Background

WRKY transcription factors (TFs) play crucial roles in plant growth, development, and response to abiotic and biotic stresses [1,2,3,4]. These abiotic and biotic stresses include drought, salt, extreme temperatures, waterlogging, ultraviolet, and various pathogen and insects [1,2,3]. The plant developmental stages regulated by WRKY TFs include flowering time, senescence, nutrient utilization, and the development of seeds, pollens, stems, and roots plant [2, 3]. WRKY TFs have a conserved WRKY domain, containing a WRKYGQK motif at the N-terminal end and a zinc finger motif at the C-terminal end [1]. WRKY TFs are classified into three groups; I, II, and III [1]. Group I WRKY TFs contain two domains and a zinc finger (C2H2) motif [1, 2], while group II and III contain a WRKY domain but have different zinc finger motifs (group II contains C2H2, whereas group III contains C2HC zinc finger motif) [1, 2]. Group II is further classified into five subgroups, IIa-IIe [1, 2], of which subgroups IIa and IIb cluster in one clade, and IId and IIe in a different clade [1, 2, 5].

Several WRKY TFs have been identified in diverse plant species at the genome level [6, 7]. Notably, most studies have mainly focused on the structural variation and evolution of WRKY and the prediction of their biological functions based on RNA-seq and quantitative real-time PCR analyses. WRKY TFs have been identified in at least 12 legumes, including Arachis duranensis, Arachis ipaensis, Cajanus cajan, Cicer arietinum, Glycine max, Lotus japonicas, Lupinus angustifolius, Medicago truncatula, Phaseolus vulgaris, Trifolium pratense, Vigna angularis, and Vigna radiate [6]. The WRKYGQK domain reportedly tends to mutate into WRKYGKK [6]. Duplicated WRKY TFs of the 12 legumes had longer polypeptides than the single WRKY TFs [6]. Synteny analysis revealed that segmental duplication event plays a major role in paralog formation in G. max, A. duranensis, and A. ipaensis [8, 9]. Moreover, accumulating evidence also demonstrated that WRKY paralogs and orthologs mainly underwent purifying selection, suggesting that WRKY homologs have conserved functions [6, 8].

Cultivated peanut (Arachis hypogaea) is an important oil and protein crop from South America [10, 11]. It is an allotetraploid plant that resulted from a cross between A. duranensis and A. ipaensis [11,12,13]. A. monticola is a wild allotetraploid plant known to be the direct progenitor of A. hypogaea [14]. To date, genome sequencing of at least six Arachis species has been completed, including A. duranensis, A. ipaensis, A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner [11, 12, 15,16,17]. The genomic information of the six Arachis species provides crucial data for evolutionary studies at the genome level. In A. duranensis, duplicated gene pairs have different responses to drought and nematode stress, and old and young duplicate genes have divergent functions [18, 19]. Old duplicate genes mainly participate in lipid and amino acid metabolism and responses to abiotic stresses, while young duplicate genes are preferentially involved in photosynthesis and biotic stress responses [19]. In A. duranensis and A. ipaensis, gradual selection and purifying pressure act on the somatic tissue-specific and sex-specific genes [20]. A comparison of the genomic structure between wild and cultivated peanuts revealed that the sub-genomes of cultivated peanuts underwent asymmetric evolution [21]. However, no homoeolog expression bias was observed in vegetative tissues between two sub-genomes of A. hypogaea except in reproductive tissues [12, 15, 16, 22]. Whole-genome re-sequencing of 203 cultivated peanut varieties was performed to verify the botanical classification of peanuts, and the results revealed that var. peruviana is possibly the earliest variant from tetraploid progenitors [23]. In addition, seed weight and length-related genes have been identified using genome-wide association analysis, and their functions have been verified in Arabidopsis [23].

In addition to genome-level analysis, gene family identification has also been used to study Arachis evolution. A. duranensis, A. ipaensis, and A. hypogaea cv. Tifrunner genomes have been used to identify gene families such as nucleotide-binding site-leucine-rich repeat (NBS-LRR), LRR-containing genes, and heat shock transcription factor (HSF) [24,25,26]. To our knowledge, only the valine-glutamine (VQ) gene family has been compared among the above-mentioned six Arachis genomes [27]. The study found that the VQs increased in A. monticola, A. hypogaea cv. Fuhuasheng, and A. hypogaea cv. Shitouqi compared to A. duranensis and A. ipaensis [27].

Previous studies identified WRKY TFs in A. duranensis, A. ipaensis, and A. hypogaea cv. Tifrunner [8, 28]. However, some coordinates changed, and a few gene models got duplicated in A. hypogaea cv. Tifrunner genome [29]. Therefore, this study aimed to identify WRKY TFs in A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner. We compared the number of WRKY TFs across the various Arachis species to determine their homologous relationships and regulatory networks. Therefore, this study provides new insights into the evolution of Arachis spp.

Methods

Identification of WRKY TFs in Arachis species

The released genome sequences of A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner were obtained from GigaDB (http://gigadb.org/dataset/100453), NCBI (ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCA/004/170/445/GCA_004170445.1_ASM417044v1), Peanut Genome Resource (http://peanutgr.fafu.edu.cn), and PeanutBase (https://www.peanutbase.org) databases [12, 15,16,17, 29]. The Hidden Markov Model (HMM) file of WRKY domains (PF03106) was downloaded from the Pfam database [30], and the HMMER program with default parameters was used to identify the WRKY domains among the Arachis spp. [31]. The WRKY sequences were extracted using in-house Perl script and were uploaded to the Pfam database to re-confirm the WRKY domains. The identification of WRKY TFs in A. duranensis and A. ipaensis was based on a previous study [8].

Phylogenetic tree construction

Six Arachis WRKY domains were aligned using MAFFT program [32], and the ProtTest program was used to estimate the best-fit model of maximum likelihood (ML) trees [33]. The ML trees were constructed using the IQ-tree program [34]. The phylogenetic tree visualized using the Figtree program.

Identification of WRKY paralogs and homoeologs in Arachis species

Paralogs occur due to gene duplication events, while homoeologs are formed via polyploidy [35, 36]. In this study, we identified paralogs and homoeologs of the Arachis species using the local BLAST program as per the following parameters: (1) the alignment region exceeds 80% of each sequence, (2) sequence identity over 80%, and (3) E-value ≦ 10–10 [18, 20, 37, 38].

W-box cis-acting elements of WRKY genes in Arachis species

WRKY TFs are auto- and cross-regulated by the W-box cis-acting elements [27, 39, 40]. In this study, the 2-kb upstream sequences of WRKY genes were extracted using the genetic feature format (GFF) by the TBtools program [41]. These sequences were uploaded to the NSITE web service to predict their WRKY binding sites [42].

Prediction of protein–protein interaction among the WRKY TFs of Arachis species

The WRKY TFs were uploaded to the STRING database, and the A. hypogaea WRKY sequences were used as a reference for predicting the protein–protein interactions.

Results

New WRKY TFs originated from tetraploid Arachis species

The WRKY domains were contained in 138, 131, 158, and 146 sequences of A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner, respectively (Fig. 1A and Table S1). Among them, 124, 131, 158, and 139 WRKY TFs in A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner, respectively, were full-length sequences (Fig. 1A). A previous study identified 75 (70 full-length sequences) and 77 (69 full-length sequences) WRKY TFs in A. duranensis and A. ipaensis, respectively [8]. In this study, the tetraploid and diploid peanut species had an equal number of WRKY TFs (Fig. 1). Similarity, an equal number of WRKY TFs was identified between the sub-genomes and their corresponding ancestral donors (Fig. 1B). The WRKY TFs were classified into three groups: I, II, and III, according to the WRKY domain number and zinc finger type [1, 2]. The number of WRKY TFs in tetraploid peanut species ranged from 16–25, 90–111, and 22–29 in groups I, II, and III, respectively (Fig. 1C). Moreover, the number of WRKY TFs in groups I, II, and III was equal between the tetraploid and diploid peanut species except for group I WRKY in A. hypogaea cv. Fuhuasheng (Fig. 1C).

Fig. 1
figure 1

Comparison of WRKY genes across various Arachis species. A Number of WRKY genes across various Arachis species. B Number of WRKY genes across various Arachis sub-genomes. The excluded WRKY genes from Arachis monticola and A. hypogaea cv. Fuhuasheng due to lack of location information. C Number of WRKY genes in groups I, II, and III across various Arachis species. Statistical analyses were executed using the Chi-square test at p ≦ 0.05

Full size image

We constructed ML phylogenetic trees using the WRKY domains of Arachis spp. and the results showed that WRKY domains were clustered in three major groups: I, II, and III. The group WRKY II domains were further classified into five subgroups: IIa, IIb, IIc, IId, and IIe, consistent with previous studies [1, 2]. However, several group II members from the tetraploid peanut species did not cluster with the corresponding group members from the diploid peanut species (Fig. 2 and Table S1). This indicated that novel WRKY TFs originated from the tetraploid peanut. In addition, several members of subgroups IIb and IIc were clustered with those in subgroups Ic and In, respectively (where In and Ic represent group I members with their WRKY domain on the N-terminal and C-terminal ends (Fig. 2). Several groups Ic and In members also clustered with group IIc and III (Fig. 2). These results indicate that Arachis WRKY TFs have multiple origins and that new WRKY TFs originated from the tetraploid peanuts as opposed to diploid peanut.

Fig. 2
figure 2

Phylogenetic analysis of WRKY genes among various Arachis species. The maximum likelihood phylogenetic tree was constructed using the IQ-tree program, and the best-fit model (JTT + I + G) was generated by the ProtTest program. I-N and I-C indicate group I members with WRKY domains from the N- and C-terminal ends

Full size image

Old WRKY TFs were lost in tetraploid Arachis species

Gene expansion and loss occur after a polyploidy event [43], and A. duranensis and A. ipaensis are the progenitors of tetraploid peanuts [11,12,13]. Ideally, tetraploid peanut species inherited all the WRKY TFs from wild diploid peanut species; however, only 44 WRKY TFs from two wild diploid peanut species had conserved orthologs with four tetraploid peanut species (Fig. 3A and B). A. monticola is known to be the direct ancestor of the cultivated peanut species [14]. In this study, 55 WRKY TFs from A. monticola had conserved orthologs in three cultivated peanuts species (Fig. 3A and B). Among the tetraploid peanut species, 96 WRKY orthologous gene pairs were distributed across three cultivated peanuts (Fig. 3A and B). These results indicate that ancestral WRKY TFs were lost after tetraploid formation. Compared with diploid peanut species, cultivated peanuts retained more WRKY TFs from A. monticola.

Fig. 3
figure 3

Homologous WRKY genes across Arachis species. A Paralogous, homoeologous, and orthologous WRKY genes among various Arachis species. B Conserved orthologous WRKY gene pairs across various Arachis species. C Paralogous and homoeologous WRKY gene pairs across various tetraploid peanut species

Full size image

Phenotypic variations such as leaf size, seed size, oil content, flowering pattern, and testa color have been observed across cultivated peanut varieties [10, 44,45,46]. Notably, A. hypogaea cv. Fuhuasheng and A. hypogaea cv. Shitouqiare are the breeding parents of about 70% of Chinese peanut cultivars [15, 16]. A. hypogaea cv. Tifrunner is a commercial cultivar in America with high disease resistance [47]. Therefore, the three peanut cultivars underwent different evolutionary processes. In this study, orthologs, paralogs, and homoeologs were identified across the tetraploid peanut species. In total, 26, 22, 31, and 23 WRKY homoeologous gene pairs were identified in A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner, respectively (Fig. 3A and C). Meanwhile, 21, 25, 41, and 31 WRKY paralogous gene pairs were identified in A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner, respectively (Fig. 3A and C). Compared with A. monticola, A. hypogaea cv. Fuhuasheng and A. hypogaea cv. Tifrunner lost WRKY homoeologs but gained WRKY paralogs, while A. hypogaea cv. Shitouqi gained WRKY homoeologs and paralogs (Fig. 3C).

Additionally, 46 WRKY orthologous gene pairs were identified between A. duranensis and A. ipaensis. Homoeologous WRKY TFs were lost in the tetraploid peanut species compared to the two wild diploid peanut species; however, paralogous WRKY TFs were produced and retained in tetraploid peanut species. Although there was no difference in the number of WRKY TFs between the tetraploid peanut species and their diploid donors, the tetraploid peanut species lost and retained some WRKY TFs, indicating a dynamic equilibrium of WRKY TFs in tetraploid peanut species.

Domestication affected WRKY evolution in peanut

WRKY TFs exert their biological functions, including auto- and cross-regulation, by binding the W-box elements of WRKY genes [3, 7]. In this study, 26, 29, 59, 69, 82, and 64 WRKY genes in A. duranensis, A. ipaensis, A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner contained at least one W-box element (Fig. 4A). W-box elements of WRKY genes were compared among the orthologs, paralogs, and homoeologs of Arachis species. The results showed that W-box elements in orthologous WRKY genes differed between the diploid and tetraploid peanut species (Fig. 4B). However, matching W-box elements of orthologous WRKY genes were found among A. monticola, A. hypogaea cv. Fuhuasheng, and A. hypogaea cv. Tifrunner (Fig. 4B). There were 26.92% (7/26), 36.36% (8/22), 48.39% (15/31), and 43.48% (10/23) homoeologous WRKY gene pairs with matching W-box elements in A. monticola, A. hypogaea cv. Fuhuasheng, A. hypogaea cv. Shitouqi, and A. hypogaea cv. Tifrunner, respectively (Fig. 4C). This indicated that cultivated peanuts retained more homoeologous W-box elements than wild tetraploid peanuts. In paralogous WRKY genes, matching W-box elements were distributed across A. monticola (38.10%, 8/21), A. hypogaea cv. Fuhuasheng (24.00%, 6/25), A. hypogaea cv. Shitouqi (26.83%, 11/41), and A. hypogaea cv. Tifrunner (25.81%, 8/31) (Fig. 4C). The results indicated that cultivated peanuts lost more paralogous WRKY genes with similar W-box elements than wild tetraploid peanuts. Overall, these results indicated that domestication possibly affected the loss and retention of W-box elements in peanuts.

Fig. 4
figure 4

The W-box elements of WRKY genes in various Arachis species. A WRKY genes containing W-box elements. B The number of orthologous WRKY gene pairs containing matching W-box elements. C The number of paralogous and homoeologous WRKY gene pairs containing matching W-box elements

Full size image

Studies showed that WRKY-WRKY protein interaction mediates certain biological functions [27, 39]. In this study, protein–protein interactions were assessed using WRKY TFs from diploid and tetraploid peanut species. Compared with WRKY TFs in A. duranensis and A. ipaensis, more WRKY-WRKY interaction complexes were formed in tetraploid peanut species (Fig. 5). These results indicated that the complex relationships could be due to allopolyploidy. Notably, cultivated peanuts had more complex WRKY protein–protein interactions than wild tetraploid peanuts (A. monticola) (Fig. 5), indicating that domestication possibly affects WRKY protein–protein interaction. Furthermore, similar WRKY protein–protein interaction patterns were detected between A. hypogaea cv. Fuhuasheng and A. hypogaea cv. Shitouqi, while A. hypogaea cv. Tifrunner exhibited a distinct WRKY protein–protein interaction relationship.

Fig. 5
figure 5

WRKY-WRKY protein interactions across various Arachis species. A WRKY-WRKY protein interactions in Arachis duranensis. B WRKY-WRKY protein interactions in A. ipaensis. C WRKY-WRKY protein interactions in A. monticola. D WRKY-WRKY protein interactions in A. hypogaea cv. Fuhuasheng. E WRKY-WRKY protein interactions in A. hypogaea cv. Shitouqi. F WRKY-WRKY protein interactions in A. hypogaea cv. Tifrunner. A. hypogaea WRKY sequences were used as a reference for predicting the protein–protein interactions

Full size image

Discussion

Several genomes of Arachis spp. have been sequenced and publicly released [11, 12, 15,16,17], accelerating the identification and comparison of gene families at the genome level [8, 28]. However, two factors should be considered when comparing gene families. First, the method used to identify the gene families should be the same because variations in the methodology have been shown to influence the final results. For example, Zhang, et al. [37] found that HMM-based methods are fast and efficient and that using full-length sequences in evolutionary analyses could eliminate false results [37]. Second, various sequencing methods and assembling strategies should be considered when analyzing different Arachis genomes, or conserved orthologs among Arachis species can be analyzed to avoid variations. A previous study identified 158 WRKY TFs in A. hypogaea cv. Tifrunner [28], whose genome was corrected and publicly released in the Peanutbase database [29]. In this study, we identified 146 WRKY TFs from the updated A. hypogaea cv. Tifrunner genome. Nevertheless, 158 WRKY TFs identified in the previous study contained 146 WRKY TFs from the updated genome. We utilized the same method to identify and analyze the evolution of WRKYs in Arachis spp.

Cultivated peanuts underwent allotetraploidy and domestication [13, 21, 23, 44]. Studies showed that cultivated peanuts have more photosynthetic pigments and larger leaves, stomata, and epidermal cells than their diploid donors because of their allotetraploid genomes [44]. Notably, Leal-Bertioli, et al. [48] compared drought tolerance among A. duranensis, A. ipaensis, synthetic allotetraploid (A. duranensis x A. ipaensis)4x, and A. hypogaea cv. Tifrunner and found that synthetic allotetraploid and A. hypogaea cv. Tifrunner had similar but lower drought tolerance than A. duranensis and A. ipaensis [48]. These findings indicate that the hybrid vigour and not allotetraploidy reduces drought tolerance in tetraploid peanuts more than in their diploid progenitors. Furthermore, a comparison of drought genes between A. hypogaea cv. Tifrunner and two diploid donors showed that A. hypogaea cv. Tifrunner lost ancestral drought genes, while new copies of drought tolerance genes lack origin function after allotetraploidy [49]. We found that tetraploid peanuts lost the old WRKY TFs and retained the new ones. Based on the changes in drought-tolerance genes of A. hypogaea cv. Tifrunner, we hypothesized that new WRKY TFs possibly have new functions and formed complex regulatory networks in tetraploid peanut species.

This study showed that domestication affected WRKY genes in cultivated peanuts. The number of W-box elements in WRKY genes was affected in cultivated peanuts compared with A. monticola. In addition, WRKY protein–protein interaction results showed that A. hypogaea cv. Fuhuasheng and Shitouqi had similar protein interaction relations, while A. hypogaea cv. Tifrunner had a different protein interaction pattern. A. hypogaea cv. Fuhuasheng and Shitouqi are the progenitors of the Chinese peanut [15, 16], suggesting that the two A. hypogaea cultivars possibly underwent a similar domestication process. A. hypogaea cv. Tifrunner was bred in the USA in 2007 and is highly resistant to various diseases, unlike the Chinese peanut [47]. This shows that the differences in the domestication process may be the main reason for the structural and functional variation of the WRKY TFs among the three cultivated A. hypogaea cultivars.

Conclusions

This study identified WRKY TFs in six Arachis species. The number of WRKY TFs and their evolutionary patterns were compared, and the results revealed dynamic equilibrium in the number of WRKY TFs across the six Arachis spp. Notably, new WRKY TFs were retained while the old ones were lost after allotetraploidy. The present study also showed that domestication affected the WRKY TFs of cultivated peanuts. The WRKY TFs of A. hypogaea cv. Fuhuasheng and Shitouqi were subjected to a similar domestication process, while those of cv. Tifrunner underwent a different domestication process based on the protein–protein interaction analysis.