Identification and expression analysis of EDR1-like genes in tobacco (Nicotiana tabacum) in response to Golovinomyces orontii

Identification of the EDR1-like gene family

The genome, gene and protein sequences of Arabidopsis, tomato and rice were downloaded from PlantGDB (http://www.plantgdb.org/), and those of tobacco from the Sol Genomics Network (http://solgenomics.net/organism/). To identify EDR1-like genes in the above-mentioned species, two EDR1 proteins from Arabidopsis (GenBank accession: ABR45974.1) and wheat (GenBank accession: AAU89661.2) were used as the query subjects in a reciprocal Basic Local Alignment Search Tool Protein (BLASTP) analysis with e-values < 1E-50. Then, the online software SMART (http://smart.embl-heidelberg.de/) was used to identify the predicted EDR1-like protein domains. Conserved EDR1 domains and kinase domains within the acquired EDR1 sequences were confirmed by searching NCBI’s conserved domain database (http://www.ncbi.nlm.nih.gov/Structure/cdd/docs/cdd_search.html).

Phylogenetic analysis and conserved domain detection

Phylogenetic trees were constructed in MEGA Version 7.0 (Kumar, Stecher & Tamura, 2016), using the Neighbor-Joining (NJ) method (Saitou, 1987) with parameters of pairwise gap deletion and 1,000 bootstraps. Multiple sequence alignment of the amino acid sequences of EDR1-like proteins was performed using the ClustalX program (version 1.83) (Thompson et al., 1997) and GeneDoc. Maximization for Motif Elicitation program (MEME, http://alternate.meme-suite.org) (Bailey et al., 2006) was used to predict the conserved motif of EDR1-like proteins in N. tabacum.

Gene structure and bioinformatic analysis of tobacco EDR1-like genes

The genomic structure of the EDR1-like gene family was analyzed using GSDS 2.0 (http://gsds.cbi.pku.edu.cn/) (Hu et al., 2015). Protparam (http://web.expasy.org/protparam/) was used to analyze the basic physical and chemical properties of tobacco EDR1-like genes. Subcellular localization of the EDR1-like genes was determined using online software Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/) (Chou & Shen, 2008, 2010a, 2010b). The nuclear localization signals (NLS) sequences were predicted by cNLS Mapper (http://nls-mapper.iab.keio.ac.jp/cgi-bin/NLS_Mapper_form.cgi#opennewwindow) (Kosugi et al., 2009a, 2009b). Transmembrane helices were predicted using OCTOPUS (http://octopus.cbr.su.se/index.php#opennewwindow) (Viklund & Elofsson, 2008). The three-dimensional (3D) structures were predicted using the online I-TASSER program (http://zhanglab.ccmb.med.umich.edu/I-TASSER/) (Zhang, 2008). The confidence score (C-score) was used to estimate the model’s global accuracy in I-TASSER (Yang & Zhang, 2015), and high C-score indicates a high-quality structure prediction (Roy, Kucukural & Zhang, 2010). The model with highest C-score in all five prediction models was further used to construct 3D model of the target genes. The constructed 3D model was examined and visualized using Chimera 1.2 (https://www.cgl.ucsf.edu/chimera/).

Plant materials and stress treatments

Common tobacco (Nicotiana tabacum cv HHDJY) seeds were surface-sterilized by 10% NaClO, and sown on sterilized plates with MS solid culture medium (PhytoTechnology Laboratories®, Kansas, MO, USA) in a greenhouse maintained at 25 °C and with aday/night cycle of 16/8 h.

For powdery mildew infection, a strain of G. orontii was maintained on 3-month-old tobacco in a greenhouse. Freshly sporulating leaves of heavily infected tobacco were washed in sterile double distilled water (ddH₂O), and the collected conidiospores were used immediately. Plants were inoculated by spraying with the inoculum suspension (about 4 × 10⁴ to 5 × 10⁴ spores ml⁻¹). The 3-month-old tobacco leaves were inoculated with G. orontii to give approximately 20–25 spores per cm² and sampled after 0, 1, 2, 12 and 24 hours post inoculation (hpi). Samples collected at 0 h were used as controls. For tissue/organ expression profiles, the root, stem and leaf were sampled on two-month-old seedlings. Flowers were taken from flowering plants and capsules were obtained during the late seed-producing period. All selected tissues and organs were stored at −80 °C.

Total RNA isolation and quantitative real-time (qRT)-PCR expression analysis

An EasyPure Plant RNA Kit (TransGen Biotech, Beijing, China) was used to extract total RNA according to the manufacturer’s protocol. The first-strand cDNA templates were synthesized from two μg of total RNA, according to the manufacturer’s protocol (Promega, Madison, WI, USA). The reverse transcription reaction was incubated at 42 °C for 65 min in a total volume of 26 μl. The reverse transcription products were diluted five times with sterile ddH₂O.

Specific primers were designed by Premier 5.0 according to tobacco EDR1-like gene sequences (Table S1). Due to the high sequence similarity between NtEDR1-1A and NtEDR1-1B, it was difficult to design specific primers for qRT-PCR to measure the expression of NtEDR1-1A and NtEDR1-1B separately. Therefore, one pair of primers targeting both NtEDR1-1A and NtEDR1-1B was synthesized to simultaneously check the expression levels of NtEDR1-1A and NtEDR1-1B. Each PCR reaction was mixed with 10 μl of SYBR Green (TaKaRa, Dalian, China), 6.4 μl of ddH₂O, two μl of synthesized cDNA product and 0.8 μl of each primer (50 μM). Relative gene expression level was analyzed according to the 2^−ΔΔCt method (Livak & Schmittgen, 2001). Tobacco EF-1a gene (GenBank: D63396.1) was used as the internal reference. The real-time PCR analyses were performed using a qTOWER2.2 real-time PCR system (Analytik Jena AG, Jena, Germany), which was programmed as follows: initial 95 °C denaturation step for 3 min, followed by 40 cycles of denaturation at 95 °C for 10 s and annealing/extension at 60 °C for 1 min (Zhang et al., 2013). Every experiment was conducted with three biological replicates. All data were expressed as mean ± SEM and analyzed by Graphpad Prism 5.0 software.

Identification of the EDR1-like gene family in tobacco and three other plant species

As there was no previous systematic identification and analysis of EDR1-like genes in plants, we identified the EDR1-like genes in tobacco and compared them with other model species. In total, we obtained 19, 12, 8 and 14 EDR1-like genes in tobacco, Arabidopsis, tomato and rice, respectively (Table S2). In addition, the basic sequence information of tobacco EDR1-like genes was analyzed, including the location of the EDR1 regulatory region and the kinase region, the number of amino acids, isoelectric points, molecular weight and predicted location (Table 1).

The NtEDR1 regulatory region was located at the N-terminal, whereas the kinase region was located at the C-terminal of the predicted protein sequence. The C-terminal kinase region shared a sequence identity of 61.48%, and the N-terminal regulatory region had a relatively low identity of 40.52% for the predicted protein sequences of tobacco EDR1-like genes (Fig. S1). The number of amino acids ranged from 648 (NtEDR1-17) to 1,240 (NtEDR1-13). The predicted molecular masses were in the range of 73.67–138.74 kDa and the isoelectric points were 5.02–6.36. Protein subcellular localization prediction is a key step in understanding the protein function and its protein network interaction pattern (Chou & Shen, 2007). In Arabidopsis, at least a fraction of the EDR1 protein has been shown to localize in the nucleus (Christiansen et al., 2011). Our results showed 15 out of 19 tobacco EDR1-like proteins were predicted to be located in the nucleus (Table 1), but NtEDR1-11 and NtEDR1-14 were predicted to be located in the cell membrane. The other two tobacco EDR1-like members, NtEDR1-6 and NtEDR1-17, were shown to be located in the chloroplast or nucleus, and cell membrane or nucleus, respectively. The NLS predictions for most genes were consistent with those of subcellular localization, which showed that 13 NtEDR1-like proteins had NLS sequences (Table 1; Table S3). However, there were differences for another six NtEDR1-like proteins (Table 1; Table S3). This might be due to the different algorithms used for the two tools.

Phylogenetic and analysis of EDR1-like genes

Reports have shown that EDR1 homologs play a negative role in the defense response against powdery mildew in Arabidopsis (Frye & Innes, 1998) and wheat (Zhang et al., 2017). To reveal the evolutionary relationships among different plant species, a phylogenetic tree including EDR1-like gene members in tobacco, Arabidopsis, wheat and rice was constructed using software MEGA 7.0 with NJ method (Fig. 1). The results divided the EDR1-like genes into six clades and the monocot and eudicot EDR1-like genes were separated into two sub-families within each clade (Fig. 1). NtEDR1-1A andNtEDR1-1B, along with EDR1 homologs involved in negative regulation of resistance against powdery mildew in Arabidopsis (AtEDR1-1) and wheat (TaEDR1), were grouped into clade I. Thus, NtEDR1-1A and NtEDR1-1B should be the first choice for further study of any negative roles in tobacco defense against powdery mildew. The exon–intron organizations of the EDR1-like genes were analyzed using online software GSDS (Fig. 1). The exon number of tobacco EDR1-like genes varied from nine (NtEDR1-18) to 22 (NtEDR1-13). Interestingly, NtEDR1-1A, AtEDR1-1, SlEDR1-1 and OsEDR1-1 had 13 exons and were all grouped into clade I in the phylogenetic tree. The similarities of sequences in clade I were further analyzed by multiple sequence alignment. The results showed that both the predicted C-terminal kinase region and N-terminal regulatory region of sequences in clade I had high identities (82.29% and 66.89%, respectively) (Fig. 2).

Conserved motif analysis and structure prediction of tobaccoEDR1-like protein

To further analyze the conserved features of the NtEDR1-like family proteins, the motif of the tobacco EDR1-like proteins were predicted by MEME web server. A total of 10 motifs were identified in the NtEDR1-like proteins (Fig. 3; Table S4). According to the prediction of SMART, motifs 1–4 and 10 were located within the kinase domain, while motif 5 was in the EDR1 domain. The results showed motifs 1, 4, 6 and 9 were necessary for all the tobacco EDR1-like members, but motifs 2, 3, 5, 7 and 8 were alternative components (Fig. 3A). The motif logo is shown in Fig. 3B, and the lengths and predicted motif models of each motif are shown in Table S4.

Homologous proteins with similar function often have a similar structure. AtEDR1-1 has been experimentally shown to be required for G. cichoracearum susceptibility (Frye & Innes, 1998), so we used AtEDR1-1 as a reference model to analyze NtEDR1-like proteins. The 3D structures of tobacco EDR1-like proteins and AtEDR1-1 were constructed by using I-TASSER web site. The results showed that the predicted 3D model of NtEDR1-1A was similar to that of AtEDR1-1, while the 3D models of the rest tobacco NtEDR1-like proteins were somewhat different (Fig. 4; Fig. S2).

Expression analysis of EDR1-like genes in tobacco

The qRT-PCR was employed to determine the expression of tobacco EDR1-like genes after plants were infected with G. orontii (Fig. 5). NtEDR1-1 was used to represent both NtEDR1-1A and NtEDR1-1B where necessary to describe the transcript abundance of NtEDR1-1A and NtEDR1-1B. The result showed that NtEDR1-1, NtEDR1-12 and NtEDR1-16 had similar expression patterns (Fig. 5B). These genes were not immediately up-regulated but began to increase after 2 hpi. NtEDR1-17 was up-regulated at 1 hpi, and maintained a high level of expression at 2 hpi. NtEDR1-3 and NtEDR1-10 weredown-regulated in the beginning of infection, but up-regulated at 12 hpi. NtEDR1-8 and NtEDR1-9 were significantly down-regulated by G. orontii infection. NtEDR1-2,NtEDR1-5 and NtEDR1-15 showed no obvious changes at all time points tested.NtEDR1-4 and NtEDR1-18 were up-regulated at 12 hpi. NtEDR1-6 and NtEDR1-11 were up-regulated at 1 hpi; subsequently, expression of NtEDR1-6 decreased, and NtEDR1-11 maintained high expression after 2 hpi. NtEDR1-7, NtEDR1-13 and NtEDR1-14 were immediately up-regulated with G. orontii infection at 1 hpi, but their expression declined to a normal level at 6 hpi. These results indicated that different EDR1-like genes showed different expression patterns after G. orontii challenge. The expression patterns of NtEDR1-1, NtEDR1-12 and NtEDR1-16 were similar after tobacco was challenged by G. orontii.

Analysis of transcriptional level of NtEDR1-like genes in different tissues/organ samples provides clues on their biological functions. The tissue/organ expression patterns of tobacco NtEDR1-like genes were also investigated using qRT-PCR (Fig. S3). Some NtEDR1-like genes showed organ-specific expression patterns in the organ sample investigated. For example, NtEDR1-13 and NtEDR1-16 were expressed highly only in roots and leaves, respectively (Fig. S3A). The expression of NtEDR1-1, NtEDR1-3, NtEDR1-7, NtEDR1-8, NtEDR1-10 and NtEDR1-15 were absent in flowers; and NtEDR1-4, NtEDR1-9, NtEDR1-11, NtEDR1-14, NtEDR1-17 and NtEDR1-18 did not show any expression in capsules. NtEDR1-2, NtEDR1-5, NtEDR1-6 and NtEDR1-12 had some transcripts in all organ samples tested (Figs. S3B–S3D).