Genome-wide identification and characterization of abiotic stress responsive GRAS family genes in oat (Avena sativa)

Identification and sequence analysis of GRAS genes in oat

To identify the GRAS genes present in oat, the first version of the whole genome sequence of oat (DNA and Gff3) was downloaded from the results of the joint sequencing study done by PepsiCo and Corteva Agriscience (https://wheat.pw.usda.gov/GG3/graingenes_downloads/oat-ot3098-pepsico). The protein sequences of A. thaliana were downloaded from the TAIR database (https://www.Arabidopsis.org/). The Hidden Markov Model (HMM) file of the GRAS domain (PF03514) was downloaded from the Pfam protein family database (http://pfam.sanger.ac.uk/). The GRAS protein sequences of oat were aligned using the HMM model in HMMER 3.1 software (National Institutes of Health, Bethesda, MD, USA). We extracted the GRAS family member sequences of oat using the TBtools software (Chen et al., 2020a). The resulting protein sequence of GRAS was verified using the SMART (http://smart.embl.de/) and PFAM (http://pfam.xfam.org/) programs. The protein sequence of AsGRAS was further verified using the NCBI protein database (https://blast.ncbi.nlm.nih.gov/Blast.cgi), BLASTp.

Phylogenetic analysis of GRAS genes in oat

To further investigate the phylogeny of the oat GRAS genes, we created a phylogenetic tree using the neighbour-joining (NJ) method. A multiple sequence alignment was performed and a phylogenetic tree of the Arabidopsis and oat GRAS gene families was constructed using the MEGA-X software (https://www.megasoftware.net/; Kumar et al., 2018). The NJ phylogenetic tree was built with 1,000 bootstrap replications. All the identified AsGRAS genes were then divided into different clusters based on AsGRAS classification. The GRAS protein sequences (A. thaliana, O. sativa, Z. mays) were downloaded from the Ensembl database (http://ensemblgenomes.org/). ITOL (https://itol.embl.de/) and Adobe Photoshop 2019 (https://www.adobe.com/products/photoshop.html?promoid=RBS7NL7F&mv=other) were used to depict the modified phylogenetic tree.

Analysis of conserved motifs of oat

The exons, coding sequences, and untranslated regions on the oat chromosome were extracted from the gff3 file and the TBtools software was used to map the structure of the gene (Chen et al., 2020a). A multiple sequence alignment of AsGRAS proteins was performed using the Jalview software package (Version 2.10.5), with the parameters of the alignment process set to default values, and the results color-coded by the BioEdit software (Hall, 2011). The sequences of the conserved proteins in AsGRAS were identified using the MEME Suit online tool (Bailey et al., 2009; Version 5.3.3, https://meme-suite.org/meme/tools/meme) and visualized using the TBtools software. The physicochemical properties of GRAS proteins in oat were predicted using the ExPASy proteomics server, and their subcellular localizations were identified using the Cell-PLoc online software (Chou & Shen, 2010).

Chromosomal location and homology analysis of GRAS genes in oat

The CDS and gene sequences corresponding to all GRAS genes in oat were retrieved from the oat genome files using the TBtools tool, and the introns of the AsGRAS genes were analyzed using the GSDS 2.0 online website (gene structure display server, http://gsds.gao-lab.org/) and the exon structure was used for predictive analyses. The corresponding positions of the AsGRAS genes on the chromosome were identified using the GFF3 data from the oat genome (Chen et al., 2020b), and TBtools (Chen et al., 2020a) was used to complete the visualization of the location of the genes on the chromosome. In order to explore the GRAS gene homology between oat and other species, we downloaded genome-wide data and the gene annotation files for Arabidopsis (Arabidopsis thaliana), rice (O. sativa), maize (Zea mays) wheat (Triticum aestivum), barley (Hordeum vulgare), foxtail (Setaria viridis), teff (Eragrostis curvula), and millet (Setaria italica) from the online website EnsemblPlants (http://ensemblgenomes.org/). We then performed an analysis of AsGRAS gene replication events using the “One-Step MCscanX” function in the TBtools software. The “Dual Synteny Plot” function in the TBtools software was used to build the multispecies collinearity analysis table, and the nonsynonymous substitution value (Ka) and synonymous substitution value (Ks) of the GRAS repeat gene were then calculated.

Analysis of cis-acting elements in the AsGRAS gene family

The 2,000 bp upstream base of the start codon of the oat GRAS gene family was used as the sequence in the PLACE (http://www.dna.affrc.go.jp/PLACE/signalscan) online software. The TBtools (Chen et al., 2020a) software was then used to analyze the sequence with the extracted elements used to perform predictive analysis on the cis-acting elements of this gene family. The cis-elements of the AsGRAS gene family were visualized by the “Simple BioSequence Viewer” function of the TBtools software.

Plant material and stress treatments

The oat variety selected in this experiment was “Qinghai 444,” and the seeds were provided by the Germplasm Restoration Laboratory of Qinghai-Tibet Plateau Research Institute, Southwest Minzu University. This experiment began on September 25, 2021 with full-grained seeds soaked in 75% ethanol for 1 min for disinfection, and then germinated on filter paper for about 48 h. The oat seedlings with consistent growth were then transferred to 1/2 Hoagland nutrient solution (PH = 5.8) in 16 h light/8 h darkness with a light intensity of about 8,000 lx. The seedlings were then grown and cultured at 28 °C until the two-leaf stage, and the nutrient solution was replaced every 2 days. The oat plants with relatively consistent growth for 3 weeks were selected to start drought stress (20% PEG-6000; Macklin, Dunmurry, UK) and salt stress (0.9% NaCl; Macklin, Dunmurry, UK) treatments (Wang et al., 2020c). The treatment time points were set at 1, 3, 6, 24, and 48 h, with one control (0 h). The second leaf from the top of each plant and a root sample from each plant were then taken and immersed in liquid nitrogen for quick freezing and then stored in a −80 °C freezer. Three independent biological replicates were performed, each with 3–6 individual plants.

Expression analysis of the AsGRAS genes by real-time PCR

Total RNA was extracted from the root and leaf tissues of oat plants under different treatment schedules using the MiniBEST Plant RNA Extraction Kit (Takara, Kusatsu, Japan), and reverse transcription was carried out using a cDNA synthesis kit (PrimeScriptTM RT reagent Kit with gDNA Eraser; Takara, Kusatsu, Japan). The cDNA was removed after the completion of the genome, and cDNA concentration was detected using a NanoDrop 2000 UV spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and then diluted uniformly to a concentration of 50 ng·μL-1 for the qRT-PCR reaction matrix. Specific primers for analysis were designed with Primer software (Version 5.0; Table S1) and the qRT-PCR analysis was performed using the StepOne Fast Real-Time PCR System (applied biosystems, Waltham, MA, USA) with three technical replicates. The amplification reaction system contained 2 µL of diluted cDNA solution, 1 µL of forward and reverse primers, 12.5 µL of SYBR Premix Ex Taq II (Japanese, Takara, Kusatsu, Japan), and 8.5 µL of sterile water for a total volume of 20 µL. The reaction program was: denaturation at 95 °C for 30 s, denaturation at 95 °C for 5 s, annealing at 60 °C for 30 s, for 40 cycles, followed by 95 °C for 15 s, 60 °C for 30 s, and 95 °C for 15 s. Using GAPDH (Tajti, Pál & Janda, 2021) as the internal reference gene, each sample was replicated three times, and the relative expression was calculated using the 2^−ΔΔCt method (Livak & Schmittgen, 2001).

Statistical analysis

All data in this study were analyzed using the Origin 2020 pro statistical analysis program (OriginLab Corporation, Northampton, MA, USA) using analysis of variance, and means were compared using least significant difference tests (LSD) at the 0.05 level of significance.

Identification of GRAS family numbers in oat

In this study, 30 GRAS proteins were identified in oat and were named AsGRAS01 to AsGRAS30 according to their physical location on the chromosome (Table S2). In this study, we analyzed the basic characteristics of these 30 AsGRAS proteins, including their coding sequence length (CDS), protein molecular weight (MW), isoelectric point (pI), and chromosomal location (Location; Table 1). Among these 30 AsGRAS proteins, AsGRAS24 was the shortest with only 291 amino acids, and AsGRAS30 was the largest with 635 amino acids. The molecular weight of the proteins ranged between 22.27 kDa (AsGRAS09) and 66.62 kDa (AsGRAS25). The pI value of the proteins ranged between 5.38 (AsGRAS17) and 9.50 (AsGRAS09), with a mean pI value of 6.01, indicating that the oat GRAS protein was acidic. More than half of the AsGRAS proteins were distributed on three chromosomes: chromosome 1A (n = 12, 40%), chromosome 1D (n = 7, 23.3%), and chromosome 2C (n = 5, 16%), and chromosome 1C had only two genes (6%).

Phylogenetic analysis and classification of AsGRAS genes

To explore the phylogenetic relationship of the GRAS proteins identified in oat, we constructed a phylogenetic tree based on the amino acid sequences of the 30 AsGRAS proteins (Table S3) and 32 AtGRAS proteins using the neighbor-joining (NJ) method (Fig. 1). According to their homology to Arabidopsis GRAS proteins, the 30 GRAS proteins of oat were divided into four subfamilies: SCR, PAT1, HAM, and DELLA. The PAT1 subfamily was the largest, with 16 AsGRAS genes. The SCR subfamily contained only one AsGRAS gene, and the HAM and DELLA subfamilies had 11 and two AsGRAS genes, respectively.

This study further investigated the diversity of the evolutionary process of the GRAS family of genes in oat by analyzing 32 Arabidopsis GRAS genes, 60 O. sativa GRAS genes, 105 Z. mays GRAS genes, and 30 A. sativa GRAS genes for genetic relationships (Table S4). According to branching angle, developmental tree topology, and Arabidopsis classification, 227 GRAS genes were divided into nine subfamilies (Fig. 2). The LISCL and HAM subfamilies contained the most GRAS genes, while the LAS and SCL4/7 subfamilies had the fewest. The LAS and SCL4/7 subfamilies were not recognized in oat, indicating that these subfamilies occur in oat as a species with specific deletions.

Gene structure and motif composition of the AsGRAS gene family

The genomic DNA sequences of the AsGRAS gene were used to create exon and intron structure diagrams of the AsGRAS gene to better understand its structural composition (Fig. 3A). All 30 AsGRAS genes contained a GRAS domain, and most (24, 80%) contained no introns. AsGRAS04, AsGRAS05, AsGRAS09, AsGRAS16, AsGRAS21, and AsGRAS24 each contained two introns. In general, members of the same subfamily had similar gene structures. The protein conserved sequences and sequence logos of the GRAS gene family members in oat were analyzed using MEME online analysis software. A total of 10 conserved motifs (named motif 1–motif 10) were identified, and more motifs were located at the C-terminus than at the N-terminus (Fig. 3B). Most AsGRAS proteins contained motif 4 and motif 10. AsGRAS02 did not have motif 4, while AsGRAS09 only included motifs 4 and 10. Some motifs were only present in specific subfamilies, for example, motif 9 was only present in the HAM subfamily. A comparison of the members of the AsGRAS gene family found that the type, number, and sequence of protein motifs of members of the same subfamily were remarkably conserved. Most closely related members had similar motifs: for example, the PAT1 group included motif 6, motif 4, motif 1, motif 8, motif 2, motif 5, motif 3, and motif 7, while the SCR group contained motif 10, motif 4, motif 1, motif 8, motif 2, motif 3, and motif 7.

Chromosomal distribution and collinearity analysis of AsGRAS genes

The distribution of GRAS genes on the oat chromosomes were also analyzed using the oat genome (Fig. 4). The results showed that the 30 AsGRAS genes were unevenly distributed on the five chromosomes in oat with chromosome 1A having the most genes (12, 40%), followed by chromosome 1D (7, accounting for about 23.3%), chromosome 2C (5, about 16.7%), chromosome 2D (4, about 13.3%), and chromosome 1C (2, about 6.7%). In this study, 17 AsGRAS genes were clustered into 12 tandem duplication event regions in oat chromosomes 1A, 1C, 2C, and 1D (Fig. 5), indicating that these regions were hot spots of AsGRAS gene distribution. Chromosome 1A had eight clusters (AsGRAS06/AsGRAS07, AsGRAS06/AsGRAS03, AsGRAS10/AsGRAS04, AsGRAS10/AsGRAS12, AsGRAS05/AsGRAS04, AsGRAS05/AsGRAS01, AsGRAS12/AsGRAS08, AsGRAS11/AsGRAS01), Chromosome 1C had one cluster (AsGRAS13/AsGRAS14), Chromosome 2C also had one cluster (AsGRAS23/AsGRAS26), and Chromosome 1D had two clusters (AsGRAS17/AsGRAS15, AsGRAS15/AsGRAS20). This suggests an evolutionary relationship between these AsGRAS members as these genes all had similar structures and functions, and originated from chromosomal duplication. The collinear analysis of AsGRAS proteins showed that the genes with collinear relationships were located in the same subfamily.

To further explore the evolutionary clues of the oat GRAS gene family, this study constructed eight representative comparative syntenic graphs of GRAS family members between oat and eight representative species (Fig. 6). The eight species included one dicot (A. thaliana) and seven monocots (Z. mays, O. sativa, T. aestivum, H. vulgare, Setaria italica, Setaria viridis, and Eragrostis curvula). A total of 30 AsGRAS gene members were associated with Arabidopsis (30), O. sativa (30), Z. mays (30), T. aestivum (30), H. vulgare (30), S. italica (29), S. viridis (30), and the genes of E. curvula (30) had syntenic relationships (Table S5). A total of 70, 57, 33, 12, 12, 11, and 2 orthologous AsGRAS gene pairs were identified between oat and Z. mays, O. sativa, T. aestivum, H. vulgare, S. italica, S. viridis, E. curvula, and A. thaliana, respectively. In general, AsGRAS genes consisted of more monocotyledonous symbiotic gene pairs. The analysis of collinearity between oat and other species have important implications for elucidating the evolution of GRAS genes.

Cis-acting element analysis of AsGRAS genes

Cis-acting elements are specific binding sites for proteins involved in transcription initiation and regulation and play a role in plant gene transcription regulation (Hernandez-Garcia & Finer, 2014). To determine the cis-acting elements in the AsGRAS promoter region, the cis-element types and potential functions of the oat GRAS gene family were analyzed using PLACE online software (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/). The results showed (Fig. 7 and Table S6) that there were a large number of cis-regulatory elements in the promoter region of the oat GRAS gene family, including light response element, abscisic acid response element, low-temperature stress response element, drought stress response cis-acting elements, and tissue-specific expression elements. The cis-acting elements contained in the promoter regions of different members of the AsGRAS gene family had similarities and differences in both types and number of elements. For example, all AsGRAS genes contained light-responsive elements, and most of the AsGRAS genes contained abscisic acid-responsive elements (26) and low-temperature stress response elements (26). Some AsGRAS genes contained drought-responsive elements (20). The AsGRAS27 gene contained more cis-acting elements (49), while the AsGRAS02 (11) and AsGRAS09 (9) genes of the DELLA subfamily contained fewer cis-acting elements. The analysis results showed that AsGRAS genes play an essential regulatory role in different stress-resistant habitats.

Expression of AsGRAS genes under salt stress and drought stress

With the aim of further investigating the expression level of AsGRAS genes under different stresses and their potential functions in the developmental process, 12 AsGRAS genes were selected from the identified GRAS genes for qRT-PCR experiment verification in oat. Each AsGRAS gene subfamily is shown in Fig. 8 and Table S2, and the specific primers used for the qRT-PCR analysis are shown in Table S1. Significant changes in AsGRAS gene expression patterns were observed in two plant tissues (leaf and root), and some of these genes exhibited significant expression in the plant tissues. Results from the qRT-PCR experiments indicated that the selected genes often exhibited different expression profiles in the different AsGRAS gene subfamilies. However, selected genes in some subfamilies showed similar expression patterns. For example, under salt stress, the expression levels of AsGRAS04 and AsGRAS12 of the HAM subfamily were upregulated in the leaf tissue from 0-HAS (hour after stress) to 3-HAS and then down-regulated in a stepwise manner starting at 6-HAS. There were also different patterns of gene expression between members of the same subfamily. For instance, under osmotic stress, the expression of AsGRAS24 of the PAT1 subfamily was upregulated in leaf tissue from 0-HAS to 6-HAS and then gradually decreased from 24-HAS onwards. In contrast, AsGRAS30 was upregulated from 0-HAS to 3-HAS in leaf tissue and then gradually down-regulated from 6-HAS. These results indicate that the selection of representative GRAS gene expression profiles in oat can describe different AsGRAS gene expression patterns, underscoring the regulation of AsGRAS genes in the process of oat growth.