Phylogenetic and Transcription Analysis of Chrysanthemum WRKY Transcription Factors

WRKY transcription factors are known to function in a number of plant processes. Here we have characterized 15 WRKY family genes of the important ornamental species chrysanthemum (Chrysanthemum morifolium). A total of 15 distinct sequences were isolated; initially internal fragments were amplified based on transcriptomic sequence, and then the full length cDNAs were obtained using RACE (rapid amplification of cDNA ends) PCR. The transcription of these 15 genes in response to a variety of phytohormone treatments and both biotic and abiotic stresses was characterized. Some of the genes behaved as would be predicted based on their homology with Arabidopsis thaliana WRKY genes, but others showed divergent behavior.


Introduction
Chrysanthemum (Chrysanthemum morifolium Ramat.) is a leading ornamental species, second only to the rose in terms of its market value [1]. In 2010, more than two billion cut chrysanthemum stems were produced in China, and almost the same number in Japan [2]. Since the major constraints faced by chrysanthemum producers are a range of biotic and abiotic stresses, enhancing the crop's resistance/tolerance to these is an important breeding aim.
The WRKY family is prominent among plant transcriptional regulators, so named because of the presence of the characteristic peptide sequence WRKYGQK [3]. Sweet potato SPF1 was the first WRKY protein to have been isolated [4]. Based on the number of WRKY domains present and the structure of the protein's zinc finger motifs, three major groups of WRKY proteins have been recognized [5]. Group I proteins contain two WRKY domains and a C 2 H 2 zinc finger motif, group IIs a single WRKY domain and a C 2 H 2 zinc finger motif, and group IIIs a C 2 HC zinc finger motif [5]. WRKY transcription factors are known to be involved in the regulation of a number of aspects of plant growth and development, as well as in the response to stress [6][7][8][9]. In Arabidopsis thaliana, for example, they participate in the response to low temperature, drought and salinity [10], while others have been implicated in signaling in the context of pathogen infection [11][12][13][14][15] and herbivore attack [16]. They have been shown to interact with phytohormones, especially salicylic acid (SA) and jasmonate (JA) [17,18].
As yet, however, the various activities of WRKY proteins in chrysanthemum have not been explored. Here, we report the isolation of 15 chrysanthemum WRKY transcription factors, based on a set of transcriptomic data, and have analyzed the effect of various stress and phytohormone treatments on their level of transcription.

The WRKY Gene Content of Chrysanthemum
The 15 WRKY sequences isolated were designated CmWRKY1 through CmWRKY15 (GenBank: KC615355-KC615369). The full length cDNAs varied in length from 757 to 1750 bp, and their predicted protein products comprised between 193 and 504 residues. Full details of the CmWRKY sequences are given in Table 1. A combination of sequence comparison, phylogenetic and structural analyses suggested that the 15 CmWRKY genes were distributed across the three known WRKY groups (Figure 1), and a schematic overview of the core motifs present is shown in Figure S1. There was a high degree of homology between the motifs present in the CmWRKYs and those in the AtWRKYs ( Figure 2). Motif 6 only featured in Group IId, while motif 3 was present in both Groups IIa and IIb ( Figure S1).    [19]. The cyan boxes represent WRKY motif, and other colored boxes each represent a specific motif with uncharacterized function.
Orthology detection is critically important for accurate functional annotation, and has been widely used to facilitate studies on comparative and evolutionary genomics. A trade-off between sensitivity and specificity in orthology detection was observed, with BLAST (Basic Local Alignment Search Tool)-based methods characterized by high sensitivity, and phylogeny-based methods by high specificity [20], so the relationship between the chrysanthemum and A. thaliana WRKY genes was analyzed using both BLAST (which delivers a local sequence alignment) and by a rooted phylogenetic tree (global sequence alignment) (Table 1, Figure 1). Some inconsistencies were noted: for example, CmWRKY7 and AtWRKY69 appeared to be closely related according to the phylogenetic analysis, but the BLAST comparison predicted that the closest A. thaliana sequence to CmWRKY7 was AtWRKY41. Gene function analysis may therefore be needed to conclude which of these relationships is the more likely to be valid.

Transcription Profiling of CmWRKY Genes
The 15 CmWRKY genes were differentially transcribed throughout the plant ( Figure 3). The transcript abundance of CmWRKY14 was more than three orders of magnitude higher than that of CmWRKY2 in the leaf, while CmWRKY10 transcript was only detectable in the leaf under non-stressed conditions. Neither CmWRKY2 nor CmWRKY5 transcript was present in the root. The level of transcription shown by CmWRKY in tube florets was at least double that present in ray florets at budding stage, besides CmWRKY7 was exception.

The Transcription of CmWRKY Genes in Plants Challenged by Phytohormones and Abiotic Stress
Twelve of the 15 genes were down-regulated by exogenous ABA (abscisic acid) (the exceptions were CmWRKY7, 9 and 15), while the abundance of CmWRKY10 transcript was below the level of detection. CmWRKY15 was induced by this treatment, while CmWRKY7 and 9 were also induced, but only after exposure of least 1 h (Figure 4a). None of CmWRKY1, 4, 6, 8, 10 or 12 were responsive to MeJA (methyl jasmonate) treatment, but CmWRKY2, 9, 13 and 15 were induced, while CmWRKY3, 5, 11 and 14 were all repressed (Figure 4b). Eleven of the genes (the exceptions were CmWRKY7, 9, 10 and 11) were repressed after 1 h exposure to SA treatment, but their transcription was triggered after 4 h (Figure 4c).
CmWRKY1, 3,6,9,12,13 and 14 were all up-regulated in the root by salinity stress, while CmWRKY4 was down-regulated. The transcript abundance of CmWRKY7, 11 and 15 was enhanced after 1 h of exposure, but later fell back (Figure 5a). The effect of moisture stress was to up-regulate CmWRKY1, 3,4,6,8,9,12 and 14 in the root by various amounts, while CmWRKY7 transcription was markedly suppressed. Transcription of CmWRKY10 was noted after 4 h of exposure to PEG (polyethylene glycol), but was not transcribed in non-stressed roots (Figure 5b). All of the CmWRKY genes were induced by exposure to low temperature, with the peak transcript abundance occurring after 8 h (Figure 5c). With the exception of CmWRKY7, 9 and 13, the genes were all down-regulated by high temperature; the abundance of CmWRKY2, 11, 12 and 15 transcript was below the level of detection (Figure 5d). Apart from CmWRKY7 and 9, the genes were all down-regulated by wounding (Figure 5e).

Differential Responses of the CmWRKY Genes to Biotic Stress
CmWRKY1, 11 and 15 were strongly induced by the presence of A. tenuissima inoculum, particular CmWRKY15, for which the level of transcript was some 80 fold higher than that of the non-inoculated control after 6 h (Figure 6a). CmWRKY1, 6 and 8 were all induced by about 10 fold when assayed 4 h after inoculation with F. oxysporum (Figure 6b). CmWRKY4, 8 and 11 were markedly suppressed by P. horiana inoculation (Figure 6c), while CmWRKY7, 9 and 12 responded positively to aphid infestation (Figure 6d). The expression changes of other CmWRKY genes were less than two fold in magnitude.

Comparison of the Expression Pattern between CmWRKY Genes and Correlated Arabidopsis Homologs
Accumulating evidence suggests that many WRKY genes are involved in the regulation of plant development and their stress response [3,21]. Here, it was clear that the CmWRKY genes differed from one another with respect to their tissue specificity and inducibility. The various stress treatments affected the transcription level of different combinations of the 15 CmWRKY genes, implying that most of them do contribute to the stress responses of chrysanthemum. The A. thaliana genes AtWRKY6 [22], 40 [23], 41 [14] and 70 [24] all participate in the defense against pathogen attack, and their homologs (as determined by BLAST) CmWRKY1, 8,11 and 15 were similarly induced by pathogen inoculation. On the other hand, AtWRKY31 (the homolog of which, based on phylogeny, is CmWRKY1) has no known involvement in the disease response. AtWRKY17 is up-regulated by salinity stress [25], as is its homolog CmWRKY12. AtWRKY11, 17 and 70, all of which are induced by wounding, drought and low temperature [26][27][28], also play a role in the interaction between SA and JA [29,30]. Their orthologs in chrysanthemum (respectively, CmWRKY3, 12 and 11) were also induced by moisture stress (PEG treatment) and low temperature, but were down-regulated by wounding; and in converse response to exogenous SA and JA. AtWRKY70 is activated by SA but suppressed by JA [29]; its constitutive expression improves the level of resistance to biotrophic, but not to necrotrophic fungi [17,31]. Its chrysanthemum homolog CmWRKY11 was inducible by inoculation with necrotrophic fungi and the exogenous supply of SA, but suppressed by biotrophic fungi and JA, which is rather different from the behavior of AtWRKY70. AtWRKY40 is an important component of WRKY-mediated ABA signaling, and exogenous ABA treatment reduces both its transcription and translation [32]. CmWRKY15, its chrysanthemum homolog, was notably up-regulated by exogenous ABA, which again suggests that some pairs of WRKY homologs have evolved different functionality between A. thaliana and chrysanthemum. The function of AtWRKY13, 21 and 32 is not well understood, so little can be concluded by comparing these with their chrysanthemum homologs, respectively, CmWRKY3, 12 and 11.

Plant Materials
Cuttings of the cut flower chrysanthemum cultivar "Jinba", maintained by the Chrysanthemum Germplasm Resource Preserving Center (Nanjing Agricultural University, Nanjing, China), were rooted in vermiculite in the absence of fertilizer in a greenhouse. After 14 days, they were transplanted into growth substrate, in preparation for exposure to a range of stress and phytohormone treatments.

Isolation and Sequencing of Full-Length CmWRKY cDNAs
Total RNA was isolated from leaves, stems, roots or florets using the RNAiso reagent (TaKaRa, Tokyo, Japan), following the manufacturer's instructions. The first cDNA strand was synthesized from 1 μg total RNA using M-MLV (Moloney murine leukemia virus) reverse transcriptase (TaKaRa), according to the manufacturer's instructions. Primer pairs (listed in Table S1) were designed to amplify internal CmWRKY fragments, based on sequences identified in a chrysanthemum transcriptome database (unpublished data). The sequences of the resulting amplicons were used to derive full length cDNAs via 5'-and 3'-RACE PCR. For the 3' reaction, the first cDNA strand was synthesized using the dT adaptor primer dT-AP, and this was followed by a nested PCR based on the primer pair CmWRKYx-3-F1/F2 and the adaptor primer AP (Table S2). For the 5' reaction, the primers consisted of AAP and AUAP (provided with the 5' RACE System kit v2.0, Invitrogen, Carlsbad, CA, USA), along with the gene-specific primer pairs CmWRKYx-5-F1/F2 (Table S3). Where required, amplicons were purified using an AxyPrep DNA Gel Extraction kit (Axygen, Hangzhou, China) and cloned into pMD19-T (TaKaRa) for sequencing. Finally, 15 pairs of gene-specific primers (CmWRKYx-ORF-F/R, Table S4) were elaborated to amplify the full open reading frame sequences.

Phylogeny of WRKY Sequences
A. thaliana WRKY sequences were downloaded from the Database of Arabidopsis Transcription Factors (DATF) [33], and combined with the newly acquired CmWRKY sequences to perform a multiple alignment analysis based on ClustalW software [34]. The subsequent phylogenetic analysis utilized the Neighbor-Joining method, and a graphical representation was produced with the help of MEGA5 software [35]. Internal branching support was estimated using 1000 bootstrap replicates. The MEME v4.8.1 program [19] served to identify the motifs present in the 15 CmWRKY proteins, using the parameter settings suggested by Huang [21], and retaining only motifs associated with an E value <e −5 .

Plant Treatments
The tissue-specific and treatment-induced transcription profiles of the 15 CmWRKY genes was explored in young seedling roots, stems and leaves, and in the tube and ray florets of inflorescences at the bud stage. A variety of abiotic stresses was imposed, namely salinity (200 mM NaCl), moisture deficit (20% w/v polyethylene glycol (PEG6000)) [36], low temperature (4 °C), high temperature (40 °C) and wounding. Plants were also inoculated separately with three different fungi, namely the necrotroph Alternaria tenuissima, the biotroph Puccinia horiana and the hemibiotroph Fusarium oxysporum; further plants were infested with the aphid Macrosiphoniella sanbourni.
For the NaCl and PEG6000 assays, young plants were transferred to a liquid medium containing the stress agent, and the roots were sampled at various time points. Other seedlings were exposed to a period of either 4 or 40 °C in a chamber delivering a 16 h photoperiod and 50 μmol·m −2 ·s −1 light, after which the second true leaves were sampled [37]. The wounding treatment involved cutting the second true leaf. The phytohormone treatments involved spraying the leaves with either 50 μM abscisic acid (ABA) [38], 1 mM methyl JA (MeJA) [39] or 200 μM SA [40]. Plants were sampled prior to the stress treatment and then after one, four, eight, 12 and 24 h.
Spore suspensions of A. tenuissima and F. oxysporum were obtained from a 14 day old PDA (potato dextrose agar) cultures, to which 10 mL of sterile water had been added; the spore suspension was then filtered through four layers of cheesecloth, and the spore concentration adjusted to 10 7 per mL using a haemocytometer [41,42]. Leaves were sampled before inoculation with A. tenuissima and then after six, 24, 48 and 72 h; the roots were sampled prior to F. oxysporum inoculation, and then after four, 12, 24 and 48 h. Infection with P. horiana was achieved using the spray method described by Zandvoort et al. [43] with the following modifications: Briefly, the infected leaves containing white rust pustules were cut into small pieces and were dispersed in deionized water and filtered through medical gauze to remove any plant debris. The concentration of the pathogenic spore suspension was then adjusted using a hemacytometer slide to a concentration of 10 6 zoosporangia·mL −1 with deionized water containing one drop of Tween 20 before application to the plants until run-off, using a hand-held sprayer. Leaves were sampled before inoculation and then after six, 12, 24, 48 and 72 h. The method used for aphid infestation followed [44] with the following minor modifications: Briefly, twenty aphids in two instar nymphs were transferred to the plants by a soft brush. A 25 cm long × 12 cm diameter polyester cylinder capped with fine gauze was placed over each plant to prevent any movement of aphids between adjacent plants. Leaves were sampled prior to infestation, and then after three, six, 12, 24 and 48 h.
After sampling, all the material was snap frozen in liquid nitrogen and stored at −70 °C. Each treatment was replicated three times.

Data Analysis
The relative transcription levels of each CmWRKY were log 2 transformed, and the profiles compared using Cluster v3.0 software [47] and visualized using Treeview [48]. SPSS v17.0 software (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses.

Conclusions
The comparative analysis of functionality in chrysanthemum and A. thaliana suggests that the local alignment method is superior to phylogenetic analysis in predicting functional homology. The present study has documented the transcription behavior of 15 CmWRKY genes in response to a range of stress treatments. These data provide a basis for identifying which individual CmWRKY genes might be usefully targeted for improving the stress tolerance of chrysanthemum.