Cytokinin oxidase gene CKX5 is modulated in the immunity of Arabidopsis to Botrytis cinerea

In our previous work, cytokinin (CK) signaling and biosynthesis were found to be modulated during Arabidopsis defense against infection by the necrotrophic pathogen Botrytis cinerea. Notably, the expression level of CYTOKININ OXIDASE/DEHYDROGENASE 5 (CKX5) was significantly induced in B. cinerea-infected leaves and later in distant B. cinerea-untreated leaves of the same plant. To confirm and determine how CKX5 is involved in the response to B. cinerea infection, transcript levels of CKX family genes were analyzed in B. cinerea-inoculated leaves, and only CKX5 was remarkably induced by B. cinerea infection. Furthermore, CKX5-overexpressing Arabidopsis plants were more resistant to B. cinerea than wild-type plants. Transcription factors (TFs) binding to the CKX5 promoter were then screened by yeast one-hybrid assays. Quantitative Real-Time Reverse Transcription PCR (qRT-PCR) analysis further showed that genes encoding TFs, including WRKY40, WRKY33, ERF6, AHL15, AHL17, ANAC003, TCP13 and ANAC019, were also strongly induced in infected leaves, similar to CKX5. Analysis of ERF6-overexpressing plants and ERF6-and AHL15-knockout mutants indicated that ERF6 and AHL15 are involved in plant immunity to B. cinerea. Furthermore, CKX5 upregulation by B. cinerea infection was affected when ERF6 or AHL15 levels were altered. Our work suggests that CKX5 levels are controlled by the plant defense system to defend against attack by the pathogen B. cinerea.


Introduction
Plants have evolved a concerted suite of defense systems involving phytohormone pathways in response to pathogen attack, among which is the backbone formed by jasmonate/ethylene (JA/ ET) and salicylic acid (SA) [1,2].Additionally, cytokinins (CKs), essential hormones controlling almost every aspect of plant growth and development, are known to be important in both the defense response and pathogen virulence [3,4].In some cases, (hemi)biotrophic pathogens can secrete CKs or induce CK production in the host plant to achieve pathogenesis.The hemibiotropic pathogen Rhodococcus fascians secretes a combination of various CK species, which are recognized by CK receptors in Arabidopsis and cause proliferation of young shoot tissues into differentiated leafy galls [5].In addition, it has been suggested that CK-secreting biotrophs or hemibiotrophs manipulate CK signaling to regulate the host cell cycle and nutrition allocation [6].Conversely, plant immunity uses CKs to resist pathogen infection through different mechanisms.It has been suggested that host-derived CKs influence plant immunity in an SAdependent manner based on the finding that the cytokinin-activated transcription factor ARR2 interacts with TGA3, a transcription factor of SA-responsive genes, to regulate defenserelated genes in response to biotrophic infection [7].On the other hand, CKs mediate resistance against Pseudomonas syringae in tobacco by increasing antimicrobial phytoalexin synthesis independent of SA signaling [8].
Studies concerning the role of CKs in plant interactions with necrotrophic pathogens such as B. cinerea, which does not produce detectable CKs [9,10], are less abundant and more complex.Expression of the CK biosynthesis gene IPT under control of the SAG12 (senescence-specific gene) promoter results in increased resistance of Arabidopsis thaliana to B. cinerea infection [11].Gupta et al. [12] found that CK promotes the resistance of tomato to B. cinerea through an SA-and ET-dependent mechanism.We recently reported that CK responds to B. cinerea infection in a variety of ways that are differentially modulated by the JA and ET pathways in Arabidopsis [13].More recently, CKs have been described as having a dual role in plant-B.cinerea interactions, inhibiting B. cinerea development and virulence while positively regulating B. cinerea energy utilization [14,15].In our previous work, we found that the expression level of CYTOKININ OXIDASE/DEHYDROGENASE 5 (CKX5), which encodes a CK degrading protein, was markedly induced in locally infected leaves and distant leaves of the same plant without pathogen inoculation [13].To explore the mechanism by which CKX5 is strongly upregulated after B. cinerea infection, we performed yeast one-hybrid (Y1H) assays to reveal the transcription factors (TFs) that bind to the CKX5 promoter, and we preliminarily investigated relationships between the screened TFs and CKX5 during B. cinerea infection in Arabidopsis.

Plant material and growth conditions
All Arabidopsis thaliana plants used in this study were of the Columbia-0 (Col-0) ecotype.Seeds of erf6 (SALK_087356, NASC stock number: N2103286), ERF6 overexpressing line (TPT_4.17490.1A,NASC stock number: N2102191) and ahl15 (SALK_040729C, NASC stock number: N668128) were obtained from the Arabidopsis seeds share center (Arashare, https:// www.arashare.cn/).Detailed information about these seeds can be found on the website of Arabidopsis Biological Resource Center (ABRC, https://abrc.osu.edu/).The coding sequence (CDS) of CKX5 was cloned using reverse transcription-PCR from total RNA obtained from Arabidopsis seedlings.The PCR products were subsequently sequenced and introduced into the binary vector pCAMBIA1300, under the control of the super promoter, which consists of Cauliflower Mosaic Virus (CaMV) 35S promoter and the mannopine synthase gene promoter.This resulting construct was designated as 1300-CKX5.To create transgenic lines of CKX5OE, 1300-CKX5 was transformed into Arabidopsis Col-0 genome using the Agrobacterium-mediated flower-dip method [16].The plants were grown in growth chamber under conditions of 16 h' white light (*100 μmol m −2 s −1 ) at 23 ˚C/8 h' darkness at 20 ˚C, with 60% humidity.Seeds of Arabidopsis plants used for B. cinerea infection experiments were surface-sterilized and grown on plates with solid Murashige and Skoog (MS) medium for two weeks after vernalization.Then, the plates were transplanted to sterile soil and grown under a 12 h light/12 h dark photoperiod.

Pathogen Bioassays
For pathogen infection, the manipulations about B. cinerea (strain B05.10) cultivation, preparation and inoculation of spore suspension, the analysis of disease symptoms were performed according to [13].After culturation on potato dextrose agar (PDA) medium (Coolaber, Beijing) for 10 days, The B. cinerea spores were harvested and suspended in a half-strength potato dextrose broth (1/2 PDB) to achieve a final concentration of 2.5 × 10 5 spores mL -1 .Subsequently, 4 μL droplet of the spore suspension was placed onto the adaxial side of the leaves of intact plants that were 4 weeks old.After the inoculations, all the plants were placed under sealed transparent hoods in a high humidity environment.
Two days after the B. cinerea infection, the symptoms on the infected leaves were examined.Lesion diameters on the leaves were measured using Image J software.Following the method described by [17], the disease symptoms on the inoculated leaves were recorded and classified into four groups: class I for lesion diameters less than 2 mm, class II for 2-mm lesions with chlorosis, class III for 2-4 mm lesions with chlorosis, and class IV for lesions with a spread greater than 4 mm.

Quantitative real-time reverse transcription PCR (qRT-PCR)
After inoculation of B. cinerea spores and mock solution, the leaves at corresponding time points were collected and frozen in liquid nitrogen for total RNA extraction and qRT-PCR.This whole process proceeded according to [13].Primers used for reference genes and genes of interest are listed in S1 Table.

Yeast one-hybrid screening
The Y1H screening assays were performed using a Matchmaker Gold Yeast One-Hybrid Library Screening System.The prey pool containing TFs cDNA library inserted into the prey vector pGADT7 was obtained from Dr. Huayan Zhao (Hubei University).Five CKX5 promoter fragments were separately inserted into pAbAi to construct the pAbAi-baits.Then the pAbAi-bait plasmids were linearized and transformed into Y1HGold.The colonies were selected on the synthetically defined medium lacking uracil (SD-Ura).After determining the minimal inhibitory concentrations of aureobasidin A (AbA) for the bait strains, the linear prey vector pool was co-transformed into the bait yeast strains and selected on synthetic dextrose (SD)/-Leu/AbA plates.The plasmids in the positive colonies were extracted and sequenced to confirm the corresponding gene number in TAIR (The Arabidopsis Information Resource, https://www.arabidopsis.org/).The primers used for cloning CKX5 promoter fragments are listed in S2 Table.

Pair-wise yeast one-hybrid
The coding DNA of the selected TFs were cloned from Arabidopsis genome by PCR and inserted into the MCS of pGADT7 to make pGADT7-TFs.Then linearized pGADT7-TFs were separately transformed into the corresponding bait-strains and grown on SD-Leu/AbA medium.The primers used to amplify TFs are listed in S3 Table.

Regulation of CKX family genes by B. cinerea infection
In our previous work, it was shown that expression of CKX5 was markedly induced in B. cinerea-infected leaves and distant leaves of the same plant [13].In the present study, we analyzed transcript levels of the Arabidopsis CKX family by qRT-PCR.CKX2 is not shown here due to the undetectable expression in B. cinerea-infected and mock leaves.Based on the results, CKX genes were differentially regulated by B. cinerea infection vs. the mock-treated leaves (Fig 1).Among them, CKX1 responded rapidly, being upregulated at 14 h and 24 h post inoculation (hpi) but downregulated at 48 hpi.CKX3 was only induced at 24 hpi.Consistent with our previous work [13], CKX4 expression was reduced at 48 hpi.Markedly and consistently, the transcript level of CKX5 was significantly increased after B. cinerea infection.The other two CKX genes, CKX6 and CKX7, were not affected after B. cinerea inoculation.Combined with the CKX5:GUS analysis in [13], our qRT-PCR results suggest an important role for CKX5 in plant immunity to B. cinerea infection.

CKX5-overexpressing Arabidopsis plants show resistance to B. cinerea infection
To further confirm the importance of CKX5 in the plant-B.cinerea interaction, we generated transgenic Arabidopsis plants with much higher transcript levels of CKX5 than wild-type (WT) plants (Figs 2a and S1).Then, four-week-old CKX5-overexpression lines were inoculated with B. cinerea spores, and disease symptoms were recorded 2 days later.The results shown in
An Arabidopsis cDNA library was transformed into the bait strains, followed by selection on SD medium lacking leucine (SD-Leu) with corresponding concentrations of AbA.Prey TFs from the positive colonies were chosen after checking that their expression was regulated by B. cinerea infection using the microarray platform GENEVESTIGATOR (https://genevestigator. com/gv/).Together, 13 potential TFs responding to B. cinerea infection according to GENE-VESTIGATOR were selected for further analysis (S4 Table ).
Next, pair-wise yeast one-hybrid assays were performed after the coding sequences of the 13 TFs were separately inserted into the pGADT7 vector (Fig 4a)  To further confirm the binding of these TFs to the CKX5 promoter sequence, JASPAR (https://jaspar.genereg.net/)was used to predict binding sites for the TFs, but only binding sites for WRKY40, WRKY33 and SPL3 were found.Then, the predicted binding sequences of the three TFs were mutated as depicted in Fig 5a and used as bait sequences for yeast onehybrid assays.The growth of yeast colonies carrying mutated bait sequences was completely arrested on SD-Leu medium with 50 ng/mL AbA; however, the growth of yeast colonies with original bait sequences was not when transformed with corresponding TFs (Fig 5b).This further suggested that WRKY40, WRKY33 and SPL3 interact with the CKX5 promoter by binding at corresponding sites.

The screened TFs are involved in plant immunity to B. cinerea
To investigate whether the TFs screened by yeast one-hybrid are truly involved in the response to B. cinerea infection, transcript levels of WRKY40, ARKY33, ERF6, AHL17, SPL3, AHL15, ANAC003, TCP13, ANAC019 and BBX14 were analyzed by qRT-PCR in Arabidopsis leaves inoculated with B. cinerea at 14, 24 and 48 hpi (Fig 6).According to the data, the transcript levels of WRKY40, WRKY33, ERF6 and ANAC003 were strongly upregulated after B. cinerea inoculation, at levels 40-fold, 29-fold, 13.4-fold and 400-fold of mock treatments at 48 hpi.Conversely, SPL13 and BBX14 expression was repressed.For AHL17, AHL15, TCP13 and

Discussion and conclusions
With qRT-PCR and analysis of the reporter line CKX5:GUS, we previously showed that Arabidopsis CKX5 expression is remarkably induced in leaves infected with B. cinerea, especially when the leaves show some necrotic areas due to B. cinerea [13].Such high expression of CKX5 was also detected in the distant leaves of the same plant not inoculated when B. cinerea successfully caused decay of part of the inoculated leaves.Therefore, we addressed the involvement of CKX5 in Arabidopsis-B.cinerea interaction and potential of TFs regulating CKX5 in response to B. cinerea infection in the present study.
In Arabidopsis thaliana, there are seven distinct CKX genes (CKX1-CKX7) that encode cytokinin oxidase/dehydrogenases responsible for the breakdown of CKs in vivo, though their preferred substrates and enzyme activity differ under different conditions [19,20].In our study, CKX genes were found to respond differently to B. cinerea infection (Fig 1).The most prominently changed CKX gene was CKX5, especially at 48 hpi, when most of the inoculated leaves had decayed.Transgenic Arabidopsis plants overexpressing CKX5 showed increased resistance to B. cinerea compared with WT plants ( Fig 2), which does not agree with reported results that tomato leaves overexpressing CKX4 are more sensitive to B. cinerea infection [12].This may be explained as follows.First, CKX4 and CKX5 responded differently to B. cinerea infection (Fig 1 ; Ref. 13).More importantly, CK has a dual role in B. cinerea biology [14,15]: (1) CK strongly inhibits B. cinerea growth and virulence by targeting the cytoskeleton and cellular trafficking in high nutritional availability; (2) CK increases the metabolism of B. cinerea under energy-restrictive conditions.Additionally, during the battle between plant and pathogen, necrotrophic fungal pathogens have been reported to cause the formation of green areas termed green necronissia in leaves, which might be induced by CKs [21].Thus, our current results regarding CKX5 suggest that to block the benefits that B. cinerea receive from CKs, CKX5 is deployed to manipulate CK levels when B. cinerea successfully attacks the host.By yeast one-hybrid screens, some TFs were found to bind to the promoter of CKX5.qPCR analysis showed that the transcript levels of selected TFs were significantly changed during B. cinerea infection (Fig 6).Among them, WRKY40, WRKY33, ERF6 and ANAC003 responded in a manner similar to CKX5 in leaves inoculated with B. cinerea.WRKY proteins are wellknown TFs that function in plant immunity to biotic stress.WRKY40 interacts with WRKY18 and WRKY60 to function in a pattern of overlapping, antagonistic, and distinct roles in plant responses to Pseudomonas syringae and B. cinerea [22].WRKY33 is a key transcriptional regulator in plant immunity, and it has been indicated that B. cinerea B05. 10 promotes disease by suppressing WRKY33-mediated host defenses [23,24].ERF6 is a positive regulator of JA/ETmediated defense against B. cinerea in Arabidopsis [25].ANAC003 is a senescence-related protein that activates expression of senescence-associated genes (SAGs), resulting in plant senescence [26].The genes encoding another four TFs, including AHL15, AHL17, TCP13 and ANAC019, were strongly induced at the late stage of B. cinerea infection (48 hpi).Transient expression of AHL15 in protoplasts blocks pathogen-associated molecular pattern (PAMP)associated gene expression, suggesting its important role in plant immunity [27].ANAC019 has been indicated to function as a transcription activator to regulate JA-induced expression of defense genes [28].Most of these TFs are involved in plant defense against pathogens.ERF6overexpressing plants and ERF6 and AHL15 mutants further proved the upregulation role of ERF6 and AHL15 in CKX5 during B. cinerea infection (Fig 7).However, the results indicate that strong upregulation of CKX5 was not only controlled by ERF6 and AHL15.It is likely that most of the TFs mentioned above are involved in this process.We therefore propose that CKX5 is manipulated by TFs belonging to the plant defense system to counteract the necrotrophic phase of infection by B. cinerea.

Fig 1 .
Fig 1.Expression profiles of cytokinin oxidase (CKX) genes in Arabidopsis plants after Botrytis cinerea inoculation.Transcript levels of CKX family were determined by qRT-PCR in 4-weeks-old leaves inoculated with B. cinerea spores at different time-points, and taking leaves inoculated with the same amount of 1/2 PDB solution as the mock treatment.Expression levels in the mock leaves at 14 h post inoculation were set to a value of 1.All data were normalized to the expression of EXP (At4g26410).Asterisks indicate significant differences between B. cinerea and mock-treated samples at the same time-point (two-tailed Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001).Error bars are standard deviations (n = 3).Three independent experiments were performed with a similar outcome; results from one representative experiment are shown.https://doi.org/10.1371/journal.pone.0298260.g001 Fig 2b-2d clearly reveal much smaller lesions and less severe disease symptoms in leaves from the plants overexpressing CKX5 compared with WT plants, indicating that the CKX5-overexpressing plants were more resistant to B. cinerea.

Fig 3 .
Fig 3. Generation of bait strains for yeast one-hybrid (Y1H) screening.(a) Physical map of the bait plasmids used for integration into yeast genome.Five short overlapping promoter fragments of CKX5 were separately inserted into the vector of pAbAi through HindIII and XhoI restriction sites.Aur R , encodes aureobasidin A-resistant mutant of inositol phosphorylceramide synthase, confers fungal resistance to Aureobasidin A (AbA); Amp R , confers resistance to ampicillin, carbenicillin, and related antibiotics; URA3, encodes orotidine5'-phosphatedecarboxylase.(b) Determination the minimal inhibitory concentrations of AbA for the bait strains.The numbers on the right side of AbA represent the concentration of AbA with unit of ng/mL.SD-Ura means synthetically defined medium lacking uracil.https://doi.org/10.1371/journal.pone.0298260.g003 . The results revealed that ANAC019 (NAC DOMAIN-CONTAINING PROTEIN 19, AT1G52890), WRKY40 (WRKY DNA-BINDING PROTEIN 40, AT1G80840), WRKY33 (WRKY DNA-BINDING PROTEIN

Fig 4 .
Fig 4. Confirmation of the 13 screened transcription factors (TF) with pair-wise yeast one hybrid.(a) Coding sequences (CDS) of the screened 13 TFs were separately cloned into the prey vector pGADT7 AD in frame with the activation domain (AD).The corresponding gene name and restriction sites used for cloning were shown.P ADH1 , constitutive ADH1 promoter; GAL4 AD, encodes GAL4 activation domain; LEU2, nutritional marker for selection in yeast.(b-d) The prey vectors were transformed into the corresponding bait strains and grown on SD medium lacking leucine (SD-Leu) with corresponding concentrations of AbA.The control stains contained empty pGADT7 AD and bait vector.https://doi.org/10.1371/journal.pone.0298260.g004

Fig 5 .
Fig 5. Mutation of the binding sequences repressed the interactions between WRKY40, WRKY33, SPL3 and CKX5 promoter fragments.(a) Physical map of the CKX5 promoter illustrating the changes of binding sites for WRKY40, WRKY33 and SPL3.The arrows showed how the original sequences were changed.(b) Determination of the interactions between the three transcription regulators and the mutated promoter of CKX5 with yeast one-hybrid assays.Y1H Gold strains successfully transformed with corresponding vectors were grown on the SD-Leu/AbA plates at 30 ˚C for 3-5 days.PCKX5-2 or PCKX5-5 represents the original fragment of CKX5 promoter.PMuCKX5-2 or PMuCKX5-5 means the fragments with mutated binding sites.The numbers on the right side of AbA represent the concentration of AbA with unit of ng/mL.https://doi.org/10.1371/journal.pone.0298260.g005

Fig 6 .
Fig 6.Expression levels of selected transcription regulator genes in response to Botrytis cinerea.Transcript accumulations of the genes were determined by real-time qPCR in 4-weeks-old wild type Arabidopsis leaves inoculated with B. cinerea at different time-points.Leaves inoculated with same amount of 1/2 PDB solution as mock.Expression levels in the mock leaves at 14 h post-inoculation (hpi) were set to a value of 1. Error bars are standard deviations (n = 3).Differences are significant at P < 0.05 (*), P < 0.01 (**) and P < 0.001 (***) by Student's t-test.https://doi.org/10.1371/journal.pone.0298260.g006

Fig 7 .
Fig 7. Alteration of in vivo EFR6 or AHL15 levels changed the susceptibility of Arabidopsis plants and the response of CKX5 to Botrytis cinerea infection.(a-c) Leaves of 4-weeks-old wild type (WT) and erf6 mutant, ERF6 overexpression plants (ERF6OE), ahl15 mutant were inoculated with B. cinerea spores and the disease symptoms were analyzed two days later.Scale bars in photographs of the infected leaves are 1 cm.The values shown for disease lesion size in the indicated genotypes are means ± SD (n = 50 inoculated leaves).Asterisks indicate significant differences between WT and mutant or transgenic plants (two-tailed Student's t-test, **P < 0.01; ***P < 0.001).The distribution of leaves in the class I-IV was calculated from 50 leaves.Class I, lesion < 2 mm; Class II, 2 mm lesion plus chlorosis; Class III, lesion 2-4 mm plus chlorosis; Class IV, lesion > 4 mm plus chlorosis.The significance of differences was analyzed by χ 2 -test.**P < 0.01; ***P < 0.001.Meanwhile, CKX5 expression levels were analyzed in leaves of the four genotypes at 24 h and 48 h post-inoculation with B. cinerea or mock solution.Expression values in WT leaves with mock treatment were set to a value of 1. Bars are the means ± SD (n = 3).Three independent experiments were performed with similar outcomes; results from one representative experiment are shown.Asterisks indicate significant differences between WT and erf6, ERF6OE, or ahl15 with same treatments (two-tailed Student's ttest, *P < 0.05; **P < 0.01).https://doi.org/10.1371/journal.pone.0298260.g007