WRKY40 , WRKY70 , and Downy Mildew Resistant 6 (DMR6)-Like Oxygenase 1 are universal marker genes for the salicylic acid pathway in bananas

Background: Banana, an important cash and staple crop worldwide, suffers from various biotrophic and In plants’ defense against these pathogens, the phytohormone salicylic acid (SA) plays a key role in the regulation of immune response. Using a specific set of SA-responsive genes as markers is frequently adopted to monitor the onset of SA-mediated immune response. However, reliable SA-responsive genes marker genes have not been well established in bananas. Results: From the transcriptome analysis of SA-treated ‘Pei-Chiao’ banana roots, we identified 19 up-regulated and 3 down-regulated genes. Four of the up-regulated genes previously reported to play crucial roles in SA-mediated immunity in other species were further analyzed for their applicability in different tissues and cultivars of bananas using real-time quantitative reverse-transcription PCR. The analysis showed that WRKY40 , WRKY70 , and Downy Mildew Resistant 6 (DMR6)-Like Oxygenase 1 ( DLO1 ) were significantly induced upon SA treatment in both the leaves and roots of ‘Pei-Chiao’ (AAA genome), ‘Pisang Awak’ (ABB genome), and ‘Lady Finger’ (AA genome) bananas. Conclusions: The uncovering of common marker genes WRKY40 , WRKY70 , and DLO1 for SA response in different banana genome types provides the stepping stone for studies towards understanding of SA-mediated immune response in bananas.

vulnerable to mass destruction in face of growing adverse abiotic and biotic stressors [3,6].
As sessile organisms, plants are prone to pathogens, insects, and changing environment without the ability to move away from incoming stresses. Thus, plants have evolved repertoire of mechanisms tuned by phytohormones in response to stresses [7]. Among the various identified phytohormones, the phenolic compound salicylic acid (SA) is critical for plant defense against a broad spectrum of pathogens. It is the major defense hormone in the interactions with biotrophic and hemi-biotrophic pathogens [8][9][10][11][12].
The plants' immune response can be stimulated from the recognition of the invading pathogen, triggering the increase in SA concentration that is both required and sufficient to activate plant defense [10,11,13]. The basal resistance involves the perception of conserved pathogen molecular signatures, pathogen-or microbe-associated molecular patterns (i.e. PAMPs or MAMPs), which activates pattern-triggered immunity (PTI) [14]. However, plant pathogens have evolved the ability to deliver counteractive effector proteins that suppresses PTI [15][16][17][18]. In turn, plants have evolved disease resistance (R) proteins, which directly or indirectly detects pathogen effectors, and activate effector-triggered immunity (ETI) [19]. Activation of both PTI and ETI is associated with increase in SA accumulation, which is important for resistance against pathogens [20][21][22]. In addition to the activation of PTI and ETI upon primary pathogen infection, plants have evolved systemic acquired resistance (SAR) that also requires SA to confer resistance in the uninfected systemic (distal) organs in response to prior (primary) infection [23][24][25]. Since endogenous SA accumulation can trigger defense reprograming and activate a specific set of genes in response to pathogen attack [26], using these SA-modulated genes as markers to monitor the onset of SA-mediated immune response is commonly adopted in the study of plant immunity [27,28].
Banana suffers from various biotrophic and hemi-biotrophic pathogens including viruses [29] and the important soil-borne disease caused by Fusarium oxysporium f. sp. cubense [3,4]. In order to expedite the research on the important SA-signaling in banana, identification of the SA marker genes with expression levels highly correlated with endogenous SA concentration will greatly help to the understanding of the banana immune response and banana-pathogen interaction in detail. In the dicotyledonous model plant Arabidopsis, the NON-EXPRESSOR OF PATHOGENESIS-RELATED GENES 1 (NPR1) has been identified as a central regulator in SA-mediated plant defense [30,31].
Upon SA treatment or pathogen infection, NPR1 translocates to the nucleus where it interacts with TGA transcription factors, to coordinate the transcriptional activation of downstream immune responsive genes such as genes encoded PATHOGENESIS-RELATED proteins (PRs). Among PRs, PR-1 gene has been used as a canonical marker to monitor the onset of SA-mediated pathway [32,33]. In bananas, a gene closely similar to Zea mays PR-1 (GenBank accession no. AAC25629), and two NPR1 homologs, MNPR1A (GenBank accession no. DQ925843), and MNPR1B (GenBank accession no. EF137717) have been shown to be induced by exogenous SA treatment or pathogen challenge [34,35]. However, a recent transcriptome analysis with treatment of a synthetic SA analogue, benzothiadiazole (BTH), identified the induction of PR-1 homolog (GenBank accession no. XM_009419475.2) [36] instead of the PR-1 (AAC25629.1) [34]. Moreover, NPR1 homolog was not recovered from the analysis of genes induced by the BTH treatment [36]. The discrepancy among the previously reported SA marker genes could be caused by the differences in growth conditions, stages plants for treatment, the SA treatment methods, the banana cultivars, and sampling time for expression profiling.
In this report, the transcriptome analysis recovered 3 reliable SA-responsive marker genes, MaWRKY40, MaWRYK70, and Downy Mildew Resistant 6-Like Oxygenase 1 (MaDLO1) from 'Pei-Chiao' (a popular Cavendish banana cultivar with AAA genome grown in Taiwan). The expression level of all three genes was highly correlated with the concentration of SA in both 'Pei-Chiao' roots and leaves.
Moreover, all of the 3 SA-responsive marker genes were also robustly induced in the roots and leaves of banana cultivars belonging to the genomic groups other than AAA genome, namely 'Pisang Awak' (ABB genome) and 'Lady Finger' (AA genome). The core SA responsive genes uncovered in this study can serve as the stepping stone for the studies of SA-mediated signaling in banana.

Results
Test of previously reported SA-responsive genes in leaves of 'Pei-Chiao' plantlets In order to identify the salicylic acid (SA) marker genes in banana that would enable the analysis of SA-induced signaling in 'Pei-Chiao', we first analyzed genes that had been previously reported to be SA-responsive (Table 1). Banana plantlets grown in potted soil were treated with buffer (Mock) or SA by foliar spray. Leaves of each treated plant were harvested at two time points, 0 (no-treatment) and 6 hour-post treatment (hpt) for SA concentration and gene expression analyses. We measured SA concentration by high-performance liquid chromatography-mass spectrometry, which revealed a significant increase in SA concentration in SA-treated plants compared to the mock or no-treatment control at 6 hpt (Fig. 1a). We next examined the transcriptional responses of the known SA-responsive genes listed in the Table 1 with real-time reverse transcriptase quantitative PCR (RT-qPCR). Among them, only MaNPR1D showed significant and reproducible induction at 6 hpt of SA (Fig. 1b). Compared to mock treatment, MaNPR1D increased by 2-fold at 6 hpt (Fig. 1b). This gene was selected for further analysis.
Analysis of SA concentration and MaNPR1D expression in leaves and roots of 'Pei-Chiao' plantlet after

SA treatment
Roots serve as the initial infection site for several hemi-biotrophic pathogen such as Fusarium oxysporium f. sp. cubense, Xanthomonas spp., and Ralstonia spp. [4,37]. To analyze whether MaNPR1D could serve as a reliable SA marker in roots of 'Pei-Chiao', we first optimized our SA treatment. To this end, we used 'Pei-Chiao' tissue culture plantlets grown on sterilize agar medium to prevent the potential infection of soil microorganism and minimize the damage during root sampling.
'Pei-Chiao' tissue culture plantlets grown in agar were treated with buffer (mock) or SA by foliar spray.
Since we do not know whether SA level is increased under the treatment method of foliar spray, we first quantified the SA concentration in mock-and SA-treated leaves and roots by the use of ultraperformance liquid chromatography chromatography (HPLC)-mass spectrometer (MS)/MS. (Fig. 2).
Quantification of SA concentration revealed the increased levels of SA in the SA-treated 'Pei-Chiao' leaves and roots harvested at 6 hpt ( Fig. 2a and b). Statistical significance was observed for the SA concentration in leaves of SA-treated plantlets at 6 hpt compared to that of no-treatment control or mock-treated plantlets (Fig. 2a); no statistical significance was found between no-treatment (0 hpt) and mock-treated plantlets (6 hpt) (Fig. 2a). In roots, the average SA concentration was also significantly higher in 6 h post SA-treated plantlets compared to no-treatment control plantlets or mock-treated plantlets (Fig. 2b). No statistical significance was found between no-treatment (0 hpt) and mock-treated plantlets (6 hpt) (Fig. 2b).
As expected, real-time RT-qPCR confirmed that MaNPR1D expression positively correlated with higher levels of SA in the leaves of 'Pei-Chiao' in repeated experiments (Fig. 2c). To identify differentially expressed genes under SA treatment, we first mapped the reads to the banana genome database (Musa acuminata DH Pahang v2; banana-genome-hub.southgreen.fr) after adaptor trimming and quality filtering. The mapping rates of the processed reads among the samples for RNA-seq analysis were similar and ranged from 83.29%-89.53% (Table 2). We then analyzed differentially expressed genes (DEGs) under SA treatment in 'Pei-Chiao' using statistic test with adjusted P value < 0.1 and obtained a total of 22 DEGs (Table 3) (Table 3 and Fig. 3). These 4 genes were selected for further analysis by real-time RT-qPCR (Table 3 Table 3 List of differentially expressed genes in 'Pei-Chiao' plantlet at 6 hour post-salicylic acid-treatment.  . 4) with the use of gene specific primers (Table 4). In the roots, all 4 genes exhibited a higher expression level in the SA treatment compared to the mock treatment at 6 hpt, MaPR1-like increased by 2-fold, MaWRKY40 increased by 56-fold, MaWRKY70 increased by 31-fold, and MaDLO1 increased by 487fold; however, the induced expression of MaPR1-like expression did not reach a statistical significance (Fig. 4a). Consistent with the RT-qPCR results of the roots, MaWRKY40, MaWRYK70, and MaDLO1 mRNA levels in the leaves were significantly higher by 92-fold, 34-fold, and 1332-fold, respectively, at 6 hour post-SA-treatment compared to mock-treatment (Fig. 4b). No obvious difference was observed for MaPR1like expression after the SA treatment in roots (Fig. 4b).
MaWRKY40, MaWRYK70, and MaDLO1 are induced by SA treatment in all tested cultivars To determine whether MaWRKY40, MaWRKY70 and MaDLO1 could represent a core set of SAresponsive genes across bananas of various genomic groups, we further compared the gene expression profiles of these genes in between mock-and SA-treated samples of 'Pisang Awak' (ABB genome) and 'Lady Finger' (AA genome) by real-time RT-qPCR at 6 hpt ( Fig. 5-6).

Discussion
We initially aimed to use reported SA marker gene for monitoring the SA-mediated immune response in our selected banana cultivar; however, only MaNPR1D are consistently induced by SA treatment in our analysis.  [46,47]. Therefore, it is not common to use NPR1 as a marker for monitoring the onset SA-mediated immune response.
In contrast to NPR1, the induction of PR1 genes has been reported to be 500-1000 fold higher at 6 hour post-SA treatment in Arabidopsis [48], thus PR1 is commonly considered as a reliable SA marker gene. However, previously report indicates the discrepancy of PR1 homolog as SA marker genes in different banana cultivars [34,36]. Moreover, MaPR1-like (NCBI accession: XM_009400035.2) that was identified through transcriptome analysis (Table 3) in 'Pei-Chiao' was only mildly induced in the roots but not leaves (Fig. 5).  [50,51]. These results suggest differences in PR1 gene as SA markers between monocots and dicots.
From our study, we identified 2 WRKY transcription factors, MaWRKY40 and MaWRYK70 that robustly corresponded to elevated SA concentration in banana leaves and roots (Fig. 5-7). WRKY transcription factors have been reported to be induced by SA treatment in both dicotyledonous and monocotylendous plants [52,53].
It was reported that the induction of WRKY70 transcripts occurred earlier and accumulated to higher levels upon pathogen infection or SA treatment in Arabidopsis. In addition, overexpression of WRKY70 resulted in the constitutive expression of SA-responsive PR genes and enhanced resistance to the biotrophic and hemi-biotrophic pathogens, while repressing jasmonic acid (JA) response [54,55]. Moreover, in the previous study of Arabidopsis [54], WRKY40 is also induced by SA and suppresses the expression of ABA-responsive genes ABI4 and ABI5 [39,41,56,57]. Hence, WRKY70 and WRKY40 function as a node of convergence for integrating signals from SA and JA-dependent pathways or SA-and ABA-dependent pathways, respectively.
The most sensitive SA-responsive gene observed in this study was DLO1, which showed strongest induction (> 100 folds) in different cultivars (Fig. 5-7). In Arabidopsis, DLO1 is strongly activated during hemi-biotrophic Hyaloperonospora arabidopsidis attack and following BTH treatment [42]. Studies showed that DLO1 catabolizes SA by converting SA to 2,3-dihydroxybenzoic acid (2,3-DHBA) and is a component of a negative feedback regulation system of SA levels during senescence [58]. Similar hydroxylation mechanisms to control levels of other plant hormones jasmonate and auxin have also been observed [59,60]. Since the induction of DLO1 in the negative feedback regulation of cytoplasmic SA levels may be critical to maintain the homeostasis of SA concentration. Our data indicates that DLO1 serve as a suitable marker gene to monitor the fine tuning of SA signaling.

Conclusion
In conclusion, we conducted a comprehensive analysis of SA-responsive genes in bananas and identified To extract SA, we used a method following previously described [61]. The LC system used for analysis was ultra-performance liquid chromatography (UPLC) system (ACQUITY UPLC, Waters, Millford, MA