Characterization of wall-associated kinase/wall-associated kinase-like (WAK/WAKL) family in rose (Rosa chinensis) reveals the role of RcWAK4 in Botrytis resistance

Wall-associated kinase (WAK)/WAK-like (WAKL) is one of the subfamily of receptor like kinases (RLK). Although previous studies reported that WAK/WAKL played an important role in plant cell elongation, response to biotic and abiotic stresses, there are no systematic studies on RcWAK/RcWAKL in rose. In this study, we identified a total of 68 RcWAK/RcWAKL gene family members within rose (Rosa chinensis) genome. The RcWAKs contained the extracellular galacturonan-binding domain and calcium-binding epidermal growth factor (EGF)-like domain, as well as an intracellular kinase domains. The RcWAKLs are missing either calcium-binding EGF-like domain or the galacturonan-binding domain in their extracellular region. The phylogenetic analysis showed the RcWAK/RcWAKL gene family has been divided into five groups, and these RcWAK/RcWAKL genes were unevenly distributed on the 7 chromosomes of rose. 12 of RcWAK/RcWAKL genes were significantly up-regulated by Botrytis cinerea-inoculated rose petals, where RcWAK4 was the most strongly expressed. Virus induced gene silencing of RcWAK4 increased the rose petal sensitivity to B. cinerea. The results indicated RcWAK4 is involved in the resistance of rose petal against B. cinerea. Our study provides useful information to further investigate the function of the RcWAK/RcWAKL gene family and breeding research for resistance to B. cinerea in rose.

structure, contacting extracellular and cytoplasmic signals. Their extracellular structure contains a galacturonan-binding domain and/or a calcium-binding epidermal growth factor (EGF)-like domain for signal perception [6]. WAKs/WAKLs play important roles in regulating plant growth, development, and responding to environmental stresses. Expression of the AtWAK4 antisense gene in Arabidopsis results in impaired cell elongation and stunted lateral root development [7]. In addition, knocked out CaWAKL20 significantly increased pepper heat tolerance, whereas overexpression of CaWAKL20 in Arabidopsis reduced heat tolerance, indicating that CaWAKL20 negatively regulated pepper heat stress response [8].
Meanwhile, WAKs/WAKLs play a vital role in plant defense response to fungal diseases. Oligogalactouronides (OGs), as the cell wall polysaccharide degradation productions, are recognized by AtWAK1. And overexpression AtWAK1 could enhance resistance to B. cinerea in Arabidopsis [9,10]. The high expression of ZmWAK, which was located in the head smut quantitative resistance locus (qHSP1), effectively suppressed the growth of Sporisorium reilianum in the mesocotyl [11]. Rice OsWAK14, OsWAK91 and OsWAK92 positively regulated the defense response to blast fungus (Magnaporthe oryzae), while OsWAK112d worked as a negative regulator of the defense response [12]. SlWAK1 is involved in the regulation of PRR-mediated immune response through FLS2/FLS3 complex in tomatoes [13].
Rose (Rosa sp.), as an important ornamental crop, accounts for more than 30% of the world's annual trade in cut flowers [14]. The long-distance transportation of cut rose flowers leads to water loss, senescence and postharvest diseases on cut flowers, resulting in serious economic losses. Gray mold disease, caused by B. cinerea, as one of the most serious postharvest fungal diseases of rose. The role WAKs/WAKLs family members in resistance against necrotrophic pathogens have been reported in Arabidopsis [9,10]. However, whether WAKs/WAKLs are involved in rose resistance to B. cinerea is largely unknown.
Our previous study found that a large number of rose WAK/WAKL genes significantly up-regulated in response to B. cinerea inoculation [15]. In this study, we comprehensively characterized the of RcWAKs/RcWAKLs in rose genome, including their gene structural features, conserved domain, evolutionary relationships, and their expression pattern upon B. cinerea inoculation. This analysis suggested RcWAK4 may play a role in the response to B. cinerea infection. We finally used virus-induced gene silencing (VIGS) to confirm that RcWAK4 is involved in the rose defense against B. cinerea.

Identification of RcWAK/ RcWAKL family members in rose
To identify the WAK/WAKL gene family in Rosa chinensis, 27 1). Finally, we have verified a total of 68 non-redundant RcWAK/RcWAKL family genes in rose genome, where 23 candidate genes contained EGF_CA, GUB_WAK_bind and Phkinase_Tyr conserved domains, and considered as RcWAK genes. The other candidate genes were RcWAKLs, of which 38 contained GUB_WAK_bind and Phkinase_Tyr domain, the remaining 7 contained EGF_CA and Pkinase_Tyr domain (Fig. 1).
The length of RcWAKs and RcWAKLs was extremely varied. The shortest is RcWAKL41, which encodes 369 amino acids. The longest is RcWAKL11, it encodes a protein of 891 amino acids. The average length of RcWAK/ RcWAKL proteins was 678 amino acids. The accession number, chromosomal location, numbers of exon and intron, length of CDS and subgroup of RcWAK/RcWAKL gene family were listed in Table 1.

Chromosomal locations of rose RcWAK/RcWAKL genes
RcWAKs/RcWAKLs were mapped to seven chromosomes of the R. chinensis genome using Mapchat2.2. RcWAK1-RcWAK23 and RcWAKL1-RcWAKL45 were named according to their order on the chromosomes (Fig. 2). RcWAK/RcWAKL family genes were unevenly distributed on the seven chromosomes of rose (Table 1; Fig. 2). The high density of RcWAKs/RcWAKLs location was observed in several specific regions, such as six RcWAKLs (RcWAKL5-RcWAKL10) distributed at 25.34-25.74 Mb on chromosome 1 and fifteen RcWAKs/RcWAKLs distributed at 11.37-11.98 Mb on chromosome 5. In contrast, chromosomes 3, 4 and 6 contained only 3, 3 and 4 of RcWAKs/RcWAKLs, respectively. RcWAK members were located on chromosomes 1, 2, 5, 6 and 7, while RcWAKLs were located on all seven chromosomes. The unbalanced distribution of RcWAK/RcWAKL genes indicated genetic variation during the evolutionary process.

Phylogenetic analysis and structure analysis of rose RcWAK/RcWAKL genes
A phylogenetic analysis of RcWAKs/RcWAKLs was performed by using neighbor-joining method (Fig. 3). The subsequent intron-exon structure analysis and conserved domain analysis of RcWAKs/RcWAKLs were consistent with the results of the phylogenetic analysis (Fig. 3). RcWAKs/RcWAKLs contained mostly 2-3 introns, while RcWAKL20 and RcWAKL24 had no introns and RcWAKL11 contained 11 introns. Most of the genes in the same evolutionary branch exhibited similar exon-intron structures, such as RcWAKL21, RcWAKL35, RcWAK15, RcWAK3, RcWAK21, RcWAK1 and RcWAKL32 all contained 2 introns (Table 1; Fig. 3). However, there were a few exceptions, for example, RcWAKL43, RcWAKL42, RcWAKL37 and RcWAK24. More than that, the intron lengths of RcWAK/RcWAKL family members were spanning a wide range, from tens to tens of thousands of nucleotides. RcWAKL45 contained the longest intron (12,233bp), while RcWAKL38 had the shortest intron (59bp).
In addition, we further analyzed the conserved domains of RcWAK/RcWAKL protein sequences by conserved domains database (CDD), and the results were similar to the previous results with Pfam database. All RcWAKs contained the highly conserved EGF_CA, GUB_WAK_bind and Pkinase, while RcWAKLs lacked the EGF_CA or GUB_WAK_bind domain. Meanwhile, the transmembrane (TM) structures and the signaling peptides (SP) were determined based on TMHMM Server (http:// www. cbs. dtu. dk/ servi ces/ TMHMM/) and SignalP (http:// www. cbs. dtu. dk/ servi ces/ Signa lP/). Several RcWAKs/RcWAKLs were missing TM or SP in structure prediction. RcWAKL4 and RcWAKL23 missed both two structures (Fig. 3). Overall, RcWAKs/RcWAKLs are highly conserved in amino acid sequences from phylogenetic analysis, although some of RcWAKs/RcWAKLs lack one or two structures. RcWAKs/RcWAKLs from different evolutionary branches exist in diversity.
In recent years, an increasing number of studies have confirmed that WAK/WAKL family genes played a critical role in defense response. We compiled a total of 15 plants defense response-related WAK/WAKL family genes, including Arabidopsis, rice (Oryza sativa), tomato (Solanum lycopersicum), cotton (Gossypium hirsutum), maize (Zea mays L.), and wheat (Triticum aestivum) (Supplemental Table S2). To evaluate the evolutionary relationship between RcWAK/RcWAKL and the defense-related WAK/WAKL genes reported in different species, a phylogenetic tree was established using the neighbor-joining method (Fig. 4). The results showed that the proteins of the WAK/WAKL family were divided into five groups, which were labeled with different colors in Fig. 4. The groups II, III, IV and V contained WAK/WAKL from other species which were involved in plant resistance. We have also performed a     Fig. S3). However, the group V consisting of RcWAKLs was lack of EGF_CA logo, which was constituted with six cysteine (C) skeleton (Supplemental Fig. S2).

Expression patterns of RcWAK/RcWAKL genes in rose petals induced by B. cinerea
Studies have shown that WAK/WAKL family members played a vital role in plant defense response to fungal pathogens. The WAKs/WAKLs which were up-regulated upon pathogen infection may involve in plant immunity. Our previous study established an RNAseq analysis of B. cinerea inoculated rose petals at 30 hpi corresponds to a later response, with the primary lesions starting to expand [15]. Twelve of RcWAK/RcWAKL genes were significantly up-regulated expression in RNA-seq data ( Table 2). Among them, RcWAK2, RcWAK4, RcWAK8, RcWAK14, RcWAK16, RcWAKL6, RcWAKL22 and RcWAKL43 were induced in rose petals at 30 hpi, while RcWAK22, RcWAKL12, RcWAKL23, and RcWAKL43 were also induced at 48 hpi. To further clarify the expression profiles from the transcriptome data, RcWAK2, RcWAK4, RcWAK22, RcWAKL12, RcWAKL22 and RcWAKL43 expression patterns were confirmed by RT-qPCR (Fig. 5).

RcWAK4 participated in rose resistance to B. cinerea
To further illustrate the potential role of RcWAK/ RcWAKL in rose petals against gray mold, we knocked down the expression of RcWAK4 using VIGS. RcWAK4 was the most significantly up-regulated RcWAKs in RNA-seq data ( Table 2) and our qPCR also confirmed RcWAK4 was the most strongly induced RcWAKs upon B. cinerea inoculation (Fig. 5). Meanwhile, RcWAK4 belonged to the group III of RcWAKs/RcWAKLs and was close to AtWAKL10 and AtWAKL22, which were known to involve in plant disease resistance. Thus, RcWAK4 was considered an important candidate gene for rose resistance to B. cinerea. To clarify whether RcWAK4 was involved in rose petal defense response, we knocked down the expression of RcWAK4 in petals. To this end, we cloned the coding sequence fragment 309 bp of RcWAK4 into the TRV2 vector to generate TRV2-RcWAK4. Mixed agrobacterium cells carried TRV2-RcWAK4 and TRV1 in a 1:1 ratio and then vacuum infiltrated into rose petal discs for RcWAK4 silencing. The silenced petals were subsequently inoculated with B. cinerea. The RcWAK4-silencing petals exhibited a significant increase in lesion area compared   Table S1. PDB, potato dextrose broth. 'T' represented the standard deviation with control petals inoculated with an empty TRV vector (TRV-00) ( Fig. 6a and b). Finally, the silencing efficiency was confirmed by qPCR (Fig. 6c). These results suggested that RcWAK4 played an important role in rose petal resistance to B. cinerea.

Discussion
Plant resistance responses are beginning from the PRRs located on cell membrane recognizing to pathogenic signals, such as microbial-associated molecular patterns/damage-associated molecular patterns (MAMPs/ DAMPs) [16,17]. WAK/WAKL is a vital class of PRRs for necrotrophic pathogens (e.g., B. cinerea). Genome-wide identification of WAKs/WAKLs has been successively reported in Arabidopsis [18], rice (Oryza sativa) [19], cotton (Gossypium hirsutum) [20] and barley (Hordeum vulgare) [5]. However, there is no systematic analysis of the RcWAK/RcWAKL gene family, and the functions of the RcWAKs/RcWAKLs were largely unclear. In this study, we used 'R. chinensis' genome as a reference to perform a genome-wide analysis of RcWAK/RcWAKL gene family. We predicted the potential function of the RcWAK/ RcWAKL gene family in rose through chromosome localization, exon-intron structural analysis, conserved domain, phylogenetic analysis, and gene expression pattern under B. cinerea infection. The rose RcWAK/RcWAKL family contained 68 members (Table 1), while in Arabidopsis, rice, and barley were identified 26, 125, and 91 WAK/WAKL family members, respectively. The large disparity in the number of WAK/WAKL family members among different species may indicate that WAK/WAKL genes underwent different degrees of expansion during the evolution of plants. However, we have identified a total of 23 RcWAKs, they have a highly conservative domain with AtWAK1-AtWAK5 [18] and GhWAK1-GhWAK29 [20], including signaling peptide, EGF_CA, transmembrane and Pkinase (Fig. 3).
In phylogenetic analysis (Fig. 4), RcWAKs/RcWAKLs were divided into five groups. Evolutionary relationship analysis of barley also divided 91 HvWAK/HvWAKL into five groups [5]. The RcWAKL11, which was located in the group V, was evolutionarily distant from the other members in the same group (RcWAKL42, RcWAKL43, RcWAKL37, RcWAKL24, RcWAKL20 and RcWAKL12), probably due to more exons in the RcWAKL11 coding sequence. RcWAK members were distributed on chromosomes 1, 2, 5, 6 and 7, and evolutionary analysis showed that none of the RcWAK members were divided in group V (Figs. 2 and 4). Nevertheless, RcWAKLs were distributed on all seven chromosomes, but none of RcWAKLs appeared in group III in the evolutionary analysis. All four groups contained plant defenses associated WAKs/WAKLs except group I. While twelve of RcWAKs/ RcWAKLs significantly up-regulated in the RNA-seq data ( Table 2) induced by B. cinerea were distributed in all groups. The results implied that there was no group specificity on the role of WAK/WAKL in plant defense responses. The expression pattern of RcWAKs/RcWAKLs induced by B. cinerea provided abundant candidate genes for its possible involvement in rose defense response to gray mold. The differential expression of RcWAK4 was more than 7-fold at the early stage after B. cinerea infection ( Table 2), and the qPCR results were also consistent (Fig. 5). Transient silencing of RcWAK4 in rose petals exhibited increased susceptibility to B. cinerea (Fig. 6). In phylogenetic analysis, AtWAKL10 and AtWAKL22 were in the same group with RcWAK4 (Fig. 4). AtWAKL10 as a candidate guanylyl cyclase (GC), involved in Arabidopsis defense response [21]. Together, these results suggested that RcWAK4 as an important RcWAKs/RcWAKLs involved in the response of rose against B. cinerea.

Conclusions
A genome-wide characterization of RcWAK/RcWAKL family was performed in this study, mainly including chromosome localization, gene structure analysis, phylogenetic relationships, gene expression analysis induced by B. cinerea. We have identified a total of 68 non-redundant RcWAK/RcWAKL family members in the whole genome of R. chinensis. RcWAK4 was confirmed to be involved in rose petal resistance to gray mold by VIGS. These results provide a piece of essential information for further study RcWAK/RcWAKL family members in rose and may facilitate the further breeding of disease resistance roses.

Gene structure and phylogenetic analyses
Extracted coding sequences region of RcWAK/RcWAKL genes, and the gene intron-exon structure was mapped using TBtools [24] with the genome-wide gene information file as reference. The conserved domain information of RcWAK/RcWAKL genes were obtained from Pfam database and Conserved domain database (CDD, https:// www. ncbi. nlm. nih. gov/ Struc ture/ cdd/ wrpsb. cgi).
Full sequence alignment of RcWAK/RcWAKL was performed using the ClustalW method. Phylogenetic analysis of the comparison results was constructed using Neighbor-Joining of MEGA6.0 [25] with a bootstrap of 1000 replicates.
The comparative analysis of conserved domains between different groups of RcWAKs/RcWAKLs. The group menbers of RcWAKs/RcWAKLs were aligned by ClustalW before performing sequences logo on WebLogo3 (http:// weblo go. three pluso ne. com/).

Expression data of RcWAKs/RcWAKLs in rose petal induced by B. cinerea
The RNA-seq data of rose petals infected by B. cinerea were obtained from the National Center for Biotechnology Information (NCBI) database with the accession number PRJNA414570 [15]. There no special permissions are necessary to collect these data from NCBI. Clean data were mapped to 'Rosa chinensis' genome, and reads per kb per million reads (RPKM) were used to value gene expression levels, and log2 (RPKM treatment/ RPKM control) was used to calculate the differential expression levels of genes.
The real time quantitative PCR were performed to confirm the RNA-seq results. To this end, cDNA synthesis were according to the RT Super Mix kit (Vazyme). qPCR was performed with TB Green Premix qRT-PCR kit (TAKARA) based on One Plus real-time PCR system (Applied Biosystems). RcUBI2 was used as the housekeeping gene, and the primers used in the experiment were shown in Supplemental Table S1.