Ethylene-Inducible AP2/ERF Transcription Factor Promotes Capsaicinoid Biosynthesis in Capsicum

Ethylene is very important in the process of plant development and regulates the biosynthesis of a lot of secondary metabolites. In these regulatory mechanisms, transcription factors that mediate ethylene signals play a very important role. Capsaicinoids (CAPs) only synthesized and accumulated in Capsicum species, causing their fruit to have a special pungent taste, which can protect against attack from herbivores and pathogens. In this study, we identied the transcription factor CcERF2, which is induced by ethylene, and demonstrated its regulatory effect on CAPs biosynthesis. Transcriptome sequencing analysis revealed the expression patterns of CcERF2 and multiple genes associated to CAPs biosynthesis were basically the same. The spatiotemporal expression results showed CcERF2 was preferentially expressed in the placenta of the spicy fruit. Ethylene can induce the expression of CcERF2 and CAPs biosynthesis genes (CBGs). CcERF2 gene silenced, 1-MCP and PZA treatments caused a decrease in expression of CBGs and a sharp decrease in content of CAPs. The results indicated CcERF2 was associated with the CAPs biosynthesis. in the MYB transcription factor MYB31 the evolution of extremely pungent in the control. Relative contents of CaP, DhCaP and CAPs in the PZA treatment fruit at 29 DAP. Ethylene release rate in the PZA treatment fruit at 29 DAP. The experiments were replicated three biological times and three technical times. Data are expressed as the mean ± SD (n = 9). Student’s t test was used to identify signicant differences compared to the control (*P < 0.05, **P < 0.01).

FatA, ACL, ACS, CoMT, pAMT and CS (Abraham-Juárez et al, 2008;Liu et al, 2013). The biosynthetic mechanism of CAPs has been widely elucidated using bioinformatics analysis and multi-omics (Qin et al, 2014). As CAPs have considerable application value and commercial use, a lot of effect have been performed to enhance their content. However, their biosynthesis changed largely in space/time, and expression level of CBGs is precisely regulated at the transcriptional level. Manipulating the expression level of crucial CBGs seems to be determined the encreasing content of CAPs (Abraham-Juárez  . TF CaMYB108 activates the CBG promoters, especially those of COMT, pAMT, and KasI, and promotes enhanced expression of CBGs, thereby promoting the biosynthesis of CAPs . By directly binding and regulating the expression of CBGs, TF CaMYB48 participates in the biosynthesis of CAPs, but the transcriptional regulation of CAPs biosynthesis has not been fully clari ed (Sun et al, 2020). Therefore, it is necessary to identify TF associated with the biosynthesis of CAPs.
Ethylene plays a signi cantly role in plant life process (Hu et al, 2020). After synthesis of ethylene, it binds to the receptor ETR and transmits the signal to the nucleus through MAPKK and EIN2 (Gray 2004). The EIN2 binds to EIN3/EIL1 Andthe EIN3/EIL1 bind to the ERF1 promoter. ERF1 binds to the GCC-boxcontaining genes in the downstream promoter region speci cally and their expression can promote secondary metabolism product synthesis (Fujimoto et al, 2000;Brown et al, 2003;Paul et al, 2019). For example, in tobacco, ERF189 and ERF163 can bind to the GCC-box in the promoter region of the tobacco nicotine synthesis-related gene PMT2 speci cally, and directly promote synthesis of tobacco alkaloids (Shoji et al, 2010). The B3 subfamily proteins ORCA3 and ORA59 of the ERF family are two typical TFs associated with the regulation of secondary metabolites. The ORCA3 can upregulate the expression level of indole alkaloid synthesis-related genes in terpenoids, and promote the terpenoid indole biosynthesis (van der Fits and Memelink 2000, 2001). Lithospermum erythrophyllum LeERF-1 affects the secondary metabolites of shikonin positively through a mechanism similar to that of ORCA3 in affecting secondary metabolites (Zhang et al, 2011). In Catharanthus roseus, CrERF5 upregulates the biosynthesis and accumulation of bisindole alkaloids (Pan et al, 2019).
The TFs and key enzymes in ethylene synthesis and its signal transduction can promote the biosynthesis of secondary metabolites. In many cases (plants, cells, and hairy roots), the addition of ethylene, addition of 1-aminocyclopropane 1-carboxylic acid (ACC), and overexpression of related TFS can increase the biosynthesis of secondary metabolites (Gantet and Memelink 2002). For example, adding 0.018 mL/L of ethylene to yew cell culture can greatly increase production of taxane (Pan et al, 2000); adding ACC to Arabidopsis can promote synthesis of avonoids (Buer et al, 2006); adding ethylene or treating cypress cells (Platycladns orientalis) with ACC can promote the biosynthesis of β-rubelin (Zhao et al, 2004); and overexpression of ORC1/ERF221 can upregulate the synthesis of pyridine alkaloids (De Boer et al, 2011).
In this paper, the ethylene-induced AP2/ERF TFs CcERF2, which was particularly expressed in the placenta was identi ed. We con rmed that CcERF2 was associated with the production of CAPs biosynthesis.

Experimental materials
The C. chinense inbred line SL08 has a high CAPs content and was derived from Shuan La, the hottest pepper genotype in China (Deng et al, 2009). The C. annuum inbred line H19 has a low CAPs content, and was derived from Xiangtan Chi Ban Jiao. Pepper seeds germinated in the soil of cell plastic ats in complete darkness at 28°C. The seedlings were grown in a greenhouse of Yunnan Agricultural University on campus under normal conditions.

Extraction and detection of CaP and DhCaP
The CaP and DhCaP were extracted and detected according to our method (Deng et al, 2009

Exogenous substance treatments
Fruit at 24 DAP was used to investigate the effect of different exogenous substances on expression level of CcERF2 and CBGs after 0.5, 1.0, 1.5, 3.5, 5.5, 8.5, and 11.5 h. Treated with exogenous substances, the 24-DAP fruits of SL08 were soaked in sterile water with 0.01 g/L 6-benzylaminopurine (6-BA), 20 g/L PEG6000, 30% H 2 O 2 , 0.17 g/L gibberellic acid (GA 3 ), 0.14 g/L salicylic acid (SA), 17.53 g/L NaCl, 90 g/L glutamate (Glu), or 0.1 mmol/L ethephon (with sterile water as control) under normal conditions. After treatment, all fruits were frozen in liquid nitrogen and stored at -80°C immediately for gene expression analysis.

CcERF2 cloning and bioinformatics analysis
The open reading frame of CcERF2 was cloned according to Deng et al (2012). The primers used in the experiment are shown in Table S1. The BLAST-protein-nucleic acid (BLASTP) analysis was performed on CcERF2 through a database (www.ncbi.nlm.nih.gov); PSORT Prediction was performed for subcellular location prediction; and MEME was used to analyze the amino acid sequence motif. The Portparam tool was used to predict the physical and chemical properties of the protein. SignalP-5.0, TMHMM, and Nepos were used for signal peptide, transmembrane structure, and phosphorylation site of the protein prediction, respectively; SOPMA was used to predict its secondary structure; and ClustalX was used for sequence alignment and phylogenetic analysis. The MEGA 6.0 software was performed to construct a phylogenetic tree based on the amino acid sequence through the neighbor joining method, and the bootstrap method was used to evaluate the reliability of each node in the tree, repeated 1,000 times.

Subcellular localization and transcriptional activation analysis
Subcellular localization and transcriptional activation analysis were performed according to Sun et al (2020) and Zhu et al (2019). The full-length coding sequences of CcERF2 was cloned into the pAN580 (green uorescent protein, GFP) vector and fused to the N-terminus of GFP under the control of the CaMV 35S promoter. The constructs were separately introduced into tobacco protoplasts for transient expression, and the GFP uorescence signals were detected using a Zeiss lsm710 confocal laser scanning microscope (Carl Zeiss, Inc, Jena, Germany)

Gene expression analysis
The RNAiso Plus (Takara, Dalian, P R China) was used to extract total RNA from pepper fruit. The synthesis of cDNA rst strand refers to the instructions of the High Fidelity PrimeScript ® RT-PCR Kit (Takara). According to the sequence of related genes revealed by transcriptome sequencing data, speci c primers were respectively designed, and the internal reference was the β-ACTIN gene. Quantitative realtime PCR (qRT-PCR) was performed according to the method provided by the SYBR ® Premix Ex TaqTM II (Tli RNaseH Plus) kit (Takara). The primers used in this study are shown in Table S1.

VIGS analysis
A fragment of the CcERF2 coding sequence (CDS) with low similarity to other genes was cloned into pTRV2 and generated the silencing vector pTRV2-CcERF2. The VIGS were carried as reported (Zhu et al, 2019). In short, the pTRV2-CcERF2 and pTVR1 vectors were co-injected into the cotyledon stage seedlings of line SL08. The empty vectors pTRV2 and pTVR1 were co-in ltrated as a control, and pTRV2-PDS and pTVR1 were co-in ltrated as a technical control. The RNA isolated from 24-DAP fruits was used for expression analysis, and 34-DAP fruits were used for CaP and DhCaP measurement. The primer information used in this study is shown in Table S1. The SL08 pepper fruit of 24DAP was used to study the effects of 1-MCP and PZA on the expression of CBGs and the content of CAPs. 1-MCP treatment: 1 μL L − 1 1-MCP fumigated for 12 hours and played in a dark environment with a temperature of 20±2°C, a humidity of 85±5%. PZA treatment: Spray 100 mM PZA on the surface of the fruit until there are droplets, and played in a dark environment with a temperature of 20±2°C, a humidity of 85±5%. The fruits after 5 days of treatment were stored in the liquid nitrogen used for subsequent analysis. All experiments were repeated 3 times.

Ethylene release rate determination
Ethylene release rate was investigated according to the method of Xu et al (2018).

Statistical analyses
Each experiment contained three biological replicates and three technical replicates. The error bars mean the standard errors. One-way ANOVA was performed to identify signi cant differences. The relative gene expression was calculated using the 2 −ΔΔCT method (Livak 2001). Without induction at the 0 h time point was used as control expression.

Accumulation of CAPs
Accumulation of CaP and DhCaP was detected in both lines, but the content of CaP was much higher (about 3-4 times) than that of DhCaP. The CaP and DhCaP contents were much higher in the placenta of SL08 than in H19 ( Figure S1). Both CaP and DhCaP began to accumulate at 9 DAP in the placenta of SL08, reached a peak at 44 DAP, and then began to decline. Both CaP and DhCaP began to accumulate at 11 DAP in the placenta of H19, and began to decrease at 49 DAP. The CaP, DhCaP, and total CAPs contents in the placenta of SL08 were 44.4, 63.1, and 48.0 times those in H19, respectively.

Expression of CcERF2 and CBGs is consistent with CAPs biosynthesis
Expression patterns of CBGs during development stages of the placenta were analyzed using transcriptome sequencing data of the placenta and pericarp of line SL08 during the early stage of the laboratory study ( Figure 1). Expressions of CcPAL, CcCa4H, Cc4CL, CcCoMT, CcpAMT, CcCS, CcBCAT, CcKAS, CcACL, CcFAT,and CcACS genes were positively associated with accumulation of CAPs in SL08. The expression of these genes (except CcACL) showed a low expression level in fruit at 4 DAP, increased, and then decreased. Gene CcCS (Capana05g000531) was not expressed in 4-DAP fruit, and reached a maximum in 24-DAP fruit. Expression of CcCoMT (Capana03g001811), CATCcBCAT (Capana04g000751), CcKAS (Capana01g000111), and CcFATA (Capana06g000197) reached their highest values in 14-DAP fruit. Expression of CcCa4H (Capana06g000272) and CcpAMT(Capana10g001341) genes reached their maximum in 34-DAP fruit. Expression of CcPAL (Capana09g002199) reached its highest value in 44-DAP fruit.
Based on the transcriptome sequencing data of lines SL08 and H19 at different developmental stages of placenta completed in the laboratory, combined with the changes in the CaP and DhCaP contents of the placenta of both lines ( Figure S1), the AP2/ERF TF CcERF2 (Capana01g000677) was identi ed. Its expression pattern was similar to that of CBGs ( Figure 1). Expression of CcERF2 was almost undetectable in the early stage of placenta (4 DAP), then rose rapidly, reached a maximum at 34 DAP, and then decreased.
We measured the expression level of CcERF2 in some tissues of line SL08 and found that it was mainly expressed in the placenta (Figure 2A). It also showed considerable level of expression in seeds and pericarp, meaning that it has a role in these tissues. We also investigated CcERF2 expression in the placenta of both lines, and found signi cantly higher expression in SL08 than in H19 ( Figure 2B). Expression of CcERF2 in the placenta has the same pattern with the content of CaP and DhCaP. Thus, our results indicated that CcERF2 may play a vital role in of CAPs biosynthesis.

CcERF2 gene cloning and bioinformatics analysis
Based on our transcriptome sequencing data of SL08 and the pepper genome data in the public database, speci c primers were designed to clone the full-length CDS of CcERF2 from SL08. Nucleotide sequence analysis showed that CcERF2 length was 795 bp. The molecular formula of CcERF2 is C 1328 H 2050 N 376 O 409 S 9 , molecular weight is 30,115.70 D, and theoretical isoelectric point is 5.74. It is a fatsoluble, hydrophilic, and unstable protein. CcERF2 had no signal peptide and no transmembrane structure, and is located in the nucleus (88.8% probability).
The protein domain prediction results showed that CcERF2 belonged to the AP2 superfamily, and the conservative amino acid sequence position was 74-131 (LYRGIRQRPWGKWAAEIRDPRKGVRVWLGTFNTAEEAARAYDKEARKIRGEKAKVNFP) ( Figure 3C). This domain speci cally bound to the 11-bp GCC box of the ethylene response element and was essential for the ethylene response. Based on BLAST, 48 amino acid sequences including CcERF2 were obtained, and 43 motifs were obtained after motif signi cance test and analysis. On the whole, the predicted motifs differed within the same family, but the conserved motifs in the same subgroup were almost the same; that is, the closer the related species, the more similar were the motifs. For peppers, they all contained 13 conserved motifs: 1-8, 11, 12, 17, 19, and 22. Of these, motifs 1, 3, and 5 covered the AP2 conserved structure domain ( Figure 3B). The secondary structure of CcERF2 protein contained 73 alpha helices (27.56%), 22 extension chains (8.42%), and 9 beta turns (3.41%). There were 160 random coils (60.61%). The evolutionary relationship of amino acid sequences showed that the selected ERF sequences were grouped together in different families. This meant that they were conserved in the evolutionary relationship. In genetic relationship, CcERF2 protein was the closest to Solanaceae. In the Solanaceae, it was closest to Capsicum and Solanum, and most distant to Petunia ( Figure 3A). The phosphorylation sites of CcERF2 protein were predicted, and 19 amino acid residue types were screened: six types of T, three types of Y, and 10 types of S. CcERF2 was localized to the cell nucleus ( Figure 4A) and the CcERF2 protein has strong activation activity ( Figure 4B).

CcERF2 is a TF induced by ethylene
. Exogenous substances were applied to the 24-DAP fruit of SL08 to test the regulation model. Expression of CcERF2 in placental tissues was analyzed using qRT-PCR ( Figure 5).
The fruits were treated with ethephon, which signi cantly induced CcERF2 expression. Compared with the control, CcERF2 expression increased by 64 times within 11.5 h of applying ethephon. At the same time, the effects of exogenous 6-BA, PEG, H 2 O 2 , GA 3 , SA, NaCl, and Glu on CcERF2 expression were also tested.
Of these, 6-BA, H 2 O 2 , GA 3 , SA, and NaCl signi cantly induced CcERF2 expression. Compared with the control, these exogenous substances increased CcERF2 expression; however, the modes of induced expression were not the same. In the test time range, under the 6-BA, H 2 O 2 , and NaCl treatments, the CcERF2 expression showed a trend of initial increase and then decrease; however, under GA 3 and SA treatment, CcERF2 expression showed an increasing trend. Treatment with Glu had no signi cant effect on CcERF2 expression and PEG treatment was inhibitory.

Effect of ethephon on expression of CBGs
Ethephon treatment signi cantly increased expression of CBGs, but the induction effect was inconsistent. Compared to controls, within 11.5 h of applying ethephon, expression levels of CcPAL, CcCoMT, and CcFAT were 2.3-5.0 times; of CcCa4H, CcpAmt, CcBCAT, CcACL,and CcACS were 5.8-8.3 times; of CcKAS gene was 11.1 times; and of CcCS was more than 100 times. Considering that ethephon signi cantly induces the expression of CcERF2, our results showed that CcERF2 was a key factor in ethylene-mediated biosynthesis of CAPs ( Figure 6).

The effect of CcERF2-silenced on CAPs content and metabolic pathways
The VIGS vector pTRV2-CcERF2 was constructed using a vector derived from tobacco rattle virus (TRV), and a CcERF2-silenced experiment was performed. The 24-DAP fruits of SL08 were used to study. It was demonstrated that Agrobacterium infection with an empty pTRV2 vector resulted indistinctive changes in the expression level of CBGs in pepper fruits (Abraham-Juárez et al, 2008). Compared with empty vector plants, the fruits infected with the pTRV2-CcERF2 construct resulted a distinctive reduction in CcERF2 expression (22.4% of the empty vector) ( Figure 7A). Consistent with the expression level of CcERF2, CBGs in silent plants were also signi cantly downregulated. In the CcERF2-silent placenta, the expression of genes derived from fatty acid metabolism pathways (such as CcKAS, CcACL, CcFAT and CcACS) only slightly changed. However, expression of genes derived from the phenylpropane pathway (such as CcPAL, CcCa4H, Cc4CL, CcComt, and CcpAmt) and CcCS underwent signi cant changes, and the genes whose expression was altered were consistent with the transcription level of CcERF2 ( Figure 7A).
The CAPs contents of the pepper fruits infected with the empty pTRV2 vector were similar to those of uninfected plants (Abraham-Juárez et al, 2008). The 34-DAP fruits were uesed to study the effect of CcERF2 silencing on the CaP and DhCaP contents. Comparing with fruits infected with empty pTRV2, the CaP and DhCaP contents in fruits infected with pTRV2-CcERF2 were signi cantly reduced by 74.2% and 73.0%, respectively ( Figure 7B). The above results strongly supported that CcERF2 regulated certain CBGs to control CAPs biosynthesis.

Effects of 1-MCP on CBGs expression and CAPs contents
The effect of blocking ethylene signal transmission by 1-MCP on CBGs expression and CAPs contents had been studied. The results showed that 1-MCP treatment could remarkably reduce the expression of PAL, C4H, COMT, pAMT and CS genes ( Figure 8A). 1-MCP treatment signi cantly reduced the content of CAPs ( Figure 8B). The results indicated that the ethylene signal transduction was involved in the regulation of capsaicinoids biosynthesis.

Effects of PZA on CBGs expression and CAPs contents
The results indicated that after ethylene biosynthesis inhibitor PZA treated pepper fruits for 12 h, the expression of ACC oxidase-2 (ACO2) gene was extremely reduced ( Figure 8C); The endogenous ethylene release rate was conspicuouse reduced ( Figure 8E); The expression of CBGs such as PAL, C4H, COMT, pAMT and CS was inhibited ( Figure 8C); The CAPs contents were also obviously reduced ( Figure 8D). It indicated that the content of endogenous ethylene signi cantly affects the biosynthesis of capsaicinoids.

Discussion
Plant secondary metabolism comprises the life activity, formed in the long-term evolutionary process. The secondary metabolites are not only associated with the regulation of plant growth, but also allow plants to effectively deal with all kinds of stresses, such as fungi, pests, and herbivores (Dangl and Jones 2001). The distribution of secondary metabolites is usually speci c to species, organs, and tissues; CAPs are unique synthesized in the fruit placenta of Capsicum and have obvious speci city (Mazourek et al, 2009). A variety of stress conditions (such as high temperature, hydropenia, and herbivore invasion) can induce the production of secondary metabolites, and this process is affected by a lot of plant endogenous hormones (such as ethylene). Studies have found that low temperature promotes the accumulation of CAPs in pepper fruits ( We investigated the mechanism of ethylene on the expression of CcERF2 and CBGs. After fruits of line SL08 were treated with ethephon, expressions of CcERF2 and CBGs were signi cantly increased, equivalent to 2.1-110.3 times that of the control. However, the response pattern of each gene induced by ethephon slightly differed ( Figure 5) . These AP2/ERF factors seem to play a similar role in regulation metabolic genes (Hu et al, 2020). In this paper, the AP2/ERF TF CcERF2, which can be used as a transcription activator to regulate CAPs biosynthesis was identi ed. There were 43 motifs obtained by analyzing the motif signi cance of the CcERF2 protein. On the whole, the predicted motifs differed within the same family, but the conserved motifs in the same subgroups were almost the same. That is, the closer the related species, the more similar were the motifs. The peppers all contained 13 conserved motifs (1-8, 11, 12, 17, 19, and 22), of which motifs 1, 3, and 5 covered the conserved AP2 structural domain. It was reported that this type of TF can regulate the synthesis of plant secondary metabolites (Paul et al, 2019). Our results showed that the function of CcERF2 was convergent and divergent among different plant species.
Various genes related to CAPs biosynthesis have been studied (Aluru et al, 2003). However, very few TFs associated to CAPs biosynthesis have been isolated and characterized, except for some MYB TFs . We need to identify some TFs, especially AP2/ERF family TFs, which can be used as positive regulators to promote accumulation of CAPs. In the current study, based on transcription data, the AP2/ERF TF CcERF2 was selected for further analysis, because CcERF2 had a similar expression pattern to the CBGs in the transcriptome data. Silencing CcERF2 downregulated the expression of CBGs, especially the expression level of genes associated with the phenylpropane pathway, and therefore reduced its CAPs content. The results showed that CcERF2 was involved in regulating this metabolic process.   Figure 1 Heat map indicatingthe CBGs and CcERF2 expression patterns in fruit of inbred line SL08. The heatmap was generated by transcriptome sequencing data with R software packages. The data were normalized in each row, and the numbers on the right side of the gure indicate the gene expression level corresponding to color.

Figure 2
Differential expression assays of CcERF2 (A) Differential expression assays of CcERF2 in different tissues of pepper inbred line SL08; Roots, stems, leaves and owers were collected from 30-day-old seedlings. Fruits at 24 DAP was collected and was divided into pericarps, placentas and seeds. The relative expression of the root was set to 1, and that of all the other tissues was measured relative to that of the root. (B) Differential expression assays of CcERF2 in pepper inbred line SL08 and H19 fruits. Fruits at 24 DAP was collected to investigate the CcERF2 expression. The relative expression of the SL08 was set to 1, and that of H19 was measured relative to that of the SL08. The experiments were replicated three biological times and three technical times. Data are expressed as the mean ± SD (n = 9). Student's t test was used to identify signi cant differences compared to the control (*P < 0.05, **P < 0.01).   Effect of elicitors (Eth, 6-BA, PEG, H2O2, GA3, SA, NaCl and Glu) on CcERF2 expression. The 24 DAP pepper fruits were treated with elicitors (Eth, 6-BA, PEG, H2O2, GA3, SA, NaCl or Glu), and the fruit were sampled after elicitation for 0.5, 1.0, 1.5, 3.5, 5.5, 8.5, and 11.5 h. Without induction at the 0 h time point was used as control expression. The experiments were replicated three biological times and three technical times. Data are expressed as the mean ± SD (n = 9). Student's t test was used to identify signi cant differences compared to the control (*P < 0.05, **P < 0.01).

Figure 6
Expression pro les of CBGs after treatment with Eth. The 24 DAP pepper fruits were treated with Eth, and the fruit were sampled after elicitation for 0.5, 1.0, 1.5, 3.5, 5.5, 8.5, and 11.5 h. Without induction at the 0 h time point was used as control expression. The experiments were replicated three biological times and three technical times. Data are expressed as the mean ± SD (n = 9). Student's t test was used to identify signi cant differences compared to the control (*P < 0.05, **P < 0.01).

Figure 7
Expression level of CBGs and CcERF2, relative contents of CaP, DhCaP and CAPs in the control (empty vector) and CcERF2-silenced fruit. (A) Expression level of CBGs and CcERF2 in the control (empty vector) and CcERF2-silenced fruit. Fruits at 24 DAP was sampled to perform the expression level of CBGs and CcERF2. The relative expression level of the CBGs and CcERF2 in the control were set to 1, and that of in the CcERF2-silenced fruit were measured relative to that in the control. (B) Relative contents of CaP, DhCaP and CAPs in the control (empty vector) and CcERF2-silenced fruit at 34 DAP. The experiments were replicated three biological times and three technical times. Data are expressed as the mean ± SD (n = 9). Student's t test was used to identify signi cant differences compared to the control (*P < 0.05, **P < 0.01). and CcERF2 in the control were set to 1, and that of in PZA treatment fruit were measured relative to that