The PIFs Redundantly Control Plant Defense Response against Botrytis cinerea in Arabidopsis

Endogenous and exogenous signals are perceived and integrated by plants to precisely control defense responses. As a crucial environmental cue, light reportedly plays vital roles in plant defenses against necrotrophic pathogens. Phytochrome-interacting factor (PIF) is one of the important transcription factors which plays essential roles in photoreceptor-mediated light response. In this study, we revealed that PIFs negatively regulate plant defenses against Botrytis cinerea. Gene expression analyses showed that the expression level of a subset of defense-response genes was higher in pifq (pif1/3/4/5) mutants than in the wild-type control, but was lower in PIF-overexpressing plants. Chromatin immunoprecipitation assays proved that PIF4/5 binds directly to the ETHYLENE RESPONSE FACTOR1 (ERF1) promoter. Moreover, genetic analyses indicated that the overexpression of ERF1 dramatically rescues the susceptibility of PIF4-HA and PIF5-GFP transgenic plants, and that PIF controls the resistance to B. cinerea in a COI1- and EIN2-dependent manner. Our results provide compelling evidence that PIF, together with the jasmonate/ethylene pathway, is important for plant resistance to B. cinerea.


Introduction
As sessile organisms, stress-resistant plants employ complex defense mechanisms to counter the adverse effects of multiple pathogens and/or herbivorous insects. The precisely controlled immunity is vital for the survivability of plants in the interaction with pathogens. According to the lifestyles, plant pathogens can be divided into biotrophs, which obtain nutrition from living host cells, and necrotrophs, which feed on the dead plant tissue [1]. As one of the typical broad host-range necrotrophic pathogens which can infect plants with no host specificity, Botrytis cinerea has been found to cause huge economical losses in agricultural production and is used as a model fungus to study the interaction between plants and pathogens.
There has been considerable progress in characterizing the molecular mechanisms underlying immune signaling networks against B. cinerea over the past several years [2]. Specifically, the plant hormones jasmonate (JA) and ethylene (ET) play crucial roles in plant defenses against necrotrophic pathogens [1,2]. Jasmonate is a lipid-derived compound, perceived by the F-box protein CORONATINE whether PIFs function in plant defenses against B. cinerea, we examined the WT plants, several single mutants (pil5-2, pif3-7, pif4-2, and pif5-3), and the pifq mutant. At six days post-inoculation, the symptom development of the pil5-2, pif3-7, and pif5-3 plants was similar to that of the WT plants. In contrast, the pif4-2 and pifq mutants had slightly and significantly less extensive disease symptoms, respectively, when compared with the WT plants (Figure 1a). Similarly, the pif4-2 and pifq mutant plants had less pathogen biomass, when compared with the WT plants at three days post-inoculation ( Figure 1c). Additionally, the pifq plants accumulated considerably less β-tubulin mRNA than the pif single mutants and the WT plants (Figure 1c). These results implied that PIFs function redundantly and negatively, to regulate plant defenses against B. cinerea.
Plants 2020, 9, x FOR PEER REVIEW 3 of 14 cinerea, we examined the WT plants, several single mutants (pil5-2, pif3-7, pif4-2, and pif5- 3), and the pifq mutant. At six days post-inoculation, the symptom development of the pil5-2, pif3-7, and pif5-3 plants was similar to that of the WT plants. In contrast, the pif4-2 and pifq mutants had slightly and significantly less extensive disease symptoms, respectively, when compared with the WT plants ( Figure 1a). Similarly, the pif4-2 and pifq mutant plants had less pathogen biomass, when compared with the WT plants at three days post-inoculation ( Figure 1c). Additionally, the pifq plants accumulated considerably less β-tubulin mRNA than the pif single mutants and the WT plants ( Figure 1c). These results implied that PIFs function redundantly and negatively, to regulate plant defenses against B. cinerea.   [11,12,38]. To explore the molecular basis of the altered responses of the pifq mutants to the necrotrophic fungal pathogen, we first analyzed the expression of several defense-related genes in these plants, after an infection by B. cinerea. The ERF1, ORA59, PDF1.2, and HEL expression levels were higher in the pifq plants than in the WT plants, following the B. cinerea infection ( Figure 1d). As shown in Figure 1d, it is obvious that pifq showed significantly high basal expression of these four genes. Thus, we hypothesized that pif mutants might have a more powerful basal defense. To further investigate this possibility, we tested the expression of defense-related genes in pif plants. The pif mutants we used showed a high expression level of defense-response genes, including ERF1, ORA59, PDF1.2, and ERF104 ( Figure 1e). Moreover, previous studies proved that the expression levels and the protein abundance of PIFs are affected by diurnal conditions [39][40][41][42]. Because of their functional redundancy, we then used pifq mutants to examine the expression levels of several defense-related genes during a 24 h period. The expression levels of defense-associated genes, such as ERF1, ORA59, PDF1.2, ERF5, and ERF6, were higher in the pifq plants than in the WT plants at the ZT0~ZT3 and ZT9~ZT12 time-points ( Figure 2). Considering the clearly close correlation between high expression of defense response genes under both conditions with and without B. cinerea treatment and the largely resistant phenotypes after its infection in pifq plants, we concluded that PIFs regulate defense-associated genes expression to control plant resistance against B. cinerea.

The Effect of PIFs in Regulating the Basal and Pathogen-Induced Expression of Defense-Related Gene
To defend against necrotrophic pathogens, plants activate a set of defense-related genes, including ERF1, ORA59, PATHOGEN INDUCIBLE PLANT DEFENSIN (PDF1.2), and HEVEIN-LIKE PROTEIN (HEL) [11,12,38]. To explore the molecular basis of the altered responses of the pifq mutants to the necrotrophic fungal pathogen, we first analyzed the expression of several defense-related genes in these plants, after an infection by B. cinerea. The ERF1, ORA59, PDF1.2, and HEL expression levels were higher in the pifq plants than in the WT plants, following the B. cinerea infection ( Figure 1d). As shown in Figure 1d, it is obvious that pifq showed significantly high basal expression of these four genes. Thus, we hypothesized that pif mutants might have a more powerful basal defense. To further investigate this possibility, we tested the expression of defense-related genes in pif plants. The pif mutants we used showed a high expression level of defense-response genes, including ERF1, ORA59, PDF1.2, and ERF104 ( Figure 1e). Moreover, previous studies proved that the expression levels and the protein abundance of PIFs are affected by diurnal conditions [39][40][41][42]. Because of their functional redundancy, we then used pifq mutants to examine the expression levels of several defense-related genes during a 24 h period. The expression levels of defense-associated genes, such as ERF1, ORA59, PDF1.2, ERF5, and ERF6, were higher in the pifq plants than in the WT plants at the ZT0~ZT3 and ZT9~ZT12 time-points ( Figure 2). Considering the clearly close correlation between high expression of defense response genes under both conditions with and without B. cinerea treatment and the largely resistant phenotypes after its infection in pifq plants, we concluded that PIFs regulate defense-associated genes expression to control plant resistance against B. cinerea.

The Effect of PIF Overexpressions on Resistance to B. cinerea
To further characterize the role of PIFs in defense responses to B. cinerea, we also examined the basal expression of defense-associated genes in transgenic plants overexpressing PIF1-GFP, MYC-PIF3, PIF4-HA, or PIF5-GFP. As shown in Figure 3c, these overexpression plants had lower expression of ERF1, ORA59, PDF1.2, and ERF104, as compared to WT. Then we compared the pathogen growth in these transgenic plants with that in WT plants. At four days after an inoculation with B. cinerea, we observed a rapid increase in the necrotic symptoms and β-tubulin mRNA accumulation in the plants overexpressing MYC-PIF3, PIF4-HA, or PIF5-GFP (Figure 3a,b). Thus, the constitutive overexpression of PIF genes decreased the resistance of transgenic plants to B. cinerea and accelerated disease symptom development, suggesting that PIFs play an important role in plant defenses against B. cinerea. Figure 2. The basal expression of defense-response genes in WT and pifq plants. The plants were grown on soil, under LD condition, for 14 days. The plants for RNA isolation were harvested at given times from ZT0. IPP2 gene was used as an internal control. Error bars indicate SD of three independent RNA extracts.

The Effect of PIF Overexpressions on Resistance to B. cinerea
To further characterize the role of PIFs in defense responses to B. cinerea, we also examined the basal expression of defense-associated genes in transgenic plants overexpressing PIF1-GFP, MYC-PIF3, PIF4-HA, or PIF5-GFP. As shown in Figure 3c, these overexpression plants had lower expression of ERF1, ORA59, PDF1.2, and ERF104, as compared to WT. Then we compared the pathogen growth in these transgenic plants with that in WT plants. At four days after an inoculation with B. cinerea, we observed a rapid increase in the necrotic symptoms and β-tubulin mRNA accumulation in the plants overexpressing MYC-PIF3, PIF4-HA, or PIF5-GFP (Figure 3a,b). Thus, the constitutive overexpression of PIF genes decreased the resistance of transgenic plants to B. cinerea and accelerated disease symptom development, suggesting that PIFs play an important role in plant defenses against B. cinerea. , asterisks indicate Student's t-test significant differences (* p < 0.05, ** p < 0.01).

Expression Profiling to Identify the Potential PIF-Involved Pathway to Control Defense Response
The obvious differences in the symptom development of the pifq mutant and PIF-overexpressing plants compelled us to further investigate the molecular basis of the altered responses. The basal expression levels of the defense-associated genes were higher and lower in the  For (b,c), asterisks indicate Student's t-test significant differences (* p < 0.05, ** p < 0.01).

Expression Profiling to Identify the Potential PIF-Involved Pathway to Control Defense Response
The obvious differences in the symptom development of the pifq mutant and PIF-overexpressing plants compelled us to further investigate the molecular basis of the altered responses. The basal expression levels of the defense-associated genes were higher and lower in the pifq and PIF-overexpressing plants, respectively (Figures 1e and 3c). Thus, we completed a genome-wide transcriptomic analysis to compare the gene expression profiles between WT and pifq plants. The analysis revealed 409 differentially expressed genes of the pifq plants with more than two-fold changes in expression, including 137 downregulated genes and 272 upregulated genes (Supplementary Materials Data S1). Consistent with the enhanced disease resistance of the pifq plants, JA-and ET-related Plants 2020, 9, 1246 6 of 13 defense genes were among the 272 upregulated genes, including ERF1, ERF5, ERF6, WRKY33, PDF1.2, and PDF1.3 (Figure 4a; Supplementary Materials Data S1). A gene ontology (GO) analysis functionally annotated genes with representative GO terms associated with defense responses to fungi or JA and the ethylene-activated signaling pathway (Figure 4b), suggesting that PIFs might regulate plant defenses against B. cinerea by modulating the transcription of defense-associated genes and that the function of PIFs to regulate defense response might be highly involved in the JA and ET signaling pathway.
pifq and PIF-overexpressing plants, respectively (Figures 1e and 3c). Thus, we completed a genome-wide transcriptomic analysis to compare the gene expression profiles between WT and pifq plants. The analysis revealed 409 differentially expressed genes of the pifq plants with more than two-fold changes in expression, including 137 downregulated genes and 272 upregulated genes (Supplementary Materials Data S1). Consistent with the enhanced disease resistance of the pifq plants, JA-and ET-related defense genes were among the 272 upregulated genes, including ERF1, ERF5, ERF6, WRKY33, PDF1.2, and PDF1.3 (Figure 4a; Supplementary Materials Data S1). A gene ontology (GO) analysis functionally annotated genes with representative GO terms associated with defense responses to fungi or JA and the ethylene-activated signaling pathway (Figure 4b), suggesting that PIFs might regulate plant defenses against B. cinerea by modulating the transcription of defense-associated genes and that the function of PIFs to regulate defense response might be highly involved in the JA and ET signaling pathway.

In Vivo Interaction between PIF4/5 and the ERF1 Promoter
Previous reports indicated that PIFs function by binding directly to the G-box (CACGTG) of their target gene promoters [30,34,[43][44][45]. Given the existence of a G-box in the ERF1 promoter ( Figure 5a) and the negative regulation of ERF1 expression by PIFs, we assumed that PIFs suppress ERF1 expression by binding directly to the promoter. To confirm this assumption, we first conducted chromatin immunoprecipitation (ChIP) assays with transgenic lines expressing pPIF4-PIF4-MYC in pifq background and PIF5-HA under the control of the cauliflower mosaic virus (CaMV) 35S promoter. The ChIP-qPCR results further demonstrated that PIF4 and PIF5 can bind to the ERF1 promoter via the G-box (Figure 5b and Supplementary Materials Figure S1a). Furthermore, to examine whether ERF1 is directly targeted by PIF4 and PIF5 in vitro, gel electrophoresis mobility shift assay (EMSA) was performed. The GST-PIF4/5 specifically bound to the ERF1 promoter region containing the normal G-box sequence (Figure 5c and Supplementary Materials Figure S1b). Additionally, with increasing concentrations, the unlabeled competitor greatly inhibited the binding of GST-PIF4/5. These observations indicated that PIFs can bind directly to the ERF1 promoter.
To confirm the negative regulatory functions of PIFs, we performed transient expression assays in WT A. thaliana mesophyll protoplasts, using a dual-luciferase reporter plasmid. As a reporter, the ERF1 promoter was fused to the firefly luciferase (LUC) gene, whereas the Renilla luciferase (REN)

In Vivo Interaction between PIF4/5 and the ERF1 Promoter
Previous reports indicated that PIFs function by binding directly to the G-box (CACGTG) of their target gene promoters [30,34,[43][44][45]. Given the existence of a G-box in the ERF1 promoter ( Figure 5a) and the negative regulation of ERF1 expression by PIFs, we assumed that PIFs suppress ERF1 expression by binding directly to the promoter. To confirm this assumption, we first conducted chromatin immunoprecipitation (ChIP) assays with transgenic lines expressing pPIF4-PIF4-MYC in pifq background and PIF5-HA under the control of the cauliflower mosaic virus (CaMV) 35S promoter. The ChIP-qPCR results further demonstrated that PIF4 and PIF5 can bind to the ERF1 promoter via the G-box (Figure 5b and Supplementary Materials Figure S1a). Furthermore, to examine whether ERF1 is directly targeted by PIF4 and PIF5 in vitro, gel electrophoresis mobility shift assay (EMSA) was performed. The GST-PIF4/5 specifically bound to the ERF1 promoter region containing the normal G-box sequence (Figure 5c and Supplementary Materials Figure S1b). Additionally, with increasing concentrations, the unlabeled competitor greatly inhibited the binding of GST-PIF4/5. These observations indicated that PIFs can bind directly to the ERF1 promoter.
gene was placed under the control of the CaMV 35S promoter. The effector constructs contained PIF genes or GFP driven by the constitutive CaMV 35S promoter. The coexpression of PIF genes with the proERF1-LUC reporter significantly decreased the LUC/REN ratio relative to the effects of GFP (Figure 5d). This result further supported that PIFs act as negative regulators to repress the expression of ERF1 and suggested that there is a possibility that PIFs control plant defense response through directly inhibiting the expression of other defense-related genes. (c) EMSA assay shows that GST-PIF4 bHLH recombinant protein binds to the promoter of ERF1 in vitro. DNA fragments of ERF1 promoter were synthesized with normal G-box motif (G-box probe) and mutated G-box motif (mutant probe), and labeled with biotin. The GST protein was used as negative control. (d) The diagram of the reporter and effectors used in the protoplast transfection assays. Transient expression assays indicate that PIFs repress ERF1 expression. The error bars indicate SD of three independent experiments. The different letters above columns indicate significant differences (one-way ANOVA; p < 0.05).

Association of PIF-Regulated Defense with JA/ET Signaling
Having ascertained that PIFs regulate the expression of several JA/ET-related defense genes following an infection by B. cinerea and the GO terms regarding JA/ET response enriched in transcriptome analysis, we examined whether the enhanced defense response of pifq plants is also associated with JA/ET signaling pathways. Specifically, coi1-16 and ein2 mutants were crossed with pifq plants, to generate coi1-16/pifq and ein2/pifq quintuple mutants. The defense responses of these quintuple mutants were similar to those of the coi1-16 and ein2 plants, with larger lesions compared with the pifq plants (Figure 6a), indicating that PIF-regulated defense response against B. cinerea relies on JA and ET signaling. Additionally, we explored the genetic relationship between PIF4/5 and ERF1. Plants overexpressing PIF4-HA and PIF5-GFP were crossed with a transgenic plant carrying the HA-ERF1 construct. The overexpression of ERF1 significantly decreased the disease To confirm the negative regulatory functions of PIFs, we performed transient expression assays in WT A. thaliana mesophyll protoplasts, using a dual-luciferase reporter plasmid. As a reporter, the ERF1 promoter was fused to the firefly luciferase (LUC) gene, whereas the Renilla luciferase (REN) gene was placed under the control of the CaMV 35S promoter. The effector constructs contained PIF genes or GFP driven by the constitutive CaMV 35S promoter. The coexpression of PIF genes with the proERF1-LUC reporter significantly decreased the LUC/REN ratio relative to the effects of GFP (Figure 5d). This result further supported that PIFs act as negative regulators to repress the expression of ERF1 and suggested that there is a possibility that PIFs control plant defense response through directly inhibiting the expression of other defense-related genes.

Association of PIF-Regulated Defense with JA/ET Signaling
Having ascertained that PIFs regulate the expression of several JA/ET-related defense genes following an infection by B. cinerea and the GO terms regarding JA/ET response enriched in transcriptome analysis, we examined whether the enhanced defense response of pifq plants is also associated with JA/ET signaling pathways. Specifically, coi1-16 and ein2 mutants were crossed with pifq plants, to generate coi1-16/pifq and ein2/pifq quintuple mutants. The defense responses of these quintuple mutants were similar to those of the coi1-16 and ein2 plants, with larger lesions compared with the pifq plants (Figure 6a), indicating that PIF-regulated defense response against B. cinerea relies on JA and ET signaling. Additionally, we explored the genetic relationship between PIF4/5 and ERF1. Plants overexpressing PIF4-HA and PIF5-GFP were crossed with a transgenic plant carrying the HA-ERF1 construct. The overexpression of ERF1 significantly decreased the disease susceptibility and pathogen biomass of the PIF4-HA and PIF5-GFP transgenic plants (Figure 6b,c and Supplementary Materials Figure S2). Thus, our results provide evidence that the PIF-mediated defense is dependent on JA and ET signaling pathway.  Figure S2). Thus, our results provide evidence that the PIF-mediated defense is dependent on JA and ET signaling pathway. For (a,c), the different letters above columns indicate significant differences (one-way ANOVA; p < 0.05).

Discussion
Previous studies proved that phyB negatively regulates PIF transcription factors, at the protein level, by enhancing their degradation and by sequestering them from their target promoters [46]. Thus, loss-of-function phyB mutants show exaggerated PIF-mediated growth [47][48][49]. Another study also revealed that phyb mutants exhibit downregulated expression of JA-inducible genes, such as HEL, ERF1, and PDF1.2, and compromised resistance to B. cinerea [49][50][51]. Additionally, PIF4 was recently reported to positively regulate the temperature-induced suppression of defense responses to Pto DC3000 [35]. Therefore, deciphering the possible roles of PIFs in defense responses to necrotrophic pathogens will provide new insights into our understanding of the phyB-PIF module-mediated plant defenses.
Pathogen attack leads to extensive transcriptional changes and the production of specialized metabolites contributing to the establishment of effective plant defenses. Transcription factors critically influence plant innate immunity. Notably, the ethylene response factor transcription factors, such as ERF1 and ORA59, are key integrators of JA and ET defense signaling pathways [52]. In the current study, the expression levels of several ERF genes, including ERF1 and ORA59, were downregulated in 35S:PIF transgenic plants and upregulated in pifq mutants. This expression model is consistent with the disease resistance of these lines (Figures 1-3), suggesting that PIFs negatively regulate ERF gene expression. Furthermore, a genetic analysis demonstrated that 35S:ERF1 can dramatically enhance disease resistance of PIF4-HA and PIF5-GFP transgenic plants (Figure 6b,c and Supplementary Materials Figure S2). All of these results, together, suggest that PIFs act upstream of ERF1 to negatively regulate the resistance to this necrotrophic pathogen. Our transcriptome For (a,c), the different letters above columns indicate significant differences (one-way ANOVA; p < 0.05).

Discussion
Previous studies proved that phyB negatively regulates PIF transcription factors, at the protein level, by enhancing their degradation and by sequestering them from their target promoters [46]. Thus, loss-of-function phyB mutants show exaggerated PIF-mediated growth [47][48][49]. Another study also revealed that phyb mutants exhibit downregulated expression of JA-inducible genes, such as HEL, ERF1, and PDF1.2, and compromised resistance to B. cinerea [49][50][51]. Additionally, PIF4 was recently reported to positively regulate the temperature-induced suppression of defense responses to Pto DC3000 [35]. Therefore, deciphering the possible roles of PIFs in defense responses to necrotrophic pathogens will provide new insights into our understanding of the phyB-PIF module-mediated plant defenses.
Pathogen attack leads to extensive transcriptional changes and the production of specialized metabolites contributing to the establishment of effective plant defenses. Transcription factors critically influence plant innate immunity. Notably, the ethylene response factor transcription factors, such as ERF1 and ORA59, are key integrators of JA and ET defense signaling pathways [52]. In the current study, the expression levels of several ERF genes, including ERF1 and ORA59, were downregulated in 35S:PIF transgenic plants and upregulated in pifq mutants. This expression model is consistent with the disease resistance of these lines (Figures 1-3), suggesting that PIFs negatively regulate ERF gene expression. Furthermore, a genetic analysis demonstrated that 35S:ERF1 can dramatically enhance disease resistance of PIF4-HA and PIF5-GFP transgenic plants (Figure 6b,c and Supplementary Materials Figure S2). All of these results, together, suggest that PIFs act upstream of ERF1 to negatively Plants 2020, 9, 1246 9 of 13 regulate the resistance to this necrotrophic pathogen. Our transcriptome sequencing analysis also indicated that the expression levels of the defense-response genes are widely upregulated. The GO functional annotations revealed that categories associated with JA/ET signaling are enriched among the 409 differentially expressed genes in pifq plants, implying the PIF-mediated defense against B. cinerea is closely related to JA/ET signaling (Figure 4). Similarly, mutations to COI1 and EIN2 in pifq plants dramatically compromised the resistance to B. cinerea (Figure 6a), further indicating that PIFs may be incorporated in the JA/ET pathway to control plant resistance to B. cinerea.
Numerous studies have demonstrated that the PIF proteins perform their biological functions by directly binding to the G-box (CACGTG) in their target promoters [30,34,[43][44][45]. The ERF1 promoter contains a G-box cis-elements. Our EMSA and ChIP experiments revealed that PIF4/5 can bind directly to the G-box in the ERF1 promoter (Figure 5b,c and Supplementary Materials Figure S1), suggesting that ERF1 is a direct target of PIF. The opposite expression patterns of ERF1 in PIF mutants and overexpressing lines, as well as the downregulation of LUC expression in the transient expression assays (Figures 1e, 3c and 5d), further suggest that PIFs are negative regulators of ERF1 expression. Thus, our results provide evidence that PIFs may function as negative regulators of plant defenses against B. cinerea via the direct inhibition of ERF1 expression. To more thoroughly characterize the biological functions of PIFs and their possible signaling pathways in defense responses to B. cinerea, their downstream target genes will need to be identified. Moreover, the 409 differentially expressed genes in pifq plants may include other PIF targets. Previous studies indicated that PIF proteins can function as both positive and negative regulators [24,[28][29][30][32][33][34][35]. Thus, PIFs extensively participate in the fine-tuning and tight control of the complex signaling and transcriptional networks that mediate plant growth and stress responses by functioning as positive and negative regulators.

Pathogen Infection
Botrytis cinerea (B05.10) was grown on Potato Dextrose Agar, under 12 h light/12 h dark, at 21 • C. B. cinerea spores' collection and inoculation of plants were performed as previously described [10]. The inoculated plants were maintained in a dark and high-humidity environment. After 3 to 7 days, the symptom development could be observed. B. cinerea growth was quantified by qRT-PCR of total RNA isolated from the inoculated plants. For drop inoculation, leaves of 28-day-old plants grown on soil were inoculated with a single 8 µL drop of a suspension of 5 × 10 5 spores/mL, in Sabouraud maltose broth (SMB) buffer. The lesion sizes of B. cinerea infected leaves were measured, using ImageJ.

Expression Analysis
For reverse-transcription PCR analysis, total RNA was extracted from Arabidopsis seedlings, using the TRIzol reagent (Invitrogen, Waltham, MA, USA). Then, cDNA was synthesized from 1 µg of total RNA, according to the reverse-transcription protocol (Takara, Beijing, China). The cDNA was subjected to qPCR, using the SYBR Premix Ex Taq (Takara, Beijing, China), on a Roche LightCycler 480. ACTIN2 or IPP2 was amplified as the reference gene. The 2 −∆∆Ct method was used for the calculation of relative expression levels. At least three biological replicates were conducted for each experiment. The gene-specific primers are provided in Supplementary Materials Table S1.

RNA Sequencing
The rosettes of 14-day-old WT and pifq plants grown on soil were collected. The total RNA was extracted from Arabidopsis seedlings, using the Trizol reagent (Invitrogen, Waltham, MA, USA). Shanghai OE Biotech Co. provided the supports of RNA sequencing and data analysis. TruSeq Stranded mRNA LTSample Prep Kit (Illumina, San Diego, CA, USA) was used to construct the libraries. Then the sequencing was performed, using Illumina HiSeqTM 2500. Trimmomatic was used to filter the raw sequencing reads. The differential genes were analyzed by DESeq, with p < 0.05 and fold change > 2 as the threshold. The GOseq was used to perform GO enrichment analysis of the differential genes. The hierarchical cluster of selected genes was constructed by using R package pheatmap.

Gel Mobility Shift Assay
For EMSA assay, PIF4 bHLH domain and the full-length PIF5 coding sequences were cloned into pGEX-TX-1 vector and transformed into E. coli strain transetta (DE3) (TransGen, Beijing, China). The recombinant proteins were induced at 37 • C for 3 h, with 0.1 mM IPTG, and purified with Glutathione-agarose beads (TransGen, Beijing, China). The 5 terminal biotin-labeled DNA fragments were synthesized. The EMSA assay was performed, using Chemiluminescent EMSA Kit (Beyotime, Shanghai, China), following the manufacturer's protocol.

Chromatin Immunoprecipitation Assay
For ChIP analyses, 18-day-old soil-grown plants, under LD conditions, were harvested at ZT0 (zeitgeber time) and cross-linked with 1% formaldehyde. The ChIP experiment was performed by using the EpiQuik Plant Chromatin Kit (Epigentek, Farmingdale, NY, USA). The normal mouse lgG was used as the negative control. The mouse anti-MYC and anti-HA antibody (Santa Cruz Biotechnology, Dallas, TX, USA) were used to immunoprecipitate the immunocomplexes. The purified chromatin fragments were subjected to qPCR. The primers used in the ChIP experiments are listed in Supplementary Materials Table S1.

Protoplast Transfection Assays
To generate reporter and effector constructs, the 5 kb promoter sequence of ERF1 was amplified by PCR and cloned into the pGreenII 0800-LUC vector; the corresponding CDS sequences of PIF1, PIF3, PIF4, PIF5, and GFP were cloned into pGreenII 62-SK vectors. Protoplast isolation from Col-0 and PEG mediated transformation was performed as described [55], and a total of 11 ug plasmid (1 ug reporter plasmid, and 10 ug of each effector plasmid) was used to perform the transformation. To measure the LUC and Renilla luciferase activities, a Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA) was used, and the luciferase activities were analyzed by GloMax-96 Microplate Luminometer (Promega, Madison, WI, USA). All the experiments were performed at least three biological times. The primers used are listed in Supplementary Materials Table S1.