ECERIFERUM1-6A is required for the synthesis of cuticular wax alkanes and promotes drought tolerance in wheat

Abstract

Cuticular waxes cover the aerial surfaces of land plants and protect them from various environmental stresses. Alkanes are major wax components and contribute to plant drought tolerance, but the biosynthesis and regulation of alkanes remain largely unknown in wheat (Triticum aestivum L.). Here, we identified and functionally characterized a key alkane biosynthesis gene ECERIFERUM1-6A (TaCER1-6A) from wheat. The clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated protein 9 (Cas9)-mediated knockout mutation in TaCER1-6A greatly reduced the contents of C27, C29, C31, and C33 alkanes in wheat leaves, while TaCER1-6A overexpression significantly increased the contents of these alkanes in wheat leaves, suggesting that TaCER1-6A is specifically involved in the biosynthesis of C27, C29, C31, and C33 alkanes on wheat leaf surfaces. TaCER1-6A knockout lines exhibited increased cuticle permeability and reduced drought tolerance, whereas TaCER1-6A overexpression lines displayed reduced cuticle permeability and enhanced drought tolerance. TaCER1-6A was highly expressed in flag leaf blades and seedling leaf blades and could respond to abiotic stresses and abscisic acid. TaCER1-6A was located in the endoplasmic reticulum, which is the subcellular compartment responsible for wax biosynthesis. A total of three haplotypes (HapI/II/III) of TaCER1-6A were identified in 43 wheat accessions, and HapI was the dominant haplotype (95%) in these wheat varieties. Additionally, we identified two R2R3-MYB transcription factors TaMYB96-2D and TaMYB96-5D that bound directly to the conserved motif CAACCA in promoters of the cuticular wax biosynthesis genes TaCER1-6A, TaCER1-1A, and fatty acyl-CoA reductase4. Collectively, these results suggest that TaCER1-6A is required for C27, C29, C31, and C33 alkanes biosynthesis and improves drought tolerance in wheat.

TaCER1-6A plays a key role in very-long-chain alkane biosynthesis and drought tolerance in wheat, and two MYB transcription factors bind directly to the TaCER1-6A promoter.

Introduction

A lipidic cuticle covers the aerial surfaces of terrestrial plants and serves as a protective barrier against various biotic and abiotic stresses (Martin and Rose, 2014; Ingram and Nawrath, 2017). The cuticle controls non-stomatal transpiration (Riederer and Schreiber, 2001; Kosma and Jenks, 2007), protects plants against UV radiation damage (Yeats and Rose, 2013), pathogens and insects invasion (Riederer, 2006), and inhibits plant adjacent organs fusions (Sieber et al., 2000; Nawrath, 2006). The cuticle is a complex structure consisting of two major components: cutin and cuticular waxes (Kunst and Samuels, 2003; Pollard et al., 2008). Cutin is an insoluble polyester composed mainly of hydroxy and epoxy fatty acids, diacids, and glycerol (Pollard et al., 2008; Beisson et al., 2012; Fich et al., 2016), whereas cuticular waxes are aliphatic monomers including very-long-chain (VLC) fatty acids, aldehydes, ketones, esters, primary and secondary alcohols, alkanes, as well as variable amounts of triterpenoids, sterols, and flavonoids (Samuels et al., 2008; Kunst and Samuels, 2009; Javelle et al., 2011). The chain lengths of wax constituents typically range from C20 to C32 (Samuels et al., 2008). Alkyl esters with C38–C70 chain lengths may be present (Buschhaus et al., 2007). Much progress has been made in understanding the biosynthesis of cuticular wax in the past decades. In Arabidopsis (Arabidopsis thaliana), wax biosynthesis begins with the formation of C16 and C18 fatty acids in plastids. C16 and C18 fatty acids are then elongated into VLC fatty acids with chain lengths ranging from C20 to C34 by the multi-enzyme fatty acid elongases (FAEs) complex in the endoplasmic reticulum (ER) (Samuels et al., 2008; Kunst and Samuels, 2009; Lee and Suh, 2013). Finally, VLC fatty acids are converted to either primary alcohols and esters by the acyl-reduction pathway or aldehydes, alkanes, secondary alcohols, and ketones by the decarbonylation pathway (Samuels et al., 2008; Bernard et al., 2012).

In many different plants and organs, alkanes are the most abundant components of cuticular waxes (Samuels et al., 2008). For example, alkanes represent 50%–80%, 11%–52%, and 45%–85% of the total wax load in Arabidopsis, wheat (Triticum aestivum L.), and cucumber (Cucumis sativus L.), respectively (Jetter and Kunst, 2008; Bourdenx et al., 2011; Wang et al., 2015a, 2015b). It is generally accepted that alkanes play an important role in plant drought tolerance (Kosma et al., 2009; Seo et al., 2011). Genes involved in alkane biosynthesis have been successfully isolated from several plant species. Arabidopsis ECERIFERUM1 (CER1) encodes an aldehyde decarbonylase that can yield alkanes (Aarts et al., 1995; Bourdenx et al., 2011; Sakuradani et al., 2013). The coexpression of CER1, CER3, and cytochrome b5 isoforms in yeast catalyzed C27–C31 alkanes synthesis (Bernard et al., 2012). Furthermore, CER1 overexpression largely increased the content of C27, C29, and C31 alkanes in Arabidopsis (Bourdenx et al., 2011). Likewise, coexpression of CER1-LIKE1 and CER3 in yeast also yielded shorter chain length alkanes biosynthesis, compared with that of the CER1 and CER3 complex (Pascal et al., 2019). In cucumber, CsCER1 overexpression causes a large increase in C25–C33 alkanes (Wang et al., 2015a). In addition, rice (Oryza sativa L.) OsCER1 (Ni et al., 2018), Kentucky bluegrass (Poa pratensis) PpCER1-2 (Wang et al., 2021), and Brachypodium distachyon BdCER1-8 (Wu et al., 2019) play key roles in wax alkanes biosynthesis.

Wheat is one of the major crops cultivated globally, supplying energy and nutrients for 30% of the world’s population (Mayer et al., 2014). Wheat cuticular waxes are mainly composed of primary alcohols, alkanes, fatty acids, aldehydes, esters, and β- and OH-β-diketones (Adamski et al., 2013; Wang et al., 2015b, 2015c). To date, a series of wax biosynthesis genes have been identified in wheat. Eight fatty acyl-CoA reductase (TaFARs) have been found to be involved in the formation of primary alcohols in wheat leaf wax (Wang et al., 2015b, 2015c, 2016; Chai et al., 2018). Besides, the W1 locus containing a metabolic gene cluster that includes Diketone Metabolism-PKS, Diketone Metabolism-Hydrolase, and Diketone Metabolism-CYP450 is required for β-diketone biosynthesis in wheat (Hen-Avivi et al., 2016). Recently, TaCER1-1A, which is responsible for alkane accumulation, has been isolated in wheat. Heterologous expression of TaCER1-1A in Arabidopsis and rice resulted in significant increases of C25–C35 and C25–C33 alkanes, respectively (Li et al., 2019b). However, the precise biochemical role of TaCER1-1A in wheat is still largely unknown. In this article, a key alkane biosynthesis gene ECERIFERUM1-6A (TaCER1-6A) was isolated from wheat. To reveal the role of TaCER1-6A in alkanes biosynthesis, overexpression and CRISPR/Cas9-mediated gene editing of TaCER1-6A were carried out in wheat, and the wax content, cuticle permeability, and drought tolerance of TaCER1-6A transgenic lines and wild-type (WT) were further analyzed. Finally, the biological function of TaCER1-6A was characterized in detail, and two R2R3-MYB transcription factors, TaMYB96-2D and TaMYB96-5D, were shown to bind directly to the promoter of TaCER1-6A.

Results

Amount and chain length distribution of alkanes on nine different wheat organs

Cuticular waxes were extracted from nine different organs of the wheat cultivar Junmai35 (JM35), including the flag leaf blade (FLB), top second leaf blade (TSB), top third leaf blade (TTB), top fourth leaf blade (TFB), flag leaf sheath (FLS), peduncle, spike, seed, and root. Gas chromatography-mass spectrometry (GC-MS) and GC-flame ionization detection (GC-FID) analysis revealed that the wax composition of the wheat cuticle consisted mainly of fatty acids, aldehydes, alkanes, primary alcohols, esters, and β- and OH-β-diketones (Supplemental Figure S1). The amounts of alkanes ranged from 1.8 to 6.5 μg cm⁻² and from 4.7 to 227.8 μg g⁻¹, and the relative proportions of alkanes ranged from 9.6% to 43.5% in all nine organs tested (Supplemental Figure S1 and Supplemental Table S1). In leaves, there was no substantial difference in the content of alkanes. Interestingly, the amounts of alkanes on FLSs and peduncles were higher than those on leaf blades (Supplemental Figure S1A). In spikes and roots, diketones and alkanes were the major wax constituents (Supplemental Figure S1, B and C), and the waxes on seeds were dominated by alkanes, followed by alkylresorcinols (ARs), fatty acids, aldehydes, alcohols, β-diketone, and methyl ARs (MARs) (Supplemental Figure S1D). Odd-numbered alkanes ranging from C23 to C33 were detected in FLBs, TSBs, TTBs, TFBs, FLSs, peduncles, and spikes, and were dominated by C29 and C31 homologs (Figure 1, A and B). Unexpectedly, both even-numbered and odd-numbered alkanes were observed in seeds and roots, with relatively sharp maxima for C23 and C25 in roots and for C27 and C29 in seeds (Figure 1, C and D).

Figure 1.

Chain length distribution of alkanes in nine different organs of the wheat cultivar JM35. A, FLB (n = 25), TSB (n = 25), TTB (n = 25), TFB (n = 25), FLS (n = 25), and peduncle (n = 30). B, Spike (n = 25). C, Root (n = 30). D, Seed (n = 30). Each value represents the mean of three replicates. Error bars = sd.

The wax crystal morphology was also investigated by scanning electron microscopy (SEM). Two crystal morphologies on the wheat cuticle were identified: platelets and tubules (Supplemental Figure S2). The adaxial sides of FLBs, TSBs, TTBs, and TFBs were covered with plate-shaped crystals (Supplemental Figure S2, A, C, E, and G). In contrast, the abaxial side of FLBs formed tubules (Supplemental Figure S2B). Interestingly, both platelets and tubules were deposited on the abaxial side of TSBs (Supplemental Figure S2D). The abaxial sides of TTBs and TFBs were covered with plate-shaped crystals similar to those found on the adaxial sides (Supplemental Figure S2, F and H). In addition, peduncles, FLSs, and glumes showed tubule-shaped wax crystals (Supplemental Figure S2, I–K).

Identification and sequence analysis of TaCER1-6s from wheat

To identify the alkane synthesis gene, we used Arabidopsis CER1 (NP_171723) to blast the International Wheat Genome Sequencing Consortium survey sequences. Initially, because the wheat genome information was not fully released, only three TaCER1s located on wheat chromosomes 6A, 6B, and 6D were found, and were then designated as TaCER1-6A, TaCER1-6B, and TaCER1-6D, respectively, for further research. When the wheat genome sequence information was fully published in the database (http://plants.ensembl.org/Triticum_aestivum/Info/Index), we finally obtained 21 putative TaCER1 homologous genes using BLASTP search with the AtCER1 amino acid sequence as a query. One of these genes was TaCER1-1A (MK214738), which has been reported as a key gene involved in wax alkane biosynthesis (Li et al., 2019b). The full-length cDNAs of TaCER1-6A, TaCER1-6B, and TaCER1-6D (hereinafter referred to as TaCER1-6s) are 2,341, 2,406, and 2,146 bp, respectively, each containing an open-reading frame of 1,860 bp in length and encoding a 619-amino acid protein (Supplemental Figure S3A). TaCER1-6 proteins contain a FA_hydroxylase domain at the N-terminus and a Wax2_C domain at the C-terminus and possess three conserved His-rich motifs (HX₃HH, HX₂HH, and HX₂HH) (Supplemental Figure S3B), which play an important role in electron transfer during alkane synthesis (Bernard et al., 2012). The predicted TaCER1-6A protein molecular mass was 71.4 kDa, which was further confirmed by SDS–PAGE analysis of the histidine-tagged TaCER1-6A protein (Supplemental Figure S4).

Furthermore, TaCER1-6A shared 55% identity with AtCER1 and 62% identity with TaCER1-1A. A neighbor-joining (NJ) phylogenetic tree was constructed to investigate the evolutionary relationship between TaCER1-6s and 21 other CER1 homologs. The phylogenetic tree was grouped into two clades. Interestingly, all dicotyledonous CER1s were grouped into the first clade, and all monocotyledonous CER1s formed the second clade (Supplemental Figure S5). TaCER1-6 proteins were more closely related to sorghum (Sorghum bicolor L.) SbCER1 and rice OsCER1, which have been implicated in wax alkane biosynthesis (Ni et al., 2018), indicating that they are homologous genes and may have similar biological functions related to wax alkane biosynthesis.

Expression pattern and subcellular localization of TaCER1-6s and TaCER1-1A

To understand the biological function of TaCER1-6s related to wax biosynthesis in wheat, we first analyzed the expression profile of TaCER1-6s and TaCER1-1A in different wheat organs by reverse transcription quantitative PCR (RT-qPCR). TaCER1-6A, TaCER1-6B, and TaCER1-6D transcripts were detected in all examined organs. The expression level of TaCER1-6A was high in FLBs and seedling leaf blades; medium in anthers and roots; and low in TSBs, glumes, FLSs, internodes, nodes, awns, seeds, and peduncles (Figure 2A). TaCER1-6B was expressed at the highest level in seedling leaf blades, and at very low levels in glumes, FLBs, TSBs, roots, and seeds. TaCER1-6D was mainly expressed in anthers, peduncles, seedling leaf blades, FLSs, internodes, nodes, and awns. TaCER1-1A was highly expressed in FLBs and peduncles, but not in roots and seedling leaf blades (Figure 2A). We then investigated the subcellular localization of TaCER1-6s and TaCER1-1A proteins. Full-length CDSs of TaCER1-6s and TaCER1-1A were fused to the N-terminal of green fluorescent protein (GFP) under the control of the cauliflower mosaic virus (CaMV) 35S promoter. The TaCER1-GFP fusion proteins and the ER marker PCX-DR-CTR3 were co-transfected into Nicotiana benthamiana leaves via Agrobacterium tumefaciens infiltration. Confocal microscopic observations revealed that TaCER1-GFPs completely colocalized with the ER marker signal (Figure 2B), suggesting that TaCER1-6A, TaCER1-6B, TaCER1-6D, and TaCER1-1A were located in the ER and further indicating that TaCER1-6A, TaCER1-6B, and TaCER1-6D may participate in wax biosynthesis.

Figure 2.

Expression and subcellular localization analysis of TaCER1-6s and TaCER1-1A. A, Expression analysis of TaCER1-6A, TaCER1-6B, TaCER1-6D, and TaCER1-1A in different organs of wheat cv JM35 by RT-qPCR. All genes were normalized using wheat Actin and Tubulin genes as internal controls. Error bars represent SD of three independent experiments. B, Subcellular localization of TaCER1-6s and TaCER1-1A proteins in N. benthamiana leaf epidermal cells. TaCER1-6A-GFP, TaCER1-6B-GFP, TaCER1-6D-GFP, and TaCER1-1A-GFP were coexpressed with the ER marker PCX-DR-CTR3 in N. benthamiana leaves, respectively. Bars = 25 μm.

CRISPR/Cas9-mediated mutagenesis and overexpression of TaCER1-6A affected alkane production on the leaf surface in transgenic wheat

To understand the biological function of TaCER1-6s in wax biosynthesis in wheat, TaCER1-6A was chosen for further analysis. CRISPR/Cas9-mediated gene editing was first used to knock out TaCER1-6A in wheat. A CRISPR/Cas9 construct was designed to target a 20-bp sequence for the single-guide RNA (sgRNA) with a protospacer adjacent motif (PAM) at the eighth exon of TaCER1-6A (Figure 3A). Then, the resulting construct was introduced into embryogenic callus tissue of wheat cv Fielder by A. tumefaciens-mediated transformation, and 16 independent transgenic T₀ plants were obtained. To detect mutations in the target region, genomic DNA from independent transgenic plants was amplified, and 15 individual clones from each PCR product were randomly selected for sequencing. The results showed that insertions and deletions were present in transgenic T₀ plants. Four types of deletions were detected in plant T₀-16. Among them, 5- and 9-bp deletions occurred in TaCER1-6A, and 3-bp deletion occurred in TaCER1-6B, and 4-bp deletion occurred in TaCER1-6D, suggesting that the three TaCER1-6 copies located on wheat chromosome 6 were all knocked out in plant T₀-16. Moreover, 3- and 9-bp deletions detected in plant T₀-6, as well as a 6-bp deletion and a 1-bp insertion detected in plant T₀-15 occurred in TaCER1-6A (Figure 3B). Because the six different mutation types all occurred in TaCER1-6A, TaCER1-6B, and TaCER1-6D, the other 18 TaCER1 homologs were not affected by CRISPR/Cas9-mediated gene editing. Simultaneously, the coding region of TaCER1-6A was overexpressed in wheat cv Fielder under the control of the CaMV 35S promoter. Cuticular wax contents of seedling leaves and flowering leaves from WT plants, T₁ transgenic knockout lines (KO#6 and KO#16), and overexpression lines (OE#10 and OE#13), which showed higher expression levels of TaCER1-6A than WT (Supplemental Figure S6A), were analyzed in detail by GC-MS and GC-FID.

Figure 3.

CRISPR/Cas9-mediated genome editing of TaCER1-6A in wheat. A, Schematic diagram of TaCER1-6A showing the target site in the eighth exon. The untranslated regions (UTRs) and exons are indicated by boxes, and introns are indicated by thin lines. B, Representative mutation types of TaCER1-6A from T₀ transgenic knockout plants. Deletions are indicated by dashes and a nucleotide insertion is detected in plant T₀-15. The numbers on the right side preceded by plus or minus signs represent the number of nucleotides deleted or inserted.

In seedling wheat leaves, cuticular wax consisted of fatty acids, aldehydes, alkanes, primary alcohols, and esters. There was no substantial difference in plant morphology or the total load among WT, T₁ transgenic knockout lines, and overexpression lines (Supplemental Figure S6, B–D). Compared with WT plants, the knockout lines KO#6 and KO#16 exhibited 31% and 43% decreases in the contents of alkanes, respectively, while the overexpression lines OE#10 and OE#13 showed 32% and 30% increases in the amounts of alkanes, respectively (Figure 4A). In particular, the C27, C29, C31, and C33 alkanes contents of knockout lines significantly decreased, but the C27, C29, C31, and C33 alkanes contents of overexpression lines largely increased when compared with those of WT (Figure 4B). For example, the knockout of TaCER1-6A led to 44%, 61%, 49%, and 35% decreases in C27, C29, C31, and C33 alkanes of KO#16, respectively, whereas the overexpression of TaCER1-6A caused a large increase of C27, C29, C31, and C33 alkanes (70%, 47%, 20%, and 22%, respectively) in OE#13 (Figure 4B). In contrast, the contents of C23 and C25 alkanes were not affected in knockout lines and overexpression lines compared with those in WT. Additionally, the amounts of other wax components, such as fatty acids, aldehydes, primary alcohols, and esters, showed no differences among WT, knockout lines, and overexpression lines.

Figure 4.

Cuticular wax analysis of leaves from WT plants, TaCER1-6A knockout lines (KO#6 and KO#16), and TaCER1-6A-overexpressing lines (OE#10 and OE#13). At the seedling stage and the flowering stage, leaves of WT plants, T₁ transgenic knockout lines, and overexpression lines grown in the greenhouse were used for analysis of cuticular wax load and composition. A, Cuticular wax content and composition on seedling leaves (n = 30). B, Alkanes content of individual chain lengths on seedling leaves (n = 30). C, Cuticular wax load and composition on flowering flag leaves (n = 25). D, Alkanes content of individual chain lengths on flag leaves (n = 25). Statistical differences between WT plants and transgenic lines are indicated by asterisks using Student’s t test (*P < 0.05, **P < 0.01, ***P < 0.001). Each value represents the mean of three independent measurements. Error bars indicate sd. Diketones include β-diketone and hydroxy-β-diketone.

The wax contents of flowering flag leaves were further analyzed. Similarly, no significant differences in the total wax load were observed among WT, knockout lines, and overexpression lines. The contents of alkanes were greatly decreased in KO#6 and KO#16 (41% and 33% decreases, respectively), whereas the contents of alkanes were significantly higher in OE#10 and OE#13 (39% and 70% increases, respectively) (Figure 4C). Likewise, TaCER1-6A knockout caused large decreases in C27, C29, C31, and C33 alkanes (41%, 42%, 42%, and 38% decreases in KO#6 and 46%, 32%, 30%, and 30% decreases in KO#16, respectively), while the contents of C27, C29, C31, and C33 alkanes in overexpression lines significantly increased (48%, 39%, 40%, and 42% increases in OE#10 and 81%, 82%, 47%, and 38% increases in OE#13, respectively) compared with those in WT (Figure 4D). In contrast, fatty acids, aldehydes, primary alcohols, esters, and β-diketones (β- and OH-β-diketones) amounts were only slightly altered in knockout lines and overexpression lines compared with those in WT. Overall, these results strongly suggested that TaCER1-6A plays an important role in VLC alkane biosynthesis and has substrate specificity for alkanes ranging from C27 to C33 on wheat leaf surfaces.

Chromosome localization and haplotype analysis of TaCER1-6A

TaCER1-6A is predicted to be located on chromosome 6A. To test this hypothesis, four genome-specific primer pairs were designed to amplify the genomic DNA from a set of Chinese Spring (CS) nullisomic–tetrasomic (N–T) lines, which lacked one pair of chromosomes and added an extra pair of chromosomes. Four 364-, 898-, 560-, and 567-bp PCR fragments were amplified from CS and 11 N–T lines, and all of these fragments disappeared in N6AT6B (Figure 5A), which lacked the entire chromosome 6A and had four copies of chromosome 6B, suggesting that TaCER1-6A is located on wheat chromosome 6A. Next, we further compared the wax contents of seedling leaves and flag leaves between CS and the N6AT6B line. Compared with CS seedling leaves, the alkanes content in N6AT6B leaves decreased by 28% (with 35%, 20%, 33%, and 27% decreases in C27, C29, C31, and C33 alkanes, respectively) (Figure 5, B and C). However, the total wax load and the amounts of other components including fatty acids, aldehydes, primary alcohols, and esters were almost unaffected between CS and N6AT6B. Similarly, in flag leaves, the alkanes content of N6AT6B was reduced by 33% compared with that of CS plants (Figure 5D). This decrease was due to a reduction of C27, C29, C31, and C33 alkanes (35%, 33%, 36%, and 23% decreases, respectively; Figure 5E). In contrast, no significant differences were observed in the contents of other wax components or the total wax coverage between CS and N6AT6B. These findings further support the conclusion that TaCER1-6A is involved in C27–C33 VLC alkanes biosynthesis in wheat leaves.

Figure 5.

Chromosomal localization and haplotype analysis of TaCER1-6A. A, Chromosome mapping of TaCER1-6A using the CS and its N–T lines. N4AT4B represents nullisomic 4A-tetrasomic 4B, and the remaining lines are named accordingly. The sizes of PCR products and primer names are shown on the left. M, DNA ladder DL2000. B and D, Composition of cuticular wax on seedling leaves (B) (n = 30) and flag leaves (D) (n = 25) between the CS and the N6AT6B line. C and E, Alkanes content with different chain lengths on seedling leaves (C) (n = 30) and flag leaves (E) (n = 25) between the CS and the N6AT6B line. The content of diketones is the sum of β-diketone and hydroxy-β-diketone. Significance was assessed by Student’s t test (*P < 0.05, **P < 0.01, and ***P < 0.001). Each value represents the mean of three replicates. Error bars = sd. F, Haplotype analysis of TaCER1-6A. The positions of SNPs are indicated in TaCER1-6A.

To analyze the haplotypes of TaCER1-6A, we sequenced the coding region of TaCER1-6A from 43 wheat accessions with wide geographic distribution and diverse genetic backgrounds, including 40 winter wheat varieties and 3 spring wheat varieties (Supplemental Table S2). As shown in Figure 5F, only three single-nucleotide polymorphisms (SNPs) were detected in the TaCER1-6A coding region, which resulted in three distinct haplotypes (HapI/II/III) in the common wheat accessions surveyed. SNP1 and SNP3 led to an amino acid substitution at 224 (Phe to Cys) and 475 (Ile to Val), respectively. These substitutions did not create a premature stop codon and were far away from the histidine boxes. However, SNP2 did not cause an amino acid substitution. Interestingly, HapI was present in 41 wheat accessions, and HapII or HapIII was only present in one wheat accession (Figure 5F), indicating that HapI was the dominant haplotype in modern cultivated wheat varieties.

TaCER1-6A expression is induced by various abiotic stresses and abscisic acid

Environmental factors such as light, moisture, and temperature are usually considered to affect cuticular waxes deposition of plants (Kosma et al., 2009). We examined whether TaCER1-6A is regulated by abiotic stresses. Six-week-old wheat cv JM35 seedlings were subjected to drought (exposure to air), polyethylene glycol (PEG), salt, and cold treatments. RT-qPCR analysis confirmed that TaCER1-6A expression was significantly increased after 2 h drought treatment and then began to decrease from 4 h until 12 h. Similarly, under the PEG stress condition, TaCER1-6A transcript reached a peak at 6 h (an approximately eight-fold increase) and was reduced at 12 h. When plants were treated with cold stress for 6 h and salt stress for 4 h, the expression of TaCER1-6A significantly increased by five- and three-fold, respectively (Figure 6A). Because TaCER1-6A expression can be induced by abiotic stresses, and abscisic acid (ABA) is an important signal in regulating the plant’s response to abiotic stress, we also investigated the effect of exogenous ABA application on TaCER1-6A expression. Treatment with 100 μM ABA for 2 and 3 h led to 2.5- and 2.7-fold increases in TaCER1-6A transcript abundance, respectively, compared with untreated plants (Figure 6A). These observations suggested that TaCER1-6A expression can be induced by abiotic stress in an ABA-dependent manner.

Figure 6.

TaCER1-6A is induced by various abiotic stresses. A, Expression analysis of TaCER1-6A under abiotic stress and ABA treatments. Six-week-old seedlings of wheat cv JM35 soil-grown in a greenhouse were subjected to drought (exposure to air), PEG (20% PEG6000), cold (4°C), salt (200 mM NaCl), and ABA (100 μM ABA) treatments. Error bars indicate sd based on three independent experiments. Different lowercase letters above the bars indicate significant differences at P < 0.05, as determined by Duncan’s multiple range test. B and C, Cuticular wax content on 6-week-old seedling leaves (B) (n = 30) and flowering flag leaves (C) (n = 25) of JM35 plants under drought and salt stress conditions. Dik, β-diketone and hydroxy-β-diketone; Unk, unknown compounds; CK, the well-watered JM35 plants; Drought, after 10 days of water deprivation on JM35 plants; Salt, after 5 days of 200 mM NaCl treatment on JM35 plants. Each value is the average from three independent measurements. Error bars = sd. Asterisks indicate significant differences (Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001).

Next, we analyzed the effects of abiotic stress on cuticular wax accumulation. Six-week-old wheat cv JM35 seedlings were treated with drought and high salinity, and cuticular wax of leaf blades was measured by GC-MS and GC-FID. Compared with non-treated (CK) plants, drought-treated and salt-treated plants showed 34% and 24% increases in the total wax amount, respectively (Figure 6B). Moreover, drought and salt treatments led to significant increases in the contents of alkanes (84% and 33% increases, respectively), primary alcohols (30% and 23% increases, respectively), fatty acids (53% and 34% increases, respectively), aldehydes (53% and 35% increases, respectively), and esters (57% and 34% increases, respectively), including most homologs with different chain lengths (Figure 6B). Additionally, flowering JM35 plants were subjected to drought and salt stresses treatments. Likewise, drought and salt stresses significantly increased the total wax amounts and the contents of most wax components such as alkanes (42% and 29% increases, respectively), primary alcohols (38% and 29% increases, respectively), β-diketones (50% and 54% increases, respectively), fatty acids (91% and 89% increases, respectively), aldehydes (45% and 36% increases, respectively), and esters (91% and 41% increases, respectively) with different chain lengths on flag leaves (Figure 6C). Taken together, our results indicated that abiotic stresses could trigger the expression of TaCER1-6A and lead to the accumulation of cuticular wax alkanes. Additionally, more wax synthesis genes were likely induced under drought and salt treatments and resulted in the production of other cuticular wax components.

Altered cuticle permeability and drought tolerance in TaCER1-6A knockout lines and overexpression lines

Previous studies have shown that cuticle permeability is associated with the components and amounts of cuticular waxes (Seo et al., 2011; Wang et al., 2021). Therefore, we measured cuticle permeability by chlorophyll leaching and water loss assays of flowering flag leaves from WT plants, knockout line KO#16, and overexpression line OE#10. The results indicated that the knockout line KO#16 exhibited a faster rate of chlorophyll leaching and water loss than WT plants, while the overexpression line OE#10 showed a lower rate of chlorophyll leaching and water loss than WT plants (Figure 7, A and B), indicating a drastic increase of cuticle permeability in the transgenic knockout line and a large reduction of cuticle permeability in the overexpression line. We next compared the drought tolerance of WT, KO#16, and OE#10 plants at the seedling stage. After drought stress treatment for 15 days, the KO#16 plants showed a wilted and curled phenotype, while the OE#10 plants appeared more green and fresh than the WT plants (Figure 7C). These results suggested that TaCER1-6A overexpression reduced cuticle permeability and substantially improved drought tolerance in wheat.

Figure 7.

Cuticle permeability and drought tolerance of WT plants, transgenic knockout line KO#16, and overexpression line OE#10. A, Chlorophyll leaching assay. Chlorophyll leaching rates are expressed as percentages of total chlorophyll extracted at 24 h after immersion in 80% ethanol solution (n = 30). The data represent the means ± sd of three replicates. B, Water loss assay. Flag leaves were excised and placed immediately in water for 3 h in the dark (n = 25). The leaves were thereafter weighed using a microbalance at individual time points. Results were obtained from three replicates. Error bars = sd. C, Drought tolerance test of WT, KO#16, and OE#10 plants. Six-week-old plants were exposed to 10 and 15 days of water deprivation (n > 12).

TaMYB96-2D and TaMYB96-5D are direct regulators of the TaCER1-6A promoter

R2R3-type MYB transcription factors, such as AtMYB94 and AtMYB96, have been reported to positively regulate the expression of cuticular wax biosynthesis genes (Seo et al., 2011; Lee and Suh, 2015b; Lee et al., 2016). We speculated that similar MYBs involved in wax regulation may exist in wheat. To verify this hypothesis, the conserved MYB DNA-binding domain (Pfam accession number PF00249) was used as a query to BLAST search for putative R2R3-MYBs in wheat database (http://plants.ensembl.org/Triticum_aestivum/Info/Index). A total of 908 sequences were identified as putative R2R3-MYBs in the wheat genome, and were further used to construct NJ phylogenetic trees with AtMYB94/96. Six TaMYB96s were grouped into a small subclade with AtMYB94/96 (Supplemental Figure S7). It seems likely that these six wheat R2R3-MYBs might act as functional homologs of AtMYB94/96. Since TraesCS5D02G234800 (renamed TaMYB96-5D) was most closely related to AtMYB94/96, and exhibited higher amino acid identity than TraesCS5A02G227400 (renamed TaMYB96-5A) and TraesCS5B02G226100 (renamed TaMYB96-5B) with AtMYB94/96, TaMYB96-5D and TraesCS2D02G163700 (renamed TaMYB96-2D) were selected to test their interaction with the promoters of three identified wax biosynthesis genes TaCER1-6A, fatty acyl-CoA reductase4 (TaFAR4), and TaCER1-1A. To this end, we first performed a dual-luciferase (LUC) assay. As shown in Figure 8A, the coexpression of TaMYB96-2D/5D effectors with ProTaCER1-6A-LUC, ProTaFAR4-LUC, or ProTaCER1-1A-LUC reporter led to significant increases in LUC activities compared with that in the control, indicating that TaMYB96-2D/5D can directly bind to the promoters of TaCER1-6A, TaFAR4, and TaCER1-1A.

Figure 8.

TaMYB96-2D and TaMYB96-5D activate TaCER1-6A, TaFAR4, and TaCER1-1A expression by directly binding to their promoters. A, Dual-LUC assay. The LUC:REN ratio of the empty vector was set to 1. Data are presented as means of three replicates, and error bars represent the sd. Different lowercase letters above the bars indicate significant differences at P < 0.05, as determined by Duncan’s multiple range test. B, Y1H analysis. Three predicted MYB-binding motifs, motif1 (CAACCA), motif2 (CAACTG), and motif3 (TAACCA), were from the promoter regions of TaFAR4, TaCER1-6A, and TaCER1-1A genes. Transformed yeast cells were grown on selective medium (SD/-Ura/-Leu) supplemented with 0.4 mg L⁻¹ AbA. Cotransformation of pGADT7 and pABAi-motif1/2/3 was used as negative control. C–E, GUS activity analysis indicated that TaMYB96-2D/5D activated TaCER1-6A expression. C, Schematic diagrams of the GUS reporter vector containing the TaCER1-6A promoter and the TaMYB96-2D/5D effector vectors. D and E, GUS staining analysis (D) and relative quantitative analysis of GUS activity (E) after reporter and effector strains were co-infiltrated into N. benthamiana. Values are presented as the mean ± sd based on three repetitions. Asterisks indicate significant differences (Student’s t test, **P < 0.01, ***P < 0.001).

Sequence analysis revealed that the promoters of TaCER1-6A, TaFAR4, and TaCER1-1A contain three potential MYB-binding motifs (“CAACCA” motif1, “CAACTG” motif2, and “TAACCA” motif3). To determine the exact binding sites of TaMYB96-2D/5D, yeast one-hybrid (Y1H) was further performed. When pGADT7-TaMYB96-2D/5D was coexpressed with pAbAi-motif1, the yeast cells grew normally on SD/-Ura/-Leu/plates containing 0.4 mg L⁻¹ aureobasidin A (AbA), while the growth of yeast cells harboring pAbAi-motif2 or pAbAi-motif3 was inhibited when cotransformed with pGADT7-TaMYB96-2D/5D (Figure 8B). These data suggested that TaMYB96-2D/5D bound specifically to “CAACCA” motif1 and failed to bind to “CAACTG” motif2 and “TAACCA” motif3. Finally, a β-glucuronidase (GUS) activation assay was performed to investigate the regulation of the TaCER1-6A promoter by TaMYB96-2D/5D. When N. benthamiana leaves were cotransformed with the ProTaCER1-6A::GUS and Pro35S::TaMYB96-2D/5D constructs, increased GUS staining and significantly increased GUS activity were observed (Figure 8, C–E). The results verified that TaMYB96-2D/5D activated the transcription of TaCER1-6A.

Discussion

TaCER1-6A is required for the biosynthesis of wax VLC alkanes in wheat leaves

Previous studies have shown that AtCER1 is involved in VLC alkane biosynthesis in Arabidopsis (Aarts et al., 1995; Bourdenx et al., 2011; Sakuradani et al., 2013). In this study, a homologous gene of AtCER1, named TaCER1-6A, was identified from wheat, which shares 55% amino acid identity with AtCER1. Similar to previously reported AtCER1 orthologs including rice OsCER1 (Ni et al., 2018), wheat TaCER1-1A (Li et al., 2019b), Brachypodium BdCER1-8 (Wu et al., 2019), cucumber CsCER1 (Wang et al., 2015a), and P. pratensis PpCER1 (Wang et al., 2021), TaCER1-6A also contains three specific His-rich motifs that are essential for VLC alkane biosynthesis (Bernard et al., 2012). Accordingly, we speculated that TaCER1-6A may have a similar function in VLC alkane biosynthesis. To directly test the biochemical function and confirm the substrate specificity of TaCER1-6A in cuticular wax metabolism, TaCER1-6A was overexpressed and knocked out via CRISPR/Cas9-mediated gene editing in the wheat cv Fielder. In seedling leaves and flowering flag leaves, the total alkane contents decreased significantly in transgenic knockout lines but drastically increased in overexpression lines compared with WT plants, suggesting that TaCER1-6A plays a key role in alkanes production in wheat leaves. Further detailed investigations of chain length indicated that the C27–C33 alkanes of TaCER1-6A knockout lines were much lower than those of WT, whereas the C27–C33 alkanes of TaCER1-6A overexpression lines were much higher than those of WT. Thus, TaCER1-6A has distinct substrate specificities and preferentially catalyzes the biosynthesis of C27–C33 alkanes, which was further supported by the cuticular wax examination of the CS N–T line N6AT6B. Similar results have been previously reported for another wheat alkane synthesis gene TaCER1-1A, which led to obvious increases in C25–C35 and C25–C33 alkanes in transgenic Arabidopsis and rice, respectively (Li et al., 2019b). Likewise, AtCER1 and CsCER1 are mainly involved in the biosynthesis of C27–C33 alkanes (Bourdenx et al., 2011; Wang et al., 2015a). In contrast, AtCER1-LIKE1 and OsCER1 primarily catalyze the biosynthesis of C25 and C27 alkanes with chain-length substrate specificities shorter than those of AtCER1 and TaCER1-6A (Ni et al., 2018; Pascal et al., 2019). Interestingly, heterologous expression of PpCER1 in Brachypodium resulted in a significant increase in C25 and C27 alkanes and a strong decrease in C29 and C31 alkanes (Wang et al., 2021). It seems likely that CER1 homologous genes from various species possess different substrate specificities depending on the host expression system. Mutation analysis revealed that three TaCER1-6 copies TaCER1-6A, TaCER1-6B, and TaCER1-6D, located on wheat chromosomes 6A, 6B, and 6D, respectively, were all knocked out in plant T₀-16, and only TaCER1-6A was knocked out in plant T₀-6. Accordingly, we further compared the difference of alkanes content between T₁ transgenic knockout lines KO#6 and KO#16. The results suggested that the total alkanes content of KO#16 seedling leaves was significantly lower than that of KO#6 seedling leaves, but there was no significant difference in total alkanes content of flag leaves between KO#6 and KO#16. It is likely that knockout of multiple TaCER1-6 homologs led to a higher reduction in the content of alkanes at least in wheat seedling leaves.

Aldehydes are considered as intermediates in the synthesis of alkanes (Aarts et al., 1995). However, the overexpression and knockout of TaCER1-6A did not significantly affect the content of aldehydes in wheat leaves. Similarly, AtCER1 overexpression and CsCER1 RNA interference specifically altered the content of alkanes, with only minor modifications of other wax components (Bourdenx et al., 2011; Wang et al., 2015a). This led us to speculate that TaCER1-6A may directly catalyze the conversion of VLC acyl-CoAs to VLC alkanes without aldehydes production. The content of alkanes was greatly influenced, but no obvious differences in the total wax amount were observed among WT plants, knockout lines, and overexpression lines. Because alkanes account for approximately 20% of total wax on wheat leaves, the increased or decreased alkanes were not sufficient to affect the total amount in transgenic wheat lines. In contrast, alkanes accounted for approximately 50% of stem total wax and approximately 70% of leaf total wax in Arabidopsis (Bernard and Joubès, 2013), and overexpression of AtCER1 led to large increases in the total wax load of stems and leaves (Bourdenx et al., 2011). In TaCER1-6A knockout lines and N–T line N6AT6B, the content of alkanes was not completely absent, suggesting that other CER1 homologous genes were involved in the synthesis of the remaining alkanes. Indeed, 21 members of the TaCER1 gene family were identified in the wheat genome. Arabidopsis CER1 and CER1-LIKE1 can interact with CER3 to form an enzymatic complex that catalyzes the biosynthesis of alkanes (Bernard et al., 2012; Pascal et al., 2019). This lead us to speculate that CER3 homologs may exist in wheat, and the further identification of TaCER3 homologs will provide more insight into the cuticular alkane-forming pathway in wheat.

TaCER1-6A was induced by abiotic stresses in an ABA-dependent manner and enhanced drought tolerance in wheat

Cuticular waxes are considered to play an important role in the interactions of plants with the environment, and wax accumulation is also influenced by a variety of environmental factors, such as water deficiency, light, and low temperature (Kosma et al., 2009). TaCER1-6A expression was induced by drought, NaCl, and cold stresses. In agreement with our results, most wax-associated genes from different species, including AtCER1 (Bourdenx et al., 2011), AtCER1-LIKE1 (Pascal et al., 2019), CsCER1 (Wang et al., 2015a), AtKCS2/AtDAISY (Lee et al., 2009), PpCER1-2 (Wang et al., 2021), TaCER1-1A (Li et al., 2019b), BdCER1-8 (Wu et al., 2019), three BdFARs (Wang et al., 2018a), and eight TaFARs (Wang et al., 2015b, 2015c, 2016; Chai et al., 2018), were up-regulated under various abiotic stresses. TaCER1-6A could also be induced by ABA, suggesting that TaCER1-6A is induced in response to abiotic stress in an ABA-dependent signaling pathway. In wheat, eight TaFARs involved in the primary alcohol-forming pathway are induced by abiotic stress in an ABA-dependent manner (Wang et al., 2015b, 2015c, 2016; Chai et al., 2018). We therefore propose that the genes involved in the two wax formation pathways may function in the ABA-dependent stress response signaling pathways in wheat. Meanwhile, consistent with the higher expression levels of TaCER1-6A under drought and salt stresses, drought- and salt-treated wheat plants contained higher amounts of total wax and each wax component than the untreated control plants (CK). Similar results have been reported in previous studies (Wang et al., 2015b, 2015c, 2016, 2018; Li et al., 2019b; Wu et al., 2019). Consequently, it was concluded that environmental signals induced wax production and deposition by upregulating the expression of key wax biosynthesis genes.

Wax alkanes are considered to play an important role in plant drought tolerance (Kosma et al., 2009; Bourdenx et al., 2011). In this study, the overexpression of TaCER1-6A caused an obvious increase in the total alkanes amount and a significant reduction in cuticle permeability and led to enhanced drought tolerance in wheat. However, TaCER1-6A knockout lines exhibited lower total alkanes with higher cuticle permeability and reduced drought tolerance, suggesting that TaCER1-6A is involved in alkane formation and affects cuticle permeability and drought tolerance in wheat. These findings further support the conclusion that alkanes play a key role in plant drought stress tolerance. Similar results have been previously reported for other alkane biosynthesis genes, such as AtCER1 (Bourdenx et al., 2011), CsCER1 (Wang et al., 2015a), PpCER1-2 (Wang et al., 2021), TaCER1-1A (Li et al., 2019b), and BdCER1-8 (Wu et al., 2019). Notably, there was no significant difference in the total wax content among WT plants and TaCER1-6A overexpression lines and knockout lines. This led us to speculate that the total wax amount may not always be related to plant drought tolerance, and specific wax components may mainly contribute to plant drought tolerance. In fact, PpCER1-2 overexpression in B. distachyon led to an increase in the alkane amount and a decrease in the amounts of total wax and primary alcohol, whereas PpCER1-2 overexpression lines displayed reduced cuticle permeability and the enhanced drought tolerance (Wang et al., 2021). In addition, we also compared the yield-related traits between transgenic lines and WT plants. Six agronomic traits including plant height, spike length, spike number per plant, grain number per spike, 1,000-grain weight, and grain yield per plant were further investigated among WT plants, knockout line KO#16, and overexpression line OE#10. As shown in Supplemental Figure S8, no significant differences in plant height, spike length, spike number per plant, grain number per spike, 1,000-grain weight, and grain yield per plant were observed among WT, KO#16, and OE#10, indicating that wheat plant height and yield-related traits were not substantially affected by CRISPR/Cas9-mediated mutation and overexpression of TaCER1-6A at least in rainfed field experiments. Because the overexpression and knockout of TaCER1-6A did not affect the morphology of wheat plants, TaCER1-6A could be used to generate new drought-tolerant wheat cultivars through a transgenic breeding strategy in the future.

TaCER1-6A is directly regulated by TaMYB96-2D and TaMYB96-5D

Tremendous progress has been made in understanding wax synthesis, and many wax-related genes have been identified. So far, the transcriptional regulation of wax biosynthesis has been reported in many plant species, and several transcription factors have been reported to positively or negatively modulate the expression of wax biosynthesis genes. In Arabidopsis, the R2R3-type transcription factor MYB30 is involved in wax biosynthesis by modulating the expression of 3-ketoacyl-CoA synthase1 (KCS1) and FIDDLEHEAD (Raffaele et al., 2008). MYB96 specifically binds to the promoters of KCS1, 3-ketoacyl-CoA synthase2 (KCS2/DAISY), 3-ketoacyl-CoA synthase6, 3-ketoacyl CoA reductase1, CER3, and wax synthase/acyl-CoA:diacylglycerol acyltransferase1 (WSD1), and regulates the expression of these target genes (Seo et al., 2011; Lee et al., 2016). The MYB96 homolog MYB94 directly activates the expression of KCS2/DAISY, CER2, FAR3, WSD1, and enoyl-CoA reductase (ECR) by interacting with the promoters of these genes (Lee and Suh, 2015b; Lee et al., 2016). Similarly, Arabidopsis WRINKLED4 (Park et al., 2016) and RAP2.4 (Yang et al., 2020) and maize (Zea mays) ZmFDL1/MYB94 (Castorina et al., 2020) positively regulate cuticular wax biosynthesis by directly binding to the promoters of wax biosynthesis genes. In addition, the APETALA2/ethylene responsive factor-type transcription factor DEWAX negatively regulates wax biosynthesis via directly interacting with wax biosynthetic gene promoters (Go et al., 2014; Kim et al., 2018; Li et al., 2019a). In this study, the two R2R3-type transcription factors TaMYB96-2D/5D directly interacted with the promoter regions of TaCER1-6A, TaCER1-1A, and TaFAR4, which participate in wax alkanes and primary alcohols biosynthesis. TaMYB96-2D/5D probably plays an important role in both the acyl-reduction pathway and the decarbonylation pathway. Additionally, TaMYB96-2D/5D specifically bound to the motif CAACCA, but not the motifs CAACTG and TAACCA. Indeed, the conserved motif CAACCA was present in the promoter regions of eight TaFARs, TaCER1-6A, and TaCER1-1A genes, leading us to hypothesize that all these genes can be transcriptionally regulated by TaMYB96-2D/5D. It seems very likely that TaMYB96-2D/5D possess the same MYB binding motif found in the promoters of wax biosynthesis genes. Similar to our inference, Lee et al. (2016) pointed out that AtMYB94 and AtMYB96 bind to the same MYB consensus motifs in wax biosynthesis gene promoters and regulate cuticular wax biosynthesis. Recently, a R2R3-type MYB transcription factor TaEPBM1-A was reported to directly interact with TaECR promoters and activate their expression in wheat (Kong et al., 2020). In fact, sequence alignment suggested that TaMYB96-2D is allelic to TaEPBM1-D, further supporting that TaMYB96-2D/5D function as transcriptional activators of wax biosynthesis genes in wheat. Thus, future work will have to concentrate on the functional characterization of TaMYB96-2D/5D involvement in the regulation of cuticular wax biosynthesis in wheat. Besides, posttranscriptional and posttranslational regulation mechanisms have been reported (Lee and Suh, 2013, 2015a, 2022; Wang et al., 2018b; Kim et al., 2019). In Arabidopsis, AtCER7 regulates stem wax by influencing AtCER3 transcript stability (Hooker et al., 2007; Lam et al., 2012, 2014). In wheat, the long noncoding miRNA gene Iw1 represses β-diketone synthesis by suppressing the transcription of the carboxylesterase-like protein gene WAX1-CARBOXYLESTERASE (Huang et al., 2017). The two E3 ubiquitin ligases MYB30-INTERACTING E3 LIGASE1 and ABA-related RING-type E3 negatively regulate wax accumulation in Arabidopsis (Lee et al., 2017; Liu et al., 2021). In rice, the E3 ubiquitin ligase DROUGHT HYPERSENSITIVE negatively regulated cuticular wax biosynthesis by interacting with the transcription factor ROC4 (Wang et al., 2018b). In wheat, the regulatory mechanisms of wax biosynthesis currently remain largely unknown, and the understanding of these mechanisms will be useful for optimizing wax contents and improving drought tolerance through crop breeding.

Materials and methods

Plant materials and growth conditions

The common wheat (T. aestivum L.) cultivar JM35 was used for gene cloning and the analysis of cuticular wax content, composition, and crystals morphology. For haplotype analysis, a total of 43 wheat accessions were grown in the research field of Northwest A&F University, Yangling, Shaanxi Province, China (34°17′ N, 108°04′ E, 506 m altitude). Twenty seeds were sown in a 1.5-m row at 30-cm apart. At the flowering date, seven different organs of JM35 plants, including FLBs, TSBs, TTBs, TFBs, FLSs, peduncles, and spikes, were sampled for SEM, GC-MS, and GC-FID analysis. Mature seeds and 4-week-old roots grown in glass tubes were evaluated for wax content. Moreover, CS N–T lines, Fielder, and transgenic lines were grown in a glasshouse with 16-h light/8-h dark, and the day/night temperature was 26°C/20°C. Six-week-old seedlings of JM35 were subjected to ABA and various abiotic stress treatments as described previously (Wang et al., 2015b).

Cuticular wax extraction

Wheat leaf blades, peduncles, and FLSs were harvested and imaged by a scanner and cut into 4-cm-long segments. The surface areas of these samples were calculated by the ImageJ software (http://rsb.info.nih.gov/ij/). All samples, including FLBs, TSBs, TTBs, TFBs, peduncles, roots, and spikes, were immersed in chloroform and shaken for 60 s at room temperature to remove cuticular waxes, except that the extraction time of seeds was 3 min instead of 60 s. For the spike, seed, and root wax analyses, the extracted wax samples were dried at 50°C for 7 days and weighed. Then, 20 μg of n-tetracosane (C24 alkane) was added to all extracts as an internal standard. Cuticular wax samples were filtered and evaporated to dryness under a stream of nitrogen. The extracts were transferred to GC autosampler vials and dried completely with N₂ gas. The residues were derivatized with 100 µL of bis-N,N-(trimethylsilyl) trifluoroacetamide (BSTFA, Sigma) in 100 µL pyridine (Fluka) for 60 min at 70°C to convert hydroxyl- and carboxyl-containing compounds to their corresponding trimethylsilyl derivatives. Finally, BSTFA and pyridine were evaporated under a nitrogen stream and the resulting derivatives were redissolved in 500 µL of chloroform for GC-MS and GC-FID analysis.

Chemical analysis of wax composition

The compositions of the cuticular waxes were analyzed by GC equipped with a Rxi-5ms column (length 30 m, i.d. 0.25 mm, df 0.25 µm; Restek, USA) attached to a mass spectrometer (GCMS-QP2010, Shimadzu, Japan). The column was operated with the helium carrier gas and splitless injection at 250°C. The oven temperature was programmed for 2 min at 50°C, increased by 5°C/min to 230°C, held for 5 min at 230°C, raised by 2°C/min to 310°C, and held for 20 min at 310°C. To identify each wax compound, the obtained mass spectra were compared with those of authentic standards and literature data. The quantitative compositions of wax extracts were studied using GC-FID (GC-2010 Plus, Shimadzu, Japan; column Rtx-1, length 60 m, i.d 0.25 mm, df 0.25 µm; Restek). The same GC program was used with hydrogen as the carrier gas. Quantification was based on FID peak areas relative to the internal standard tetracosane. The total amount of cuticular wax is expressed per unit of leaf blades, peduncles or FLSs surface area, and the total amount of spikes, seeds, and roots wax is expressed in micrograms of wax per grams of spikes, seeds, and roots (dry weight).

CRISPR/Cas9-mediated knockout and overexpression of TaCER1-6A in wheat

For genome editing of TaCER1-6A, a target site located at the eighth exon of TaCER1-6A was chosen for designing the sgRNA sequence. The target sequence was first ligated to a corresponding sgRNA expression cassette driven by the OsU6a promoter and then combined into the binary pYLCRISPR/Cas9 vector based on Golden Gate Cloning (Ma et al., 2015). Finally, the pYLCRISPR/Cas9 plasmid was introduced into the A. tumefaciens strain EHA105 and transformed into the immature embryogenic callus of wheat cv Fielder as described previously (Ishida et al., 2015). Mutations in the target region were confirmed by sequencing 15 individual clones from each T₀ transgenic plant. For TaCER1-6A overexpression, the coding region of TaCER1-6A was inserted into the modified pCAMBIA3301 vector using XbaI and BglII sites under the control of the constitutive maize ubiquitin promoter. The generated plasmids were transferred into Fielder callus by A. tumefaciens-mediated transformation using the methods of Ishida et al. (2015).

TaCER1-6s and TaCER1-1A expression analysis

Total RNA was extracted from different organs of wheat cv JM35 using an RNAprep Pure Plant Kit (TIANGEN, Beijing, China) and reverse-transcribed into the first-strand cDNA using a FastKing RT Kit (TIANGEN) for further analysis. Quantitative PCR was carried out on an ABI7500 Real-Time PCR System (Applied Biosystems) using SYBR Green Realtime PCR Master Mix (TOYOBO, Japan). Wheat Actin (AB181991) and Tubulin (U76558) genes were used as the internal references, and the relative expression levels of TaCER1-6s and TaCER1-1A were calculated using the 2^−ΔΔCT method (Livak and Schmittgen, 2001). RT-qPCR analysis was performed using three biological replicates and three technical replicates for each sample.

Sequence alignment and phylogenetic analysis

A BLAST analysis with the deduced CER1 protein sequence was used to search for the closest homologs from the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). Multiple sequences were aligned with the ClustalW 1.83 program using default parameters. The aligned sequences were imported into Bioedit for manual editing (Hall, 1999). The phylogenetic tree was constructed using the NJ method (Saitou and Nei, 1987) with 1,000 bootstrap replicates through MEGA7 software (Kumar et al., 2016).

Subcellular localization of TaCER1-6s and TaCER1-1A

The full-length coding sequences of TaCER1-6s and TaCER1-1A without a stop codon were in-frame fused upstream of the GFP vector pCAMBIA1300-221-GFP under the control of a 35S promoter. The resulting construct TaCER1-GFPs and the red fluorescent protein (RFP) ER marker PCX-DR-CTR3 were introduced into A. tumefaciens strain GV3101, and then coinjected into leaf epidermal cells of 3-week-old N. benthamiana plants according to protocols described previously (Sparkes et al., 2006). GFP and RFP fluorescent signals were detected and imaged using a confocal laser scanning microscope (Leica TCS SP8 SR) after infiltration for 2 days. GFP was excited using a 488-nm laser under the 505–550-nm emission range, and the intensity and the gain were 10% and 200, respectively. RFP was excited using a 561-nm laser under the 590–650-nm emission range, and the intensity and the gain were 10% and 100, respectively.

Chlorophyll leaching and water loss assays

For the chlorophyll leaching experiment, flag leaves from WT plants, knockout lines, and overexpression lines were placed in the dark for 3 h and cut into pieces of approximately 5 cm in length. Leaves were immediately immersed in 50 mL of 80% ethanol (v/v) in glass tubes. Then, 100 μL of aliquots was sucked out at 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, and 24 h, and the absorbance of each sample was measured at 647 and 664 nm. The concentration (M) of chlorophyll in the leachate was calculated using the equation: M = 19.53 × A₆₄₇ + 7.93 × A₆₆₄ as described previously (Lolle et al., 1998). To measure the water loss rate, flag leaves were collected and initially soaked in deionized water in the dark for 3 h. The excess water was removed by a napkin and the leaves were returned to the dark. Subsequently, leaves were weighed at 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12 h with a microbalance. The water loss rate was calculated as the percentage of the initial fresh weight (Chen et al., 2003).

Y1H assay

The Matchmaker Gold Yeast One-Hybrid Library Screening System (Clontech) was used to carry out the Y1H assay. Three tandem copies of the “CAACCA” motif1, “CAACTG” motif2, and “TAACCA” motif3 from the promoters of TaFAR4, TaCER1-6A, and TaCER1-1A genes were inserted into the pABAi vector as the bait vectors. The pABAi-motifs constructs were linearized by the BstBI restriction enzyme and transformed into the Y1HGold yeast strain. The coding sequences of TaMYB96-2D/5D were fused into the pGADT7 vector using EcoRI and BamHI sites as the prey constructs. Subsequently, the pGADT7-TaMYB96-2D/5D constructs and the empty pGADT7 vector were separately transformed into the bait strain. The transformed yeast cells were grown in SD/–Ura–Leu medium containing 0.4 mg L⁻¹ AbA at 30°C for 4 days. Primers used in this assay are listed in Supplemental Table S3.

LUC assay

The 1,141-, 1,514-, and 1,826-bp promoter sequences of TaCER1-6A, TaFAR4, and TaCER1-1A were inserted into the pGreenII 0800-LUC vector using HindIII and BamHI restriction enzyme sites. The generated constructs ProTaCER1-6A-LUC, ProTaFAR4-LUC, and ProTaCER1-1A-LUC were used as reporters. Meanwhile, the coding sequences of TaMYB96-2D/5D were cloned into the pGreenII 62-SK vector through EcoRI and HindIII sites to generate the effector constructs TaMYB96-2D-SK and TaMYB96-5D-SK. The effector, reporter, and empty vector constructs were transformed into A. tumefaciens strain GV3101. Each reporter strain, together with the empty vector or either the TaMYB96-2D-SK effector or the TaMYB96-5D-SK effector, was co-infiltrated into 4-week-old N. benthamiana leaves according to previous protocols (Sparkes et al., 2006). After 3 days of infiltration, the activities of firefly LUC and Renilla LUC (REN) were measured using the TransDetect Double-Luciferase Reporter Assay Kit (Transgene, Beijing, China) according to the manufacturer’s instructions. The relative LUC activity was determined by the LUC:REN ratio.

GUS activity assay

The TaMYB96-2D/5D coding regions were cloned into the pC1300s vector using restriction enzyme sites SacI and BamHI to generate the effector constructs. The 1,141-bp promoter sequence of TaCER1-6A was cloned into the HindIII and BamHI sites upstream of the GUS reporter gene in the pBI121 vector to generate a reporter construct. The reporter and effector vectors were transformed into A. tumefaciens strain GV3101 and co-infiltrated into 4-week-old N. benthamiana leaves. The examination of GUS activity was performed as previously described by Jefferson et al. (1987).

Statistical analysis

All experiments were carried out with three biological replicates and three independent times. Data are presented as means ± standard deviation (SD) of three independent experiments. Statistical analysis was performed with SPSS 19.0 software. Statistical differences of wax contents and GUS activity were determined using Student’s t test. Asterisks indicate significant differences (*P < 0.05, **P < 0.01, and ***P < 0.001). For multiple comparisons of expression levels at different time points and LUC/REN values, Duncan’s multiple range test (P < 0.05) was performed. Different letters above the error bars indicate significant differences.

Accession numbers

Genes referenced in this article are available in the GenBank database with the accession numbers: TaCER1-6A (OM249960), TaCER1-6B (OM249961), TaCER1-6D (OM249962), TaMYB96-2D (OM249963), and TaMYB96-5D (OM249964).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. Contents of individual wax components on nine different organs of the flowering wheat cultivar JM35.

Supplemental Figure S2. Wax crystal morphology on seven different organs of the flowering wheat cv JM35.

Supplemental Figure S3. Cloning and structure characterization of TaCER1-6s.

Supplemental Figure S4. Protein expression of TaCER1-6A.

Supplemental Figure S5. Amino acid sequence alignment of TaCER1-6s and phylogenetic analysis of plant CER1 homologous proteins.

Supplemental Figure S6. Relative expression of TaCER1-6A and plant morphology comparison in WT plants and TaCER1-6A transgenic lines.

Supplemental Figure S7. Phylogenetic analysis of MYB transcription factors from Arabidopsis and wheat.

Supplemental Figure S8. Comparison of agronomic traits from WT plants, transgenic knockout line KO#16, and overexpression line OE#10 in the field experiment.

Supplemental Table S1. The percentage (%) of cuticular wax components in nine different organs of flowering wheat cv JM35.

Supplemental Table S2. The 43 wheat accessions used for TaCER1-6A haplotype analysis.

Supplemental Table S3. Sequences of the PCR primers used in this study.

 

Y.W. and Z.W. conceived, designed, and supervised the experiments. Y.W., J.H., C.L., N.H., Y.Z., and Z.H. performed the experiments. Y.W. wrote the manuscript. J.H., C.L., N.H., Y.S., and Z.W. revised the manuscript. All authors have read and approved the final version of the manuscript.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) is: Yong Wang (wangyong2114@163.com).