Identification and Analysis of the AP2 Subfamily Transcription Factors in the Pecan (Carya illinoinensis)

The AP2 transcriptional factors (TFs) belong to the APETALA2/ ethylene-responsive factor (AP2/ERF) superfamily and regulate various biological processes of plant growth and development, as well as response to biotic and abiotic stresses. However, genome-wide research on the AP2 subfamily TFs in the pecan (Carya illinoinensis) is rarely reported. In this paper, we identify 30 AP2 subfamily genes from pecans through a genome-wide search, and they were unevenly distributed on the pecan chromosomes. Then, a phylogenetic tree, gene structure and conserved motifs were further analyzed. The 30 AP2 genes were divided into euAP2, euANT and basalANT three clades. Moreover, the cis-acting elements analysis showed many light responsive elements, plant hormone-responsive elements and abiotic stress responsive elements are found in CiAP2 promoters. Furthermore, a qPCR analysis showed that genes clustered together usually shared similar expression patterns in euAP2 and basalANT clades, while the expression pattern in the euANT clade varied greatly. In developing pecan fruits, CiAP2-5, CiANT1 and CiANT2 shared similar expression patterns, and their expression levels decreased with fruit development. CiANT5 displayed the highest expression levels in developing fruits. The subcellular localization and transcriptional activation activity assay demonstrated that CiANT5 is located in the nucleus and functions as a transcription factor with transcriptional activation activity. These results help to comprehensively understand the pecan AP2 subfamily TFs and lay the foundation for further functional research on pecan AP2 family genes.


Introduction
The APETALA2/ethylene-responsive factor (AP2/ERF) is a superfamily of transcription factors (TFs), which were characterized by conserved AP2 domains. AP2/ERF superfamily TFs are mainly divided into the ERF (ethylene-responsive-element-binding protein), DREB (dehydration responsive element binding), AP2 (APETALA2), RAV (Related to ABI3/VP) and Soloists five major subfamilies [1,2]. Both the ERF and DREB subfamily members only contain one AP2 domain, and both are further subdivided into six subgroups [3]. The AP2 subfamily members containing two consecutive AP2 domains are further subdivided into euAP2, basalANT and euANT clades depending on the nuclear localization and amino acid sequences as well as other conserved motifs [4]. The RAV subfamily members contain a single AP2 domain and a B3 DNA-binding domain [5]. Although the Soloist subfamily with a single AP2 domain is displayed in most plants, its sequence and structure distinguish it from other AP2/ERF superfamily TFs [2,3]. The AP2/ERF domain consisted of a 60-70 amino acid functioning DNA binding domain; however, each of its subfamily recognized specific cis-acting elements. For example, ERF subfamily members usually bind to GCC-box responding to ethylene, pathogens and wounding [6], while DREB subfamily TFs usually bind to the DRE/CRT(A/GCCGAC) cis-element to activate drought and salt responsive genes [7]. The second AP2 domain in the AP2 subfamily binding to T/A-rich elements positively regulates floral organ identity [8], while in RAV1 it binds to the CAACA motif involved in ABA signaling [9,10]. Taken together, the diversification of gene structure and DNA-binding affinities may contribute to the different functions of the AP2/ERF superfamily members.
The AP2/ERF superfamily members have been reported to serve as regulators of stress responses and development processes, especially in reproductive organ development. APETALA2 (AP2), expressed in floral organs and developing ovules as well as nonfloral organs, was initially identified as controlling flower and seed development in Arabidopsis [11]. Further research revealed that AP2 could bind to the AT-rich target sequence of AGAMOUS (AG) and directly activate its expression through the second AP2 domain [8]. AINTEGUMENTA (ANT), which belongs to the euANT clade, regulated cell division and organ size. The ectopic expression of ANT enlarged embryonic and all shoot organs by increasing cell numbers in Arabidopsis [12,13]. Moreover, seven high sequence similarity genes, AINTEGUMENTA-like (AIL), were identified, and they primarily expressed in young actively dividing tissues. The expression analysis by in situ hybridization demonstrated that AIL5, AIL6 and AIL7 expressed distinct spatial and temporal patterns in inflorescence meristems and flowers [14]. The overexpression of AIL5 resulted in similar large floral patterns with ectopic expressions of ANT, while AIL6's functioned redundancy with ANT took place during flower development [14,15]. In rice, the AP2 factor gene RSR1 was identified as a negative regulator of type I starch syntheses genes, and the loss of function of RSR1 resulted in an altered amylopectin structure and increased gelatinization temperature [16]. Interestingly, the same genetic locus was identified as the candidate of major QTL qHD5, controlling the heading date in rice [17]. FRIZZY PANICLE (FZP) was required to establish rice floral meristem identity and determine grain size [18,19]. SUPERNUMERARY BRACT (SNB) encoding APETALA2-like TF controlled rice seed shattering and seed size. A point mutation in SNB resulted in alternative splicing and reduced shattering in its mutant lines [20]. EXCESSIVE NUMBEROF FLORAL ORGANS (ENO) was identified as a tomato fruit regulator and regulating WUSCHEL expression. The disruption of ENO expanded the expression domain of SlWUS in a flower-specific manner, resulting in enlarged fruit size [21]. These studies demonstrated that the AP2 family TFs participated in the regulation of development processes, such as flower development, heading date, spikelet development and seed development.
With advances in sequencing technology, whole genomic sequences of more plants are available. The AP2 family genes were genome-wide identified and analyzed in increasingly more plant species, such as barley [22], common wheat [23] and Dendrobium officinale [24], which contributed to the understanding of the molecular function of AP2 family genes. However, the AP2 family TFs have not been reported in pecan (Carya illinoinensis (Wangenh.) K. Koch), which is considered a famous nut and woody oil tree species rich in unsaturated fatty acids [25]. In this study, we identify 30 AP2 subfamily TFs from pecans through a genome-wide search and mapped them on pecan chromosomes. Then, we construct a phylogenetic tree, analyzed gene structure and conserved motifs as well as promoter cis-acting elements. Furthermore, a qPCR is conducted to analyze their expression pattern, then subcellular localization and transcriptional activation activity are investigated. These results enrich our understanding of the AP2 family in pecans and lay the foundation for further functional research on pecan AP2 family genes.

Identification of AP2 Subfamily Genes in the Pecan
To identify AP2/ERF family genes, an HMM search was employed to ascertain the genes containing the AP2 domain (Pfam ID:PF00847) from rice, Arabidopsis and pecan protein datasets. Then, 207 putative genes were retrieved from the pecan, and an evolutionary tree of AP2 family genes was initially constructed ( Figure S1). According to the number of AP2 domains contained, AP2/ERF family genes can be usually divided into ERF, AP2 and RAV subfamilies [4]. Furthermore, we found that both Arabidopsis AINTEGUMENTA (ANT) [26] and rice qHD5 [17], containing two AP2 domains, were grouped into the AP2 subfamily. In this study, we specially focused on the pecan AP2 subfamily genes and further checked the candidate sequences through Smart (http://smart.embl-heidelberg.de, accessed on 12 July 2021). Interestingly, although OF21942-RA, OF26251-RA, OF27743-RA and OF31010-RA were clustered into the AP2 subfamily ( Figure S1), they do not contain two AP2 domains. In this study, we preferentially focused on members containing two AP2 domains, and finally obtained 30 AP2 subfamily genes from pecan renamed from CiAP2-1 to CiAP2-7 and CiANT1 to CiANT23, depending on their structural characteristics (Table 1). Furthermore, we analyzed the genomic distribution of AP2 genes, and results demonstrated that CiAP2 genes were mainly distributed on chromosome 1, 6, 8 and 13 ( Figure 1). Unfortunately, CiANT13, 16, 19, 20 and 22 were not anchored to the corresponding chromosomes.

Phylogenetic Analysis of Pecan AP2 Subfamily Genes
To further reveal the evolutionary relationships among pecan AP2 subfamily proteins, we constructed a phylogenetic tree using the sequences of 30 pecan, 17 rice and 14 Arabidopsis AP2 subfamily proteins ( Figure 2). The results were credible with high bootstrap values and demonstrated that pecan AP2 subfamily proteins could also be classified into the euAP2, euANT and basalANT three subgroups consistent with the previous study on AP2 family genes in Arabidopsis and rice [4]. We noted that Arabidopsis ANT (AT4G37750) and rice qHD5 (Os05g03040) were clustered into euANT and euAP2, respectively. ANT is involved in ovule and female gametophyte development and controlled plant organ size by regulating growth and cell numbers during organogenesis in Arabidopsis [12,13]. qHD5 was reported as a candidate gene for heading date in rice [17]. Therefore, although all AP2 subfamily genes contain two AP2 domains, they may regulate different biological processes.   (14) and Arabidopsis (17) were adopted to construct the phylogenetic tree. The tree was constructed using FastTree according to the maximum likelihood method. The AP2 subfamily proteins were clustered into three groups: euAP2, euANT and basal ANT. Proteins from pecan, rice and Arabidopsis are marked by green, blue and red colors, respectively.

Conserved Motif Analysis in AP2 Proteins
The motifs in 30 AP2 protein sequences were analyzed through MEME. The results displayed the first ten conserved motifs, and they varied among euAP2, basalANT and euANT evolution branches ( Figure 3). In euAP2, motifs 1, 2, 3, 4, 5 and 6 were found almost in all members. Among them, motif 6 was located in the first AP2 domain and motifs 1 and 3 were parts of the first AP2 domain. Motifs 2 and 4 were located in the second AP2 domain and motif 5 was part of the second AP2 between two AP2. Compared with the euAP2 group, basalANT and euANT contained the unique motif 10 in the first AP2 domain. BasalANT contained the unique motifs 7 and 8 on the C terminal, while euANT contained the unique motif 9 on the N terminal. Overall, the whole AP2 domain is presented in the middle of the euAP2 group, while near the N terminal in the basalANT group and in a diverse position in the euANT group. Moreover, the gene structures were visualized to further reveal the variations between CiAP2 genes. As shown in Figure 3, the genes in the euANT group exhibited highly similar gene structures with 8 or 9 exons. Although the exon number ranged from 8 to 11 in euAP2 and 5 to 14 in the basalANT group, most genes showed similar gene structures in their respective groups except for CiAP2-2 and CiANT19, 21, 23.

cis-Element Analysis of AP2 Subfamily Genes
AP2 family genes have been demonstrated to contribute to the processes of plant growth and development as well as various environmental stimuli. Therefore, detecting the characteristic of the cis-element in the promoter will deepen our understanding of the expression pattern of pecan AP2 genes. We analyzed the promoter through the PlantCARE server and clustered all motifs into eight groups depending on the function annotation ( Figure 4). Evidently, promotor and enhancer elements accounted for the largest proportion, with TATA-box and CAAT-box occurring the most frequently. Light responsive elements and plant hormone-responsive elements ranked as the second and third largest proportion, respectively. Box4, G-box and GT1-motif were the top three in light responsive elements. We further predicted the plant hormone-responsive elements and found that abscisic acid, salicylic acid, MeJA, gibberellin and auxin responsive elements were identified in their promotors ( Figure S2). Among them, the abscisic acid-responsive element ABRE and MeJA-responsiveness CGTCA-motif appeared the most frequently. We also found that abiotic stress responsive elements, such as the MYB binding site (MBS), were involved in drought inducibility, which implied CiAP2 genes may participate in drought-stress response through the MYB transcription factor. Interestingly, we found tissue-specific elements were all cis-regulatory elements involved in seed-specific regulation, which was consistent with the function of AP2 genes participating in embryo development. Furthermore, we only identified one circadian-responsive and two-cell-cycle regulation elements in all AP2 subfamily genes. Overall, our results indicate that AP2 subfamily genes may participate in growth and development processes by responding to plant hormones, light and abiotic stress.

Expression Analysis of CiAP2 Genes
To investigate the putative roles of pecan AP2 subfamily genes, a qPCR assay was employed to analyze their expression patterns in different tissues. Among the euAP2 clade, CiAP2-5 had the highest relative expression level in leaves, stamens and pistils. The expression level of CiAP2-7 was second in stamens and pistils. CiAP2-1 and CiAP2-2 were clustered together and presented consistent expression patterns (Figure 5a). Genes in basalANT clades, such as CiANT15, 16 and 18, displayed similar expression patterns with relatively high expression levels in pistils, while CiANT17, 21, 22 and 23 showed higher expression levels in stamens, especially CiANT23 (Figure 5b). The expression pattern in the euANT clade varied greatly (Figure 5c,d). Although CiANT1, 5, 11 and 12 were clustered together, CiANT1 and 5 expressed higher, while CiANT11 and 12 showed almost no expression in the tested tissues. The expression levels of CiANT2 and CiANT6 were higher in different tissues, while CiANT3 displayed almost no expression in the tested tissues and CiANT4 showed trace expression in leaves. Previous studies have demonstrated that some AP2 family genes participate in seed development in Arabidopsis. To identify the potential genes that are involved in pecan fruit development, we selected genes with a relatively high expression in the pistil for further quantitative analysis in the developing pecan fruits. As the results show (Figure 5e), CiAP2-5, CiANT1 and 2 shared similar expression patterns in developing fruits, and their expression levels decreased with fruit development. The expression of CiANT6 was lowest in developing fruits. CiANT5 and CiANT23 displayed the highest expression levels in developing fruits; CiANT5 expressed higher in the first three weeks, while CiANT23 expressed higher in the last three weeks during pecan fruit development. The difference in the expression levels implied that CiANT5 and CiANT23 were likely to function diversely during pecan fruit development.

CiANT5 Function as a Transcription Activator in Yeast
As mentioned above, CiANT5 demonstrated a high expression level in pecan leaves, pistil and developing fruits. Therefore, we selected it for further analysis. To explore whether CiANT5 is located in the nucleus as expected of a transcription factor, we conducted a subcellular localization analysis first. We obtained the full-length CDS of CiANT5 without a stop code through PCR amplification and fused CiANT5 to the N terminal of GFP. Then, the CiANT5-GFP and nuclear maker Ghd7-mCherry were transiently co-expressed in tobacco leaves. The results demonstrated that the CiANT5-GFP fluorescence signals localized in the nucleus overlapped perfectly with the red fluorescence of the nuclear maker Ghd7-mCherry, while free GFPs were detected throughout the nucleus and cytoplasm (Figure 6a,b). Subsequently, the transcriptional activation activity was analyzed in yeast cells. The full-length of CiANT5 as well as seven truncated fragments were ligated into the pGBKT7 vector to generate BD-CiANT5(1-682aa), BD-N(1-324aa), BD-N1AP2(1-397aa), BD-∆C(1-491aa), BD-C(492-682aa), BD-C2AP2(428-682aa), BD-∆N(325-682aa) and BD-AP2(325-491aa) fusion constructions. These constructions as well as the empty pGBKT7 were transformed into yeast AH109 and cultured on selective medium SD/-Trp and SD/-Trp/-His/-Ade, respectively. As the result show, CiANT5 displayed transcriptional activation activity in yeast (Figure 6c). Among the six truncated fragments, except C2AP2, ∆N and AP2, others showed transcriptional activity. Furthermore, the results were confirmed by quantitatively measuring β-galactosidase (β-gal) activity. When the truncated fragments contained the AP2 domain, β-gal activity decreased and remained lower than the negative control in the BD-AP2 fragment. Taken together, these results demonstrate that CiANT5 is a transcription factor localized in the nucleus with transcriptional activity, while the AP2 domain has a negative effect on transcriptional activity.

Characterization of Pecan AP2 Subfamily Genes
In this study, we initially identified 207 AP2/ ERF family genes in the pecan, which were classified into three subfamilies, ERF, RAV and AP2, depending on the number of AP2 domains ( Figure S1). Of them, both ERF and RAV members contain a single AP2 domain, while members in the AP2 subfamily contain two tandem AP2 domains. Further, we verified the sequences and finally obtained 30 AP2 members with two AP2 domains, which was higher than that in Pinus massoniana (7) [27], Pineapple (24) [28], Dendrobium officinale (14) [24], while less than that in Elaeis guineensis (34) [29] and Saccharum spontaneum (43) [30]. Pecan AP2 members were also grouped into three clades, euAP2, euANT and basalANT, which is consistent with previous research reported in angiosperms, such as wheat and Dendrobium officinale [4,22,24]. Furthermore, the analysis of genomic distribution displayed that most genes were distributed on chromosomes 1, 6, 8 and 13 (Figure 1). Unfortunately, five genes in the basalANT clade, namely CiANT13, 16, 19, 20 and 22, failed to be anchored on corresponding chromosomes. Similarly, three genes were not mapped to any chromosomes in pineapple [28], perhaps due to the imprecise chromosome assembly.
Proteins in the same family usually share a common characterized domain; however, members may contain diverse domains and motifs contributing to a variety of functions. A motif analysis showed that the top ten conserved motifs in the pecan AP2 subfamily TFs, and motifs 1, 2, 3, 4, 5 and 6 were found in almost all members (Figure 3), while motifs 7 and 8 were specific to basalANT and motif 9 was specific to euANT. This difference may lead to a diversity of gene functions. As reported, miR172 binding to specific motif sites regulates euAP2 TF expression through transcript cleavage and translational repression [31,32]. In Arabidopsis, AP2 family members, such as AP2, TOE1, TOE2, TOE3, SMZ and SNZ, were miR172 target genes. AP2 participated in floral organ formation, while others mainly suppressed floral formation [31,33]. Moreover, the gene structure reflected gene function and evolution to some extent. Our results show that the euANT group exhibited a highly similar gene structure with 8 or 9 exons, and the basalANT group had more diversified structures with 5 to 14 exons, which is consistent with previous findings in wheat [23]. In short, our results implied that different conserved motifs and gene structures may result in a diversity of functions responding to evolutionary selection pressures.

Promoter cis-Acting Elements and Expression Analysis Implied the Functions of Pecan AP2 TFs
AP2/ERF, as a plant-specific TF family, has been proven to play important roles during plant growth, abiotic stress and hormone response [2,34]. As expected, we found many tissue-specific elements, abiotic stress-responsive elements and plant hormoneresponsive elements in the promoters of AP2 family genes (Figure 4). The tissue-specific elements mainly are the RY-element and AACA_motif involving seed and endospermspecific expression regulation. Previous studies have demonstrated that petunia AP2-like genes PhAp2A, PhAp2B and PhAp2C are strongly expressed in endosperm and function in flower and seed development [35]. Wang et al. identified 57 putative TFs during maize seed development by expressed sequence tag (EST) analysis and found that three AP2/ERF members, ZmAP2-1, ZmAP2-2 and ZmAP2-2, played important regulating roles in maize seed development [36]. In Brassica napus, BABY BOOM (BBM) belongs to the AP2/ERF family, preferentially expressed in developing embryos and seeds, and participates in cell proliferation and morphogenesis during embryogenesis [37]. Our results demonstrate that CiAP2-5, CiANT1, 2, 5 and 23 displayed relatively higher expression in pistils and developing pecan fruits ( Figure 5); they may participate in pecan fruit development. Besides, abiotic stress-responsive elements are another common element in the AP2 gene promoters, which may be related to the AP2 members regulating drought, heat, salt and freezing. Arabidopsis DREB1A/CBF1, DREB1B/CBF2 and DREB1C/CBF3 and their orthologous genes OsDREB1A and OsDREB1B in rice were involved in a lowtemperature and drought response [7,[38][39][40]. Zeng et al. [24] also found that the expression of DoAP2 genes was changed to different degrees by being treated with cold, PEG and NaCl. Furthermore, we found that many elements respond to hormones in the promoters of pecan AP2 genes, of which ABA-responsive elements (ABRE) were the largest group (Figures 4 and S2). This implied that the expression of CiAP2 genes may be induced by ABA and that they may participate in various stress regulatory networks. Interestingly, MeJA-responsive elements were the largest group in DoAP2 promoters [24]. This difference may be related to the optimum growing environment required by different species.

CiANT5 as a Member of AP2 TFs Dispayed Transcription Activity
Transcription factors can bind to specific DNA sequences and activate or repress the transcription of corresponding genes in the nucleus. Nuclear localization is the basic feature of transcription factors. In this study, we selected a homologous gene of Arabidopsis ANT, CiANT5, to conduct a subcellular localization assay. As expected, CiANT5, in the manner of SNB [20] and DoAP2-1, DoAP2-2, DoAP2-6 and DoAP2-11 [24], was part of the AP2 family TFs located in the nucleus (Figure 6b). Furthermore, we analyzed the transcriptional activity and found ANT5 functioned as a transcriptional activator in yeast. A truncated fragment analysis revealed that the activating domain was located in BD-N(1-324aa) and BD-C(492-682aa) (Figure 6c). While the two tandem AP2 or respective domains showed no activating function, perhaps they acted as binding domains bound to specific DNA motifs. Similarly, APETALA2-like TF SNB presented transcriptional activity in yeast, but the activating domain was located in residues 278 to 324 [20]. The C-terminal region (107-318) of FZP is indispensable for transcriptional activation activity by LUC activity in rice protoplasts [19]. These results implied the activating domain may be located in a diverse position in different AP2 TFs. Moreover, some AP2 TFs displayed transcriptional inhibitory activity. Arabidopsis AP2 negatively regulated the homeotic gene AGAMOUS, which is involved in floral organ development [41]. DoAP2-1, DoAP2-2, DoAP2-6 and DoAP2-11 displayed transcriptional repression activity in tobacco leaves [24]. Whether transcriptional activation or repression, both are the precise regulation of TFs to its target genes. Thus, the related transcriptional characteristics of the remaining pecan AP2 TFs still need additional research.

Identification of AP2 Subfamily Genes in Pecan
The genome sequences of the pecan were derived from our re-sequenced data. The AP2 family sequences of Arabidopsis and rice were downloaded from Phytozome (https://phytozome.jgi.doe.gov/pz/portal.html, accessed on 25 March 2021). The Hidden Markov Model of AP2 was downloaded from Pfam (http://pfam.xfam.org/, accessed on 27 March 2021), according to the Pfam id (PF00847). The HMM search from the HMMER program (v3.2.1) was employed to ascertain the presence of the AP2 domain in these three species. Furthermore, all candidate sequences were checked by Smart (http://smart.embl-heidelberg.de, accessed on 12 July 2021) to remove sequences without two AP2 domains and redundant sequences. The amino acid sequences of AP2 subfamily members are listed in Supplementary File S1.

Gene Structure and Conserved Motif Analysis
According to the annotation files, the coding sequences and corresponding genomic sequences were extracted from the pecan genome. The conserved motif structure of AP2 subfamily genes was found by MEME (version 5.0.2) [46]. Then, all data were input into tbtools to visualize the gene structure and conserved motif.

cis-Element Analysis
A total of 2.0 Kb upstream sequences from initiation codon were retrieved as promoter sequences, and then the cis-regulatory elements were predicted through the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 22 August 2021). Furthermore, we clustered all motifs into different groups depending on the function annotation and counted the frequency distribution of promotor and enhancer elements, protein-binding sites, light responsive elements, tissue specific elements, abiotic stress responsive elements, circadian responsive elements, cell cycle regulation and plant hormone responsive elements.

Plant Material and Sample Collection
The pecan cultivar Western used as plant material was grown in Lin'an City (30 • N, 119 • E), Zhejiang Province, China. The leaves, stems, stamens and pistils were collected from Western in late April 2021, when the stigmas were acceptable. Moreover, we collected developing fruits one to ten weeks after controlling for pollination. All samples were immerged in liquid nitrogen immediately after collection and stored at −80 • C in a refrigerator for its use.

RNA Isolation and qPCR Analysis
The RNA was isolated from pecan samples (leave, stems, stamens and pistils, as well as developing fruits) with modified CTAB combining TRIzol reagent (Invitrogen, Grand Island, NY, USA), as we used previously in [47]. The cDNA was synthesized using the Prime-Script™ RT Reagent Kit (TaKaRa, Kyoto, Japan), according to the manufacturer's instructions.
qPCR was performed on a Bio-Rad CFX 92 Touch device using SYBR Green PCR Kit (TaKaRa). The procedure was as follows: 95 • C, 90 s; 40 cycles of 95 • C, 5 s; 55 • C, 20 s; 72 • C, 10 s. The relative expression level was calculated using the 2 −∆∆Ct method. The pecan actin gene (OF23284) was used as the internal reference gene. All primers used for the qPCR analysis are listed in supplementary Table S1.

Subcellular Localization Analysis
The full-length CDS of CiANT5 without a stop code was amplified and inserted into the vector pYBA1132 between the XbaI and HindIII sites to form the fusion construct CiANT5-GFP. Ghd7-mCherry was used as nuclear maker. Different combinations of fusion constructs were co-transformed in N. benthamiana leaves. The EGFP signals were detected with a laser scanning confocal microscope (LSM880, Karl Zeiss, Oberkochen, Germany) 48 h after transfection. The primers used in the analysis are listed in Supplementary Table S1.

Transcriptional Activity Analysis
The transcriptional activity assay was performed using the Matchmaker GAL4 Yeast Two-Hybrid System 3 (Clontech, CA, USA). The full-length CDS and seven CiANT5 cDNA fragments were amplified and fused with pGBKT7 plasmid linearized by EcoRI. Empty pGBKT7 and fused plasmids, which were marked as BD-ANT5, BD-N, BD-N1AP2, BD-∆C, BD-C, BD-C2AP2, BD-∆N and BD-AP2 were transformed into the yeast strain AH109. All transformants were grown in a tryptophan-negative dropout medium (SD/-Trp), then incubated in selective media plates (SD/-Trp/-His/-Ade) at 30 • C for four days to assay transcriptional activity. Furthermore, chlorophenol red-β-D-galactopyranoside (CPRG, Roche, Basel, Switzerland) was used as a substrate to measure β-galactosidase activity via liquid culture assays, as we did previously in [48]. All the primers used for the vector construction are listed in supplementary Table S1.