in drought stress resistance.
This study reported that OsGL1-6 is involved in both leaf cuticular wax accumulation and drought resistance. Thus, genetic modification of OsGL1-6 may have great potential for improving the drought resistance of rice. In addition, the availability of these OsGL1-6 antisense-RNA transgenic plants will be convenient for further studies of the role of waxes in response to other types of environmental stress.