Genome-wide analysis of evolution and expression profiles of NAC transcription factor gene family in Juglans regia L.

Abstract

Key message

NAC transcription factors may play important roles in the biological processes in Persian walnut. A total of 102 JrNAC s were identified in Persian walnut. The conserved domains, transcriptome profile, expression analysis, and interaction network suggest that JrNAC1-4 plays potential roles in Persian walnut flowering development.

Context

NACs are plant-specific transcription factors that participate in various plant developmental processes such as flowering, plant growth regulation, and development. We identified and analyzed the evolution and expression profiles of NAC genes in Juglans regia.

Aims

The main objectives were to identify the NAC transcription factors and verify the expression levels in different tissues and female flowers in developmental stages in Persian walnut.

Methods

We identified NAC transcription factors in J. regia based on the genome-wide analysis. We analyzed the phylogenetic relationships, conserved domain, chromosome location, gene structure, and gene collinearity of JrNACs. We also verified the JrNAC expression levels based on transcriptome analysis and qRT-PCR.

Results

We identified 102 NAC genes in J. regia and divided them into ten subfamilies. A total of 30 pairs of JrNAC genes were expanded by whole-genome duplications (WGDs) and one pair of genes (JrNAC2-10 and JrNAC9-8) as a tandem duplication in Persian walnut. Collinearity analysis results indicate that a large number of syntenic relationship events existed between J. regia and Populus trichocarpa. We found that JrNAC1-4 and JrNAC2-6 were expressed significantly higher in female flowers based on both transcriptome and qPCR analysis. We further identified that JrNAC2-9 and JrNAC9-6 were highly expressed at the end period of flowering stages.

Conclusion

A total of 102 JrNACs were identified in the Persian walnut genome. These genes were conserved in plants for collinearity analysis, which was performed within the genome and other genomes (P. trichocarpa, Olea europaea, and Quercus robur). A total of 24 NAC transcription factors were highly expressed in female and male flowers, and these transcription factors play a role in J. regia flowering.

1 Introduction

The NAC (NAM, ATAF1/2, and CUC) transcription factor (TF) family contains about ~ 150 amino acid residues and is the most important and largest TF family in plants that contains a conserved NAM domain at their N-terminus (Souer et al. 1996; Tran et al. 2010; Zhu et al. 2012). NAM was first identified in petunia and determines the position of the primordium and shoot apical meristem (Souer et al. 1996; Duval et al. 2002; Ooka et al. 2003). The NAC genes at their C-terminal end also contain a-helical transmembrane motifs (TMs) (Puranik et al. 2012). The C-terminal domain contains a non-conserved transcriptional activation region (TAR) in plants and acts as a transcription repressor or activator, sometimes displaying protein-binding activity. In addition, an α-helical transmembrane motif (named NTL) is present in some NACs that are required for plasma membrane or endoplasmic reticulum membrane anchoring the C-terminal region (Ooka et al. 2003; Tran et al. 2004). To date, NAC transcription factors in many plants have been detected, such as in the model plant Arabidopsis, where a total of 105 NAC genes have been identified (Ooka et al. 2003); in important annual crops Brassica pekinensis, Glycine max, Oryza sativa, Cajanus cajan, Setaria italica, Medicago truncatula, and Triticum turgidum (Nuruzzaman et al. 2010; Le et al. 2011; Puranik et al. 2013; Liu et al. 2014; Ling et al. 2017; Saidi et al. 2017); and perennial woody plants including Populus trichocarpa, Vitis vinifera, Prunus mume, Morus notabilis, and Malus domestica (Hu et al. 2010; Su et al. 2013; Satheesh et al. 2014; Baranwal and Khurana 2016; Zhuo et al. 2018).

NAC genes are involved in many biological processes, including stress responses (Wang et al. 2013; Satheesh et al. 2014), apical meristem development (Souer et al. 1996; Wang et al. 2009), hormone signaling (Ooka et al. 2003), leaf senescence (Guo and Gan 2006), fruit ripening (Duval et al. 2002; Shan et al. 2012), and flower formation (Souer et al. 1996; Sablowski and Meyerowitz 1998; Liu et al. 2009). Expression analysis showed that the NAC gene family is involved in both reproductive and vegetative tissues (Hu et al. 2003; Hennig et al. 2004). In woody plants, for example, a microarray transcriptomic analysis showed six of these genes to be expressed during the development and ripening of the Fragaria x ananassa fruit (Moyano et al. 2018), P. trichocarpa NAC genes have putative functional roles in wood-forming and secondary cell wall biosynthesis (Hu et al. 2010), and the grapevine NAC genes play a potential role in response to stress (Wang et al. 2013). MdNAC1 overexpressing apple plants maintained a higher photosynthetic rate under drought conditions and accumulated lower levels of reactive oxygen species under drought conditions (Su et al. 2013). Sixteen PmNACs (Prunus mume) exhibited downregulation during flower bud opening in apricot (Zhuo et al. 2018). For the NAC in flowering development, SlNAM2 participates in the establishment of tomato flower whorl and sepal boundaries (Han et al. 2012). There are some previous studies on woody plants (Hu et al. 2010; Su et al. 2013; Wang et al. 2013; Moyano et al. 2018; Zhuo et al. 2018), including the plant development and flowering process; however, there are no reports of NAC genes in Persian walnut (Juglans regia) that focus on the flowering and development.

Persian walnut is a diploid (2n = 32), large, wind-pollinated, monoecious, dichogamous, enduring, perennial tree and is the most monetarily vital nut tree on earth belonging to the family Juglandaceae (Han et al. 2016; Martínez-García et al. 2016; Feng et al. 2018). It has been an important tree species since ancient times, valued for both wood and nuts (Feng et al. 2018; Zhao et al. 2018; Yan et al. 2019a). As we know, the flowering development is important for the Persian walnut nuts, and the NAC gene is involved in flower formation (Souer et al. 1996; Sablowski and Meyerowitz 1998; Liu et al. 2009). In this study, to better understand the potential role and characteristics of NAC transcription factors in Persian walnut flowering, we performed phylogenetic analysis and analyzed the tandem and segmental duplications and intron–exon structures of JrNAC genes. To better comprehend whether the NAC gene might play an important role in flowering, we analyzed the transcriptome expression level in reproductive and vegetative tissues and carried out qRT-PCR analysis for three genes (JrNAC1-4, JrNAC2-6, and JrNAC13-5) in male and female flowers and leaves. This study provides the first genome-wide analysis of the Persian walnut NAC transcription factor family, and these findings will be useful for understanding the putative functions of Persian walnut NAC genes.

2 Materials and methods

2.1 Identification of NAC transcription factors in J. regia

The Persian walnut whole protein sequence was downloaded from National Center for Biotechnology Information (NCBI) (Martínez-García et al. 2016). Members of the Arabidopsis NAC gene family were downloaded from The Arabidopsis Information Resource (TAIR) website (Garcia-Hernandez et al. 2002). To search against Persian walnut protein sequences, we used Arabidopsis NAC protein sequences as a query using a local alignment search tool Basic Local Alignment Search Tool (BLAST), considering those with an E value less than 1 × 10⁻¹⁰. We implemented a profile hidden Markov model (HMM) in HMMER v.3.2.1 for the window (Prakash et al. 2017) with default parameters to search for NAC proteins and NAC domains in the protein family (Pfam) database (El-Gebali et al. 2018).

2.2 Phylogenetic, chromosome location, domain analysis, motif, and gene structure analysis of JrNAC transcription factors

We constructed a neighbor-joining (NJ) tree of 102 JrNAC transcription factors using MEGA v.7.0 software (Kumar et al. 2008; Yan et al. 2019b) with the pairwise deletion of 1000 bootstraps and a Poisson model (Lescot et al. 2002) by using the Pfam webserver (El-Gebali et al. 2018) to search for the presence of potential domains. The simple modular architecture research tool (SMART) program (Schultz et al. 2000) also detected the same domains obtained from Pfam with an E value cutoff of 1.0 to validate the final result. The chromosomal locations of NAC transcription factors were searched against J. regia whole-genome sequence using BLASTN. A conserved domain database (CDD) search was conducted in NCBI (Marchler-Bauer et al. 2016). The whole coding sequence (CDS) database was downloaded (https://treegenesdb.org/FTP/Genomes/Jure/v1.4/annotation/). The exon and intron structures were displayed using the online gene structure display server (Hu et al. 2014). The genome browser was used to search for related Persian walnut gene sequences. The motif identification used the MEME program with default parameters, the maximum number of motifs (20), and the optimum motif width (30–50) (Bailey et al. 2015).

2.3 Synteny analysis and calculating K_a, K_s, and K_a/K_s values of duplicated gene pairs

Identification of potential pairs of homologous genes across multiple genomes (E < 1 × 10⁻⁵, top 5 matches) was performed using BLASTP. We used homologous gene pairs to identify syntenic chains through MCScanX (Wang et al. 2012). We detected duplicate gene pairs by using MCScanX, which included whole-genome duplication (WGD), tandem duplication, segmental, and other types of gene pairs. To evaluate the type of NAC gene selected, we used the ratio of non-synonymous substitutions to synonymous substitutions (K_a/K_s) using DnaSP software (Librado and Rozas 2009).

2.4 Cis-element analysis and GO annotation

To conduct cis-element analysis, 1500 bp upstream of the NAC genes of each species were analyzed using Plant-CARE (plant cis-acting regulatory element) with default parameters (Lescot et al. 2002). Blast2GO v2.5 with a cutoff E value of 1 × 10⁻⁶ was used to performed gene ontology (GO) annotations (Conesa et al. 2005). First, we performed BLASTP analysis with an E value of 1e-05 using the JrNAC protein sequence. The analysis was then carried out with the GO mapping. After that, the GO annotation program was used to obtain the GO annotation of the JrNAC members. Finally, the GO enrichment analysis was carried out through the online GO enrichment program on the OmicShare website (https://www.omicshare.com/ tools/Home/Sof/gogsea) (Table 1).

Table 1 The synonymous (K_s) and non-synonymous (K_a) substitution rates for each gene pair and the estimated time of the tandem and segmental replication events of NAC genes

2.5 Plant materials, treatments, and collections

In this study, the first opening of female flowers occurred on 10 April, 15 April, and 22 April, and full opening of female flowers occurred specifically on 15 April and 22 April, when the stigma was not fully developed, and 1 May was the end date. A total of 26 samples were collected from Persian walnut in this study, including 15 female flowers at different stages, 3 male flowers, 3 leaves, and 3 hulls. The female flowers were collected on 23 March, 1 April, 8 April, 16 April, and 23 April as 3 replicates (Table 2), and the male flowers were collected on 10 April, 11 April, and 2 May. The leaves were collected on 23 April as 3 replicates, and the hulls were downloaded from https://treegenesdb.org/FTP/Genomes/Jure.v1.0/transcriptome/rawreads/ (Martínez-García et al. 2016). After harvesting, the pericarp was immediately dissected, and the flesh was frozen in liquid nitrogen and stored at − 80 °C. The three leaves downloaded from the website and the other 23 tissues’ total RNA were isolated using RNA-prep Pre-Plant Kit (Tiangen, Beijing, China) (Zhu et al. 2011). Finally, 23 libraries were constructed and sequenced on an Illumina HiSeq 2500 platform. Analysis of differential gene expression (DESeq) was carried out using the package DESeq R v.1.1.1. Genes found by DESeq with adjusted P values > 0.05 were allocated as differentially expressed (Khan et al. 2020).

Table 2 A total of 26 samples of common walnut used for expression profiling in this study

2.6 Quantitative real-time PCR

To verify the expression pattern, we used the tissues of female flowers, male flowers, and leaves ( for details, see Table 2). Upon dilution to 1:10 with sterile water, the synthetic cDNA was used as the qRT-PCR template. iQ™ SYBR® Green Supermix was used to performed qRT-PCR (Cat. 170-8880AP; Bio-Rad). PCR was conducted on a Light Cycler 480 Real-Time PCR system (Roche Diagnostics, Laval, QC, Canada). For internal control of gene, 18S rRNA was used (Xu et al. 2012). Details of primer information are listed in Table 3. Before the experiment, primer specificities and corresponding melting curves were verified. Each experiment was conducted in triplicate.

Table 3 Primers used for quantitative real-time PCR

2.7 Interaction network of JrNAC proteins

Persian walnut NAC protein matched a homologous Arabidopsis NAC protein in the BLASTP program with an E value of 1e-05 (Aoki et al. 2007, Camacho et al. 2009). Of the Arabidopsis NAC proteins that represent the walnut NAC proteins, 102 were uploaded to the STRING website to predict protein interactions (https://string-db.org/) using the input proteins of J. regia and six predicted input proteins (Szklarczyk et al. 2016).

3 Results

3.1 Identification, phylogenetic relationship, and chromosome location of NAC transcription factor family in J. regia

We identified a total of 102 NAC genes based on the J. regia whole reference genome (Fig. 1; Khan et al. 2020). The neighbor-joining (NJ) phylogenetic tree showed that the JrNAC genes are divided into ten subfamilies in Persian walnut (Fig. 1). In the phylogenetic tree, subfamilies IV and VII with 16 NAC family members were the largest clades. III and X, with five NAC family members, were the smallest clades. Moreover, the numbers of subfamilies I, II, V, VI, VIII, and IX were 15, 15, 10, 6, 7, and 6, respectively. The JrNAC genes were then renamed according to their location on the chromosome (Khan et al. 2020). A map of the NAC genes’ physical positions was created based on the Persian walnut genome physical location information (Khan et al. 2020). Our results show that the NAC genes were distributed unevenly on 16 chromosomes of Persian walnut (Khan et al. 2020). A maximum number of 14 JrNAC genes were present on chromosome 10, followed by 13 on chromosomes 1, representing 27/102 (26.5%) of the total JrNAC genes, followed by 11 each on chromosomes 2 and 11; 2 each on chromosomes 8, 11, and 12; and a minimum of 1 JrNAC on chromosome 5 (Khan et al. 2020).

Fig. 1

Phylogenetic analysis of NAC proteins in Persian walnut (102). A neighbor-joining (NJ) tree was constructed using 102 NAC sequences. The tree further clustered into 10 subfamilies

3.2 Conserved motifs and gene structure analysis

A total of 20 different motifs were detected (Fig. 2 a and b). Ninety-eight of 102 NAC genes contained at least four main motifs (motifs 1, 2, 3, and 5) (Fig. 2b). Motifs 11, 12, 17, and 19 were present in 7 JrNACs (JrNAC13-3, JrNAC16-2, JrNAC16-3, JrNAC12-2, JrNAC7-3, JrNAC2-2, and JrNAC9-3) (Fig. 2b). Structural analysis of exons–introns indicates that the number of exons varies from 1 (on JrNAC3-7) to 22 (on JrNAC1-6) (Fig. 2c). Moreover, two exons were found on one NAC, three exons on 62 NAC genes, four exons on 12 NAC genes, and five exons on six NACs (Fig. 2c). The results show that genes on the same branch might show similar organization of exons and introns (Fig. 2c).

Fig. 2

The phylogenetic relationship, motif compositions, and gene structure of JrNAC transcription factors. a The phylogenetic relationships of JrNACs based on the NJ method. The various colors characterize the ten subfamilies. b Motif compositions of JrNACs. Gray lines indicate non-conserved sequences, and colored boxes represent conserved motifs. The motifs are displayed proportionally in each protein. c Gene structure of JrNACs. CDS represents coding sequence, green boxes indicate CDS, and gray lines represent introns; 0, 1, and 2 represent different types of phase. Phase 0, located between two consecutive codons; phase 1, splitting codon between the first and second nucleotides; Phase 2, between the second and third nucleotides of a codon

3.3 Paralogous NACs, gene duplication, and synteny analysis of JrNAC genes

We identified nine pairs of paralogous NAC genes in Persian walnut (Fig. 3a and Table 1). The ratio of K_a/K_s of nine pairs of paralogous JrNAC gene pairs was less than 1 based on the synonymous (K_s) and non-synonymous (K_a) estimation (Table 1), indicating that these genes are under negative selection. The MCScanX analysis showed that a total of 30 pairs of genes undergo whole-genome duplications (WGDs), while tandem duplication was observed for one pair of genes (JrNAC2-10 and JrNAC9-8), and 60 genes were dispersed (Fig. 3b; Khan et al. 2020).

Fig. 3

Paralogous genes, duplicate type, and synteny analysis of NAC genes between J. regia and three groups of representative plant species. a Schematic representation for the chromosomal distribution and interchromosomal relationships of Persian walnut NAC genes. Gray lines indicate duplicated NAC gene pairs. The chromosome number is indicated at the bottom of each chromosome. b The 102 JrNAC member duplicate type in Persian walnut. c Synteny analysis of NAC genes between J. regia and three plant species. Gray lines in the background indicate the collinear blocks within Persian walnut and other plant genomes, while the red lines highlight the syntenic NAC gene pairs. The species names with the prefixes “P. trichocarpa,” “O. europaea,” and “Q. robur” indicate Populus trichocarpa, Olea europaea, and Quercus robur, respectively

Syntenic analysis was conducted for three plants (Populus trichocarpa, Olea europaea, and Quercus robur) to infer the NAC genes’ evolutionary relationship among these three species (Fig. 3c). We identified 42 pairs of orthologous NAC genes between J. regia and P. trichocarpa, a total of 23 orthologous gene pairs between J. regia and Q. robur, and only 4 orthologous gene pairs between J. regia and O. europaea (Fig. 3c). Interestingly, 16 collinear gene pairs identified between J. regia and Q. robur were not found between J. regia and O. europaea/P. trichocarpa, 7 collinear gene pairs identified between J. regia and P. trichocarpa were not found between J. regia and O. europaea/Q. robur, and 4 collinear gene pairs identified between J. regia and O. europaea were not found between J. regia and P. trichocarpa/Q. robur (Fig. 3c). However, one collinear gene pair was found in three species, J. regia, P. trichocarpa, and O. europaea, and three collinear gene pairs were found in three species, J. regia, P. trichocarpa, and Q. robur (Fig. 3c).

3.4 Promoter region analysis and conserved domains of JrNACs

We found that JrNAC1-4 contained auxin response component, while the gene JrNAC2-6 has a protein-binding site based on in silico analysis (Fig. 4a). The JrNAC promoter regions contain 77% of light-responsive elements, 13% of site-binding-related elements, 7% of environmental stress-related elements, and 3% of the response to plant growth (Fig. 4b).

Fig. 4

Cis-element analysis in the JrNAC gene family. a Each gene cis-element in the phylogenetic tree of walnut. b The percentage of responsive elements: hormone-responsive elements, environmental stress-related elements, plant growth responsive elements, and site-binding responsive elements (except for TATA and CAAT binding sites) in all JrNAC genes

All the JrNACs contained the NAM conserved domain, and six JrNAC transcription factors contained special domains at their C-terminal (PEP_TPR_lipo, FBA_1, BAR, SET, Rubis-subs-bind, and CBF5) based on conserved domains analysis (Khan et al. 2020). The analysis of GO enrichment was divided into three categories. In the category of biological processes, bio-regulation, metabolic processes, cellular processes, and stimulus response are significantly enriched terms. In the cellular component category, cell, organelle, and cell parts are significant. GO terms for the transcription factor activity of nucleic acid binding were highly represented in the molecular function category. The most enriched GO term in members of the JrNAC is GO: 003674 (molecular function).

3.5 Expression profiles and qRT-PCR analysis of JrNACs among different tissues and flowering development stages

We investigated the transcriptome expression profiles in different tissues of the Persian walnut to further provide information on the function of the JrNAC transcription factor family. The JrNACs were differently expressed in different tissues, indicating that these genes have a function variation. Twenty-four of 102 Persian walnut NAC members were expressed highly in female flowers compared with other tissues (M, L, and H), particularly JrNAC1-4, JrNAC16-8, JrNAC16-9, JrNAC16-1, JrNAC6-5, JrNAC10-13, JrNAC9-6, and JrNAC2-6. Two JrNACs (JrNAC2-9 and JrNAC9-6) were highly expressed in male flowers, four genes in leaves and two genes highly expressed in the hull; these results show that the NAC genes in different tissues have a different expression pattern (Fig. 5a, Khan et al. 2020). The expression of some genes increases as the flowers grow, such as JrNAC1-13, JrNAC16-9, JrNAC16-5, JrNAC16-8, JrNAC7-1, JrNAC2-9, and JrNAC9-6 (Fig. 5b).

Fig. 5

Expression profiles of the Persian walnut NAC genes. a Hierarchical clustering of expression profiles of 14 Persian walnut NAC genes in 12 samples including F, M, L, and H represents female flower, male flower, leaves, and hull as 3 biological replicates. b The heatmap exhibits the ratio of the expression levels of 12 NAC genes in five developmental stages; 1, 2, 3, 4, and 5 represent different stages of Persian walnut flower. c Expression analysis of 3 JrNACs in three representative samples by qRT-PCR. Data were normalized to the β-actin gene, and vertical bars indicate standard deviation

The transcription levels of JrNAC1-4 and JrNAC2-6 in flowering and vegetative tissues were analyzed using qRT-PCR. JrNAC1-4 and JrNAC2-6 were highly expressed in female compared with male flowers, while JrNAC13-5 was expressed highly in leaves of J. regia (Fig. 5c). These genes are differentially expressed in female and male flowers and leaves (Fig. 5c), which could be subsequently prioritized in plant functional studies for further analysis.

3.6 The interaction network of JrNAC proteins

Each JrNAC protein was in close association with at least one NAC protein from Arabidopsis. Some JrNAC proteins were closely aligned with the same NAC protein in Arabidopsis. We downloaded NAC proteins from the Arabidopsis to detect the predicted role of highly expressed genes in the flowering of Persian walnut. The previous study claimed that these genes regulate the development of the flower. Therefore, we detected the interaction relationship between these genes, and the results indicate a strong relationship between the JrNAC1-4 proteins and AtWRKY12 (Fig. 6) (Li et al. 2016).

Fig. 6

JrNAC protein interaction network. The JrNAC protein interaction network was constructed using Arabidopsis homologous NAC proteins. The corresponding relationship between the walnut NAC protein and the Arabidopsis NAC protein is shown at the top left of the figure. Proteins are represented by network nodes. The 3D protein structure is displayed inside the nodes. Edges represent associations of proteins

4 Discussion

4.1 The number, phylogenetic relationships, and location of JrNACs in Persian walnut

In this study, we identified a total of 102 NAC genes in Persian walnut. Previously, a large number of NAC genes identified in other plants and contained over 100 members (Ooka et al. 2003; Nuruzzaman et al. 2010; Liu et al. 2014; Peng et al. 2015; Saidi et al. 2017). The number of NAC genes in Persian walnut is lower as compared with other plants including B. pekinensis (204) (Liu et al. 2014), O. sativa (151) (Nuruzzaman et al. 2010), G. max (152) (Le et al. 2011), P. trichocarpa (163) (Hu et al. 2010), Arabidopsis (105) (Ooka et al. 2003), P. mume (113) (Zhuo et al. 2018), and M. domestica (180) (Jia et al. 2019).

Based on phylogenetic analysis, the JrNACs were divided into 10 distinct subfamilies. The phylogenetic tree obtained in this study mainly aligned with previous reports (Ooka et al. 2003; Shen et al. 2009). In Persian walnut, all NAC gene motifs 1, 2, 3, and 5 were frequent. The gene structure analysis showed that exon numbers vary from 1 to 22, and this number is greater than in M. acuminata, in which the number of exons varies from 0 to six (Cenci et al. 2014); in O. sativa, where the number ranges from 0 to 16; and in G. hirsutum, where the number is 0 to nine (Nuruzzaman et al. 2010; Zhu et al. 2011). The Persian walnut 62 NAC genes mainly contain three exons and two introns. In Arabidopsis, G. hirsutum, M. acuminata, and O. sativa, this phenomenon was observed, and three exons were present in the majority of NAC genes (Nuruzzaman et al. 2010; Zhu et al. 2011; Cenci et al. 2014). The motif compositions, exon–intron structure, and intron phase 0 of JrNACs indicate that NAC was highly conserved in each subfamily.

In our study, the NAC genes were unevenly distributed on 16 chromosomes but most commonly observed on chromosome 10 (14 NAC genes) with comparatively few NACs on chromosome 5 (1 NAC gene) (Khan et al. 2020). In the previous studies, for example, in O. sativa and Z. mays NAC genes, uneven distribution on chromosomes was also reported (Nuruzzaman et al. 2010; Peng et al. 2015).

4.2 The evolution and expansion of JrNACs

Tandem duplication, segmental duplication, and WGD were most likely chosen by gene families as forms of expansion (Cannon et al. 2004; Dong et al. 2017). However, a total of 30 JrNACs genes pair were duplicated by WGD, and only one gene pair (JrNAC2-10 and JrNAC9-8) experienced tandem duplication events. These results indicate that the evolutionary expansion patterns of NAC transcription factor family members were duplicated by WGD events (Khan et al. 2020; Cannon et al. 2004). This finding contrasts with several previous reports in which a similar phenomenon was analyzed (Zhu et al. 2011; Satheesh et al. 2014). For example, the expansion of the Populus and Gossypium NAC genes were tandem duplication events (Zhu et al. 2011; Satheesh et al. 2014).

There are collinear genes between the Persian walnut and P. trichocarpa, O. europaea, and Q. robur; these results suggest that the NAC genes may have evolved from a common ancestor in different plants (Khan et al. 2020; Satheesh et al. 2014). The K_a/K_s values of the paralogous gene pairs were calculated to estimate their evolution history, and we found that JrNAC genes evolved through negative selection as the K_a/K_s values were less than 1. We found that the JrNAC promoter regions contain 77% of light-responsive elements, 7% of environmental stress-related elements, and 3% of the response to plant growth, while the JrNAC1-4 gene promotes the auxin response component based on in silico analysis (Khan et al. 2020). These results indicate that NAC members of Persian walnut experienced the selection of ecological factors, especially for light and environmental stress (Tran et al. 2004; Tran et al. 2010; Puranik et al. 2012; Satheesh et al. 2014).

4.3 The expression profile of NAC members of Persian walnut

The NAC transcription factor family has diverse functions and plays important roles in plant development and physiological processes (Souer et al. 1996; Sablowski and Meyerowitz 1998; Ooka et al. 2003; Guo and Gan 2006; Liu et al. 2009; Wang et al. 2009; Shan et al. 2012; Wang et al. 2013; Satheesh et al. 2014), involving in at least four types of processes, such as flower formation, cotyledon development, shoot apical meristem maintenance, and subsequent development of the root (Sablowski and Meyerowitz 1998; Takada et al. 2001; He et al. 2005). In this study, a total of seven conserved domains, including one conserved domain (NAM) and six special domains, were identified (Khan et al. 2020). In Z. mays, 13 distinct conserved domains and at least six conserved domains were identified in M. notabilis (Peng et al. 2015; Baranwal and Khurana 2016). Our results are similar to those for O. sativa and S. officinarum, which also have seven conserved domains (Ooka et al. 2003; Ramaswamy et al. 2017). All the JrNACs contain the NAM domain, which is consistent with previous studies and is mainly associated with DNA binding and flower development (Hussain et al. 2017; Ramaswamy et al. 2017). In Petunia, G. max, and P. trichocarpa, the NAM domain has a role in flower formation, primordia, and embryo development (Souer et al. 1996; Hu et al. 2010; Hussain et al. 2017), indicating that the JrNACs may play a potential role in the flowering of J. regia. A comprehensive analysis was conducted to evaluate the expression patterns of the JrNAC gene family in vegetative and reproductive tissues of Persian walnut. Based on the expression analysis, a total of 24 JrNAC genes showed higher expression levels in female flowers, indicating that these genes play a key role in the development of flowers (Ikeda et al. 2004; Tran et al. 2009; Zhou et al. 2010; Su et al. 2013; Singh et al. 2013; Kou et al. 2014; Zhuo et al. 2018), and the results for two genes (JrNAC1-4 and JrNAC2-6) were also supported by qRT-PCR.

The results of the interaction relationship indicate a strong relationship between the JrNAC1-4 proteins and the AtWRKY12 proteins (Li et al. 2016). Notably, JrNAC1-4 contained part of an auxin-responsive element and has a systemic relationship with P. trichocarpa (Potri.007G099400.1) and Q. robur (Qrob_P0657980.2). JrNAC1-4 was highly expressed in female and male flowers, with extremely low expression in leaves, which is consistent with the previous studies. For example, O. sativa (ONAC300) (Zhou et al. 2010), G. max (GmNAC016 and GmNAC14) (Tran et al. 2009), Arabidopsis (NAP) (Sablowski and Meyerowitz 1998), S. tuberosum (StNAC034 and StNAC075) (Singh et al. 2013), M. domestica (MdNAC42, MdNAC110, and MdNAC138) (Su et al. 2013), S. lycopersicum (SNAC8) (Kou et al. 2014), Fragaria (FaNAC021, FaNAC022, FaNAC042, and FaNAC092) (Moyano et al. 2018), and P. mume (PmNAC) (Zhuo et al. 2018) also showed higher expression in flowers and also showed high similarity with AtWRKY12, a gene that regulates the development of flowers (Li et al. 2016). These results show that JrNAC1-4 might play a potential role in Persian walnut flowering. Taken together, our results suggest that JrNAC1-4 may play important roles in Persian walnut flowering development.

5 Conclusion

In Persian walnut (J. regia), we identified a total of 102 NAC transcription factors. Phylogenetic analysis showed that the NAC transcription factors are clustered into 10 subfamilies. Based on the conserved domains, NAC transcription factors contain a conserved domain (NAM). The analysis of the expression profile showed that the NAC transcription factors reveal diverse patterns of expression in different Persian walnut tissues. Most of the Persian walnut NAC transcription factors are expressed highly in female and male flowers. A total of 24 NAC transcription factors were highly expressed in female and male flowers, which might play a role in J. regia flowering. The transcription data and qRT-PCR analysis indicated that two NAC transcription factors (JrNAC1-4 and JrNAC2-6) were highly expressed in female and male flowers, while JrNAC13-5 was expressed highly in leaves. In conclusion, these results provide a base for studying the potential function of Persian walnut NAC transcription factors.