Identification and characterization of two APETALA2 homolog genes in lotus (Nelumbo nucifera) involved in sepal and petal development Authors: Liu Xuelian, Qin Yu, Cao Dingding, Deng Jiao, Chen Songbiao, Yang Pingfang, Lin Zhongyuan Published: 2024 ABSTRACT AbstractBackgroundLotus (Nelumbo nucifera) is a significant aquatic ornamental genus widely utilized in horticulture for its decorative, culinary, medicinal, and other practical uses. It presents a variety of flower shapes, including few-petalled, semi-double-petalled, double-petalled and thousand-petalled flowers, making it an ideal candidate for studying the flower development of aquatic plants. However, the molecular mechanism of floral development in lotus remains elusive.ResultsIn this study, two APETALA2 (AP2) homologues, NnAP2a and NnAP2b, were identified in lotus. Interestingly, both NnAP2a and NnAP2b proteins contained two conserved AP2 domains and were verified to be located primarily in the nucleus. Both NnAP2a and NnAP2b showed high expression levels in the floral buds and petals. Ectopic expression of NnAP2a and NnAP2b in Arabidopsis led to an increase in the number of petals and sepals compared to the wild type (WT). Meanwhile, each of the two NnAP2 genes was able to rescue the sepal and petal defective phenotype of the ap2-6 mutant in Arabidopsis. Furthermore, protein–protein interaction assays indicated that NnAP2s could form a protein complex with other proteins involved in floral organ development, such as AP3, PISTILLATA (PI), and SEPALLATA3 (SEP3).ConclusionsThese results suggest that NnAP2s could influence sepal and petal development in N. nucifera. Our findings not only provide some insights into molecular mechanism underlying sepal and petal development and formation of lotus, but also might help its breeding in improving flower morphology.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12870-024-05923-7.Keywords: Nelumbo nucifera, Flower development, AP2, Sepal, Petal Background Lotus (Nelumbo nucifera) is a significant aquatic ornamental plant. It belongs to the Nelumbonaceae family and the Nelumbo genus. Through evolutionary and phylogenetic research, it has been suggested that the lotus is one of the earliest eudicots, as evidenced by its slow branching and retention of primitive morphological characteristics similar to monocotyledons [1–3]. There are two species within the Nelumbo genus: Asian lotus (N. nucefera Gaertn.) and American lotus (N. lutea Willd.) [4]. The Asian lotus is primarily found in China and holds a significant cultural importance and history of cultivation [5, 6]. There are three distinct types of lotus: flowering lotus, seed lotus, and rhizome lotus. Among them, the flowering lotus is the most diverse and valued for its ornamental appeal [6]. As an emerging horticultural model plant, lotus is attracting more and more attention from the scientific community [7]. Various types of lotus flowers, including semi-double-petaled, double-petaled, and thousand-petalled flowers have been progressively appearing [7]. The double flower phenotype, known for its ornamental beauty, is actively sought after by breeders in numerous species. Previous studies have tackled the underlying molecular mechanisms associated with double flower formation. The ABC model was originally proposed following the study of various types of floral organ homozygous mutants in Arabidopsis thaliana and Antirrhinum majus [8]. The molecular mechanisms of floral organ development were elucidated, showing that floral organ formation is controlled by homozygous heterozygous genes belonging to three distinct classes, namely Class A, Class B, and Class C. Then, the ABCDE model was further defined with the identification of Class D and E homozygous and heterozygous genes [9–11]. Specifically, Sepal formation is controlled by Class A and E genes, petal development by the A, B, and E genes, stamen formation by the B, C, and E genes, and carpel formation by the C and E genes. APETALA2 (AP2) is a Class A group gene involved in sepals and petals development in first two whorls of a flower. It antagonizes the Class C gene (AGAMOUS, AG) during flower development [12]. Thus, their expressions are restricted in floral organs [13]. Belonging to the AP2 subfamily of the APETALA2/ethylene-responsive element binding factor (AP2/ERF) superfamily, AP2 is known to play very important roles in diverse processes of plant development [14]. For instance, AP2 controlled seed size by influencing embryo, endosperm, and seed coat development [15, 16]. The regulation of AP2 by miR172 affect on inflorescence meristem size, floral organ identity, and the control of flowering time, servs as floral repressors [17–19]. Most of the studies on AP2 function were focused on its role in floral development. In Arabidopsis, the AP2 gene was found to be required early in flower development to direct the first and second whorl of floral primordia that develop as sepals and petals [20]. A lack of AP2 function can result in stamens replacing petals, while the overexpression of AP2 leads to the production of double flowers [21, 22]. For example, 35S::AP2m1 flower with numerous petals and loss of floral determinacy were found [19]. The function of AP2 homologues also has been identified in many other species, such as Rosa chinensis [23], Dianthus chinensis [24], Hordeum vulgare [25, 26], and Prunus mume [27]. In addition, AP2 co-regulates the expression of multiple downstream genes through the recruitment of synergistic repressors, which ultimately control floral formation, especially floral homeotic genes. AP2 is involved in regulating the expression of class B genes APETALA3 (AP3) and PISTILLATA (PI) and the class E gene SEPALLATA3 (SEP3) at the floral organ boundary [28]. Moreover, AP2 acts as a repressor of AG, and binds an AT-rich sequence element in AG [29, 30]. Our previous studies revealed that AP2-like gene may be involved in stamen petaloidy and carpel petaloidy [31, 32]. However, the specific regulatory function of AP2 in floral development in Nelumbo nucifera remains unknown. Therefore, in this study, we isolated and identified two AP2 orthologs, NnAP2a and NnAP2b, from Nelumbo nucifera. Their expression profiles were analyzed in the double-petalled flower cultivar lotus ‘Fenhonglingxiao’. Furthermore, transgenic Arabidopsis plants overexpressing NnAP2a and NnAP2b genes showed an increased number of sepals and petals compared to wild-type plants. And we showed that both NnAP2a and NnAP2b were able to rescue the loss of sepals and petals when transformed into the Arabidopsis ap2-6 mutant. Additionally, Protein–protein interaction assays indicated that they were interacting with other floral homeotic genes, including NnAP3s, NnPI, and NnSEP3. Our results might contribute to reveal the regulatory mechanism associated with lotus floral organ formation and help the molecular breeding of Nelumbo species. Results Sequence analysis of two NnAP2 genes in N. nucifera In a previous transcriptomic study, differentially expressed genes associated with lotus stamen variation were obtained, one of which was the class A gene AP2 [32]. Through a BLAST search against the NCBI database (https://www.ncbi.nlm.nih.gov/), two AP2 homologs in lotus were discovered as LOC104606822 and LOC104601729, respectively. LOC104606822 has three isoforms, while LOC104601729 has two isoforms. One of the isoforms from LOC104606822, consisting of 1557 nucleotides encoding 519 amino acids, was successfully amplified and designated as NnAP2a. On the other hand, one of the isoforms from LOC104601729, containing 1707 nucleotides encoding 569 amino acids, was cloned and named as NnAP2b. The other sequences were not obtained in our study. The coding region of NnAP2a and NnAP2b shared 81.7% identity at the amino acid sequence. Gene structure analysis revealed that both NnAP2a and NnAP2b genes in lotus consist of ten exons and nine introns, displaying a structural pattern similar to that of the AtAP2 gene (Fig. 1A). Analyses on the gene structure, protein sequence conservative and phylogeny of NnAP2a and NnAP2b. A The gene structure of NnAP2 with the untranslated regions (UTRs). Analysis was carried out using the Gene Structure Display Server tool (http://gsds.cbi.pku.edu.cn/). Exons, introns and UTRs are shown in yellow boxes, black lines, and blue boxes, respectively. The size scale is indicated at the bottom. B Alignment of NnAP2a and NnAP2b amino acid sequences with homolog proteins from other plant species; The GenBank accession numbers are as follows: AtAP2 (Arabidopsis thaliana, NP_001190938.1), BdAP2 (Brachypodium distachyon, XP_003580693.1), CaAP2 (Capsicum annuum, PHT90362.1), CsAP2 (Citrus sinensis, XP_006494466.2), CsaAP2 (Cucumis sativus, XP_004141908.1), EgAP2 (Eucalyptus grandis, XP_010028218.1), GmAP2 (Glycine max, XP_003524643.1), GhAP2 (Gossypium hirsutum, XP_016679121.1), JrAP2 (Juglans regia, XP_018845006.1), LsAP2 (Lactuca sativa, XP_023754311.1), MtAP2 (Medicago truncatula, XP_013457432.1), NnAP2a (Nelumbo nucifera, XP_010270518.1), NnAP2b (Nelumbo nucifera, XP_010263475.1), NtAP2 (Nicotiana tabacum, XP_016459840.1), OsAP2 (Oryza sativa Japonica Group, XP_015636848.1), PdAP2 (Phoenix dactylifera, XP_008788845.2), PtAP2 (Populus trichocarpa, XP_002310715.1), PpAP2 (Prunus persica, XP_007208004.1), RcAP2 (Ricinus communis, XP_025015758.2), SiAP2 (Setaria italica, XP_004955852.1), StAP2(Solanum tuberosum, XP_006339870.1), SbAP2 (Sorghum bicolor, XP_002459580.1), SoAP2 (Spinacia oleracea, XP_021861759.1), TcAP2 (Theobroma cacao, A0A061DGG4), TaAP2 (Triticum aestivum, KAF7011720.1), ZmAP2 (Zea mays, XP_008651022.1), VvAP2 (Vitis vinifera, XP_010652782.1), NcAP2 (Nymphaea colorata, XP_031481069.1), AmtAP2 (Amborella trichopoda, XP_011625606.1). C Phylogenetic relationship among AP2 of lotus and other plant species. The phylogenetic tree was constructed by MEGA11 software according to the Neighbor-Joining method. NnAP2a and NnAP2b are marked with green triangle and AtAP2 is marked with a red spot Conceptual translation reveals that NnAP2a and NnAP2b contain two highly conserved AP2 DNA binding domains, in alignment with other AP2 orthologous proteins (Fig. 1B). Phylogenetic analysis of the amino acid sequences of the putative AP2 family genes of N.nucifera and A.thaliana (Fig. S1). Among eighteen AP2 family proteins of lotus, NnAP2a and NnAP2b was grouped into the predicted AP2. To further explore the evolutionary relationships of NnAP2, a phylogenetic tree was constructed, which classified NnAP2a and NnAP2b into two categories (Fig. 1C). NnAP2a and NnAP2b showed a higher homology with 81.5%, and both of them exhibited a closer relationship with VvAP2 (Fig. 1C). The level of sequence identity between NnAP2a and VvAP2 was found to be 72.9%. NnAP2b and VvAP2 was shown to have 74.4% sequence similarity (Table S1). The tertiary structures of NnAP2a and NnAP2b were predicted by SWISS-MODEL, demonstrating that their protein spatial structures are extremely similar (Fig. 2A and B). The 3D structural model of NnAP2 proteins and promoter sequence analysis of NnAP2 genes. A The 3D structural of NnAP2a protein. B The 3D structural of NnAP2b protein. C Cis-elements identified in the promoter region of NnAP2a and NnAP2b, respectively In this study, our focus was on the promoter region of the NnAP2a and NnAP2b genes, specifically selecting a sequence approximately 2000 bp upstream of the start codon. Thus, using PlantCARE software, we identified cis-elements in the promoter region of these genes (Fig. 2C). In addition to the typical core cis-elements, we also found cis-elements associated with abiotic stress, light response, meristem expression, and hormone regulation. Among these elements, the most prevalent cis-element were found related to light response, suggesting that a potentially significant role of light in regulating the expression of the two NnAP2 genes. Expression pattern analysis of NnAP2a and NnAP2b To analyze the gene expression in various lotus tissues, we collected different organs from ‘Fenhonglingxiao’, including leaf, flower bud, petal, petaloid stamen, stamen, carpel, and receptacle. The expression patterns of NnAP2a and NnAP2b were analyzed (Fig. 3). Our findings revealed that the expression level of NnAP2a was the highest in petals, while NnAP2b exhibited the most strongly expression in flower buds. Both genes exhibited a similar expression trend, gradually decreasing from the outer to the inner parts of the flower organs. Floral morphology of ‘Fenhonglingxiao’ varieties and expressional patterns of NnAP2 genes. A Relative Expression of NnAP2a in various tissues of ‘Fenhonglingxiao’ varieties. Le, leaf; Fb, flower bud; Pe, petal; PSt, petaloid stamen; St, stamen; Ca, carpel; Re, receptacle. B Expression of NnAP2b in various tissues of ‘Fenhonglingxiao’ varieties; C Section of ‘Fenhonglingxiao’; D The morphology of the various tissues of the lotus. The genes expression was detected using the qRT-PCR method and NnActin as the internal reference gene. All data are means of three replicates, with error bars indicating standard deviations. The values are means ± SD of triplicate analysis. Scale bars = 1 cm Subcellular localization analysis of NnAP2a and NnAP2b Using WoLF PSORT analysis, it was determined that both NnAP2a and NnAP2b are predominantly localized in the nucleus. To further verify this subcellular localization, NnAP2a-GFP and NnAP2b-GFP fusion proteins were transiently expressed in rice protoplasts. Confocal microscopy analysis confirmed that both NnAP2a and NnAP2b were indeed localized to the nucleus, which is consistent with the prediction (Fig. 4). Subcellular localization analysis of two NnAP2s. NnAP2a and NnAP2b were constructed into plasmid with fusions of GFP and driven by the CaMV35S promoter, respectively. GFP: represents gene localization, DAPI: represents nuclear staining. Scale bars = 20 μm Functional analysis of NnAP2a and NnAP2b via ectopic overexpression in Arabidopsis To validate the function of NnAP2a and NnAP2b, they were ectopically expressed in wild-type and homozygous ap2-6 mutant Arabidopsis plants. The majority of 35S::NnAP2a/WT and 35S::NnAP2b/WT plants exhibited normal flowers. Additionally, overexpression of NnAP2a in Arabidopsis negative influenced flowering time, while overexpression of NnAP2b in Arabidopsis triggered early flowering time (Fig. S2). Surprisingly, some of the transgenic plants overexpressing NnAP2a or NnAP2b showed an increase in the number of sepals and petals. Specifically, the number of sepals in 35S::NnAP2a/WT increased from 4 to 5, and the number of petals in 35S::NnAP2b/WT increased from 4 to 5–6 (Fig. 5 and Table S2). However, compared to normal plants, there were no significant differences in the morphology of stamens and pistils (Fig. 5 and Table S2). The positive transgenic lines were confirmed by RT-PCR (Fig. S3). These results revealed that the AtAP2 gene was expressed in all phenotypic Arabidopsis plants, while the NnAP2a and NnAP2b genes were specifically over-expressed in the corresponding transgenic plants. Meanwhile, in 35S::NnAP2a/WT and 35S::NnAP2b/WT over-expressing transgenic lines, the expression levels of AtAP3, AtPI, AtAG, and AtSEP3 were repressed in flowers (Fig. S4). AP2-defective Arabidopsis mutant ap2-6 shows perianth organ to carpel and stamenoid structure fused to perianth homeotic transitions. Compared to untransformed homozygous ap2-6 lines, both 35S::NnAP2a/ap2-6 and 35S::NnAP2b/ap2-6 Arabidopsis lines completely rescued petals and returned to normal flower patterns (Fig. 6). These results showed that the NnAP2a and NnAP2b positively regulate petal development in transgenic Arabidopsis plants. The phenotypes of flower pattern in the wild-type (WT) and 35S::NnAP2s (NnAP2a and NnAP2b) transgenic lines of Arabidopsis. A-C Flower of the WT Arabidopsis, showing normal four-petalled flower pattern; D-F Flower from 35S::NnAP2a/WT or 35S::NnAP2b/WT with five-petalled flower pattern. G-I Flower from 35S::NnAP2a/WT or 35S::NnAP2b/WT with six-petalled flower pattern. The red, yellow, white, and green numbers are the number of pistil, stamens, petals and sepals, respectively. Bars = 500 μm The phenotypes of flower pattern in the homozygous ap2-6 mutant and 35S::NnAP2s (NnAP2a and NnAP2b) transgenic homozygous of ap2-6 Arabidopsis. A, B mutant atap2-6. C, D 35S::NnAP2a/atap2-6. E, F 35S::NnAP2b/atap2-6. Bars = 500 μm in (A), (C), (E) and Bars = 10 mm in (B), (D), (F) Protein interaction between NnAP2s and other floral organ identity genes Using the online website String-DB, four proteins associated with floral organ identity genes were predicted to interact with NnAP2a and NnAP2b, including NnPI (XP_010269077.1), and NnAP3 (NnAP3a, XP_010263856.1; NnAP3b, XP_010271509.1; NnAP3c, XP_010240899.1) (Fig. S5). Moreover, it was found in a previous study that SEP3 was bound by AP2 [28]. The ortholog NnSEP3 (XP_010250667) is also predicted to interact with NnAP2. Furthermore, the online tools WoLF PSORT predicted that all of them are localized in the nucleus, thereby increasing the likelihood of protein interactions. Class A genes play a fundamental role in floral organogenesis by interacting with various proteins. To investigate the protein–protein interactions between NnAP2a and NnAP2b, we employed in vitro Y2H assay (Fig. 7). The results showed that NnAP2a interacted with NnPI, NnAP3a, NnAP3b, NnAP3c and NnSEP3, with NnAP2a and NnAP3c showing a weak interaction. On the other hand, NnAP2b only weakly interacted with NnAP3a and NnSEP3, while it did not show any interaction with NnPI, NnAP3a, and NnAP3b. These interactions were further confirmed by BiFC experiments conducted in onions (Fig. S6, Fig. S7, and Fig. S8). These findings suggest that NnAP2a and NnAP2b may have different regulatory pathways during the development of floral organ. However, their specific biological functions still need further understand. Protein interactions between NnAP2s and NnAP3a, NnAP3b, NnAP3c, NnPI, NnSEP3 by the yeast two-hybrid system. Interactions were determined in yeast on selective medium SD/-Trp-Leu-His-Ade, and the experiments were repeated three times with the same results. Yeast cells transformed with pGBKT7 empty vector or pGBKT7-53/pGADT7-T vectors were used as negative or positive controls, respectively Discussion Previous studies have shown that a single copy of the AP2 gene is sufficient for floral development and floral specification in Arabidopsis. However, in basal eudicot, such as B.frutescens, P.somniferum, and C.coccinea, it has been found that there are two copies of AP2 ortholog [33]. In our studies, we found two AP2 genes in N.nucifera. Additionally, three AP2-like genes have been reported in louts [34]. However, one of them is more similar to RAP2.7 than to the real AP2 homeotic gene. The NnAP2 proteins were mostly restricted to both AP2-R1 and AP2-R2 domains, supporting their characterization as AP2 floral homeotic genes and suggesting similar functionality (Fig. 1B). Interestingly, phylogenetic analysis revealed that the amino acid sequence of NnAP2a and NnAP2b protein exhibits a high degree of identity with homologous protein from Vitis vinifera (Fig. 1C). This is consistent with extensive collinearity of homologous genes in lotus genome and grape genome [2]. The findings are consistent with previous reports on lotus being the basal eudicot order of Proteales [7]. In this order, the closest relatives of lotus are mainly shrubs and woody trees [35]. Expression patterns of AP2 paralogs may vary significantly in different plant species [33]. In our study, the expression profile of NnAP2a and NnAP2b was found to be conserved. Specifically, the transcript of NnAP2a and NnAP2b showed significant accumulation in floral buds and petals (Fig. 3). The expression of NnAP2a showed peak in floral buds, while the transcript of NnAP2b was peak in petals. These results suggest that these two homologous genes, NnAP2a and NnAP2b, may have similar or dissimilar functions in regulation of floral development. It is known that a lack of AP2 function will cause the transformation of petals into stamens, while the ectopic expression of AP2 will result in the production of double flowers [21]. Our transgenic results demonstrated that both normal AP2 genes from N. nucifera were able to rescue the sepal and petal developmental defects of the ap2-6 mutant in Arabidopsis (Fig. 6). Specifically, the increasing number of sepals and petals is found in the ectopic expression NnAP2a and NnAP2b (Fig. 5). These results indicate that both normal AP2 genes have the ability to regulate sepal and petal development. Similar results have been reported in oilseed rape, where AP2 homologs BrAP2a and BrAP2b restored the phenotype of the Arabidopsis mutant (atap2-5) with the sepal defect [36]. In water lily, overexpression of AP2 homologue NsAP2 in Arabidopsis triggered more petal numbers [37]. AP2 orthologs in other species have also been shown to regulate flowering time [17, 38, 39]. Consistent with these findings, overexpression of NnAP2 genes in Arabidopsis influenced flowering time (Fig. S1). Moreover, the majority of 35S::NnAP2 plants exhibited normal flowers, which is consistent with previous studies [19]. Further studies are needed to provide more detailed information on the regulation of flowering time by NnAP2s. AP2 plays a key role in controlling flowering and flower development by repressing the expression of key flowering loci like SOC1 and AGAMOUS, as well as miR172 [17, 29]. Moreover, AP2 also acts as a transcription factor, activating other floral repressors such as AGL15 and miR156 [29]. Moreover, AP2 negatively regulates multiple floral organ identity genes, including AP3, PI, and SEP3 [28]. The flowering promoting transcripts and key floral genes AP1, FUL, SEP3 and SOC1 are negative regulated by AP2 [29]. In peper, AP2 homologs gene CaFFN also repress the expression of CaSEP3 [38]. Compared to the other SEP proteins, the SEP3 assumes a significantly more significant role. SEP3 interacts with the class B proteins AP3 and PI [40]. The members of the AP2-TPL-HDA19 complex bind to the regulatory regions of AP3 and SEP3 [28]. The orthologue of PI, HvMADS2, is a potential target of HvAP2. HvMADS2 expression increased in hvap2-2 mutant [26]. However, other study demonstrated that increasing AP2 expression levels did not lead to the induction of AP3 and PI expression [41]. Altogether, the regulation of AP2 associated with a specific MADS-box gene in floral organ development varies among different plants. Similarly in N. nucifera, NnAP2a interacts with NnAP3s, NnPI, and NnSEP3, while NnAP2b only interacts with NnAP3a and SEP3 (Fig. 7). This finding further supported the divergent role of AP2 proteins in regulating floral development in N. nucifera. Conclusions In this study, we characterized two AP2 genes in N.nucifera, and named as NnAP2a and NnAP2b. The analysis of their phylogenetic relationships, protein sequences, and expression patterns revealed that NnAP2a and NnAP2b may have both conserved and divergent roles in flower development. This hypothesis was further verified through functional analysis and identification of their protein interaction partners in transgenic Arabidopsis. Our study suggests that NnAP2 genes played an important role in regulating sepal and petal development in N.nucifera. Materials and methods Plant materials Nelumbo nucifera cultivar ‘Fenhonglingxiao’ was grown in experimental pools in Fuzhou, China. The seven different tissues including leaf, floral bud, petal, petaloidy stamen, stamen, receptacle and carpel were sampled before 10:00 am in July. Following sample collection, all the samples were immediately frozen in liquid nitrogen and stored at −80℃ for subsequent uses. For the transgenic experiments, the Arabidopsis wild-type and ap2-6 mutant were used as transgenic receptor material. The ap2-6 mutant was kindly provided by Mr. Pang Yongzhen from the Chinese Academy of Agricultural Sciences. Total RNA extraction and cDNA synthesis RNA extraction was performed following the protocol provided with the OminiPlant RNA Kit (DNaseI) (CWBIO, CW2598S, Fuzhou, China). The quality and purity of the RNA samples were assessed using the Thermo NanoDrop One C (Thermo Fisher Scientific, Designed in American). Additionally, the total RNA quality was checked by electrophoresis using a 2% agarose gel. The first strand cDNA was synthesized according to the instructions of the HiScript Q RT SuperMix for qPCR (+ gDNA wiper) (Vazyme, R123-01, Fuzhou, China). Gene clone and sequence identification To search for homologous gene sequences in Nelumbo nucifera, the AtAP2 (AT4G36920) sequence was used as a query in a blast search on the National Center for Biotechnology (NCBI) database (https://www.ncbi.nlm.nih.gov/). The obtained sequences included coding sequence (CDS), protein, full-length sequence, and promoters. The full length sequences of the NnAP2a and NnAP2b genes were amplified using Phanta Max Super-Fidelity DNA Polymerase (Vazyme, China). The primers used for amplification are listed in Table S1. The PCR products were cloned into the pCE2 TA/Blunt-Zero Vector (Vazyme, China), and then trans-formed into DH5α E.coli cells. Three positive clones were selected for identification and sequenced by Tsingke Biotechnology Company (Xiamen, China). Sequence analysis The Gene Structure Display Server 2.0 (GSDS, http://gsds.gao-lab.org/index.php) was used to analyze the exon and intron structures of NnAP2a, NnAP2b, and AtAP2. The AP2 protein sequences of A. thaliana and others plants were downloaded from the NCBI database (https://www.ncbi.nlm.nih.gov/) and the Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org/). The sequences were aligned using the Jalview tool, and the phylogenetic tree was constructed using the Neighbor-Joining method (NJ) with a bootstrap value of 1000 in MEGA11 software. To analyze the cis-regulatory elements, the 2.0 kb upstream sequences of the start codon related to NnAP2a and NnAP2b were submitted to PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/htmL/). For tertiary structure prediction modeling of proteins, the online website SWISS-MODEL (https://www.swissmodel.expasy.org/interactive) was used. The subcellular localization of NnAP2a and NnAP2b was predicted using the WoLF PSORT (https://wolfpsort.hgc.jp/). Interaction proteins of NnAP2a and NnAP2b were predicted using the online website string-db (https://cn.string-db.org/). Gene expression analysis To verify the spatiotemporal expression patterns of NnAP2a and NnAP2b, we performed real-time quantitative RT-PCR analysis on various tissues of lotus, including leaves, flower buds, petals, stamens, petaloid stamens, carpels, and receptacles. The primers used for real-time quantitative RT-PCR analysis were listed in Table S3. The quantitative PCR reactions were carried out on a BIO-RAD quantitative PCR instrument using the ChamQ Universal SYBR qPCR Master Mix kit (Vazyme, China). In addition, β-actin (LOC104586251) was selected as an internal reference gene in N. nucifera. Each experiment was conducted with three technical repeats of each of three biologic replicates. The relative expressions of NnAP2 genes at the mRNA level were calculated using the 2−ΔΔCT method [42]. Subcellular localization To obtain the full-length CDSs of NnAP2a and NnAP2b genes, PCR amplification was performed using the primers listed in Table S3. Then, after the fragments digested with BamH I and Hind III, they were ligated with the pRTVcGFP vector to construct the fusion vectors GFP-NnAP2a and GFP-NnAP2b. For the transient expression test, the fusion plasmids were transformed into rice protoplast. Rice seedlings (Zhonghua 11) were cultured at 28°C for 2 weeks under darkness for preparing protoplasts. The plasmid DNA was transformed into rice protoplasts via a PEG-mediated method as previously described [43]. To detect the expression of GFP, the GFP fluorescence signal was detected by a laser confocal microscope (Leica SP8). Ectopic expression of NnAP2 genes in Arabidopsis The full-length CDSs of NnAP2a and NnAP2b genes were cloned and constructed into the PMDC83 vectors. The primers used for cloning were listed in Table S3. These plasmids were transformed into the Agrobacterium strain GV3101 by the freeze–thaw method and then infiltrated into Arabidopsis plants using the floral dip method. The T1 seeds obtained from the infiltrated plants were selected and plated on 1/2 MS medium containing 25 mg/L hygromycin B. The T2 generation homozygous plants were used for phenotype analysis. For the phenotype analysis, the plants were grown under controlled conditions with a temperature range of 21–24°C and a photoperiod of 16 h light and 8 h dark. Yeast two-hybrid (Y2H) assay The full-length CDSs of NnAP2a, NnAP2b, NnAP3a, NnAP3b, NnAP3c, PI, and SEP3 were successfully cloned and constructed into the pGBKT7 and pGADT7 vectors (details of primers are given in Table S3). The fusion plasmids containing the bait and prey constructs were simultaneously introduced into the AH109 strain using the lithium acetate method. The transformed colonies were incubated on selective SD medium (SD/ − Trp/ − Leu and SD/ − Trp/ − Leu/ − His/ − Ade) in triplicates for interaction analysis. Yeast cells carrying the pGBKT7-53 and pGADT7-T plasmids were used as positive controls, while pGBKT7-Lam and pGADT7-T were co-transformed and served as negative controls. Bimolecular fluorescence complementation (BiFC) analysis The full-length CDSs of NnAP2a, NnAP2b, NnAP3a, NnAP3b, NnAP3c, PI, and SEP3 were cloned and constructed into the pCAMBIA1300S-YN and pCAMBIA1300S-YC vectors (the primers are listed in Table S3). The fusion plasmids were transformed into Agrobacterium GV3101 competent cells. Positive clones were selected and used for the transformation into onion for BiFC test. The transformed cells were resuspended in 100 mL MS medium containing 250 mg glucose, 100 mM acetosyringone, 20 mM Na3PO4, and 0.5 M MES. The suspension was then incubated in a shaker at 28°C and 220 rpm for 14–16 h, until OD600 reached 0.6. Afterwards, the Agrobacterium GV3101 strains carrying the recombinant expression plasmid and the control were separately injected into onion inner epidermal cells. The treated onions were cultured in the darkness for 24 h and then grown for 2 days at 28°C under 16 h light / 8 h dark. And the GFP fluorescence signal was detected by a laser confocal microscope (Leica SP8). Supplementary Information Additional file 1: Fig. S1. Phylogenetic tree of the AP2 family in lotus and Arabidopsis. The phylogenetic tree was constructed by MEGA11 software according to the Neighbor-Joining method. Fig. S2. The phenotypes of flowering time in the wild-type (WT) and 35S::NnAP2s (NnAP2a and NnAP2b) transgenic lines of Arabidopsis. Bars = 10 mm. Additional file 2: Fig. S3. The expression levels of NnAP2s and AtAP2 in wild-type and transgenic plants as determined by semi-quantitative RT-PCR. Atactin gene was used as the internal control. Additional file 3: Fig. S4. The expressions of AtAP3, AtPI, AtAG, and AtSEP3 endogenous genes related to floral development were examined with qRT-PCR in wild-type (WT), 35S::NnAP2a, and 35S::NnAP2b transgenic Arabidopsis. Significant difference was assessed by One-Way ANOVA and indicated by asterisks; (*) p ≤ 0.05, (**) p ≤ 0.01. Data are expressed as the mean of three biological replicates, with error bars indicating the SD values. Fig. S5. Predicted functional partners with NnAP2a (A) and NnAP2b (B) in Nelumbo nucifera, respectively. AP3 also named NnAP3a; LOC104606822: NnAP2a; LOC104601729: NnAP2b; LOC104607542: NnAP3b, LOC104585644: NnAP3c. PI also named as NnPI. Fig. S6. Protein interactions between NnAP2a and NnAP3a, NnAP3b, NnAP3c, NnPI, NnSEP3 by the BiFC system. Note: YFP is Yellow Fluorescent Signal Protein; DAPI: fluorescent dye; BF: bright field; Nu is the nucleus of the cell. Scale bars = 50 μm. Fig. S7. Protein interactions between NnAP2b and NnAP3a, NnSEP3 by the BiFC system. Note: YFP is Yellow Fluorescent Signal Protein; DAPI: fluorescent dye; BF: bright field; Nu is the nucleus of the cell. Scale bars = 50 μm. Fig. S8. Protein interactions between empty vector YN and NnAP3a, NnAP3b, NnAP3c, NnPI, NnSEP3 by the BiFC system as negative controll. Note: YFP is Yellow Fluorescent Signal Protein; DAPI: fluorescent dye; BF: bright field; Nu is the nucleus of the cell. Scale bars = 50 μm. Additional file 4: Table S1 Identity matrix of the AP2 for the full-length protein sequences in different species. Table S2 The number of floral organs after over-expression NnAP2s in Arabidopsis. Table S3 Primers used in this study. Acknowledgements The authors would like to thank Dr. Pang Yongzhen at Chinese Academy of Agricultural Sciences for providing Arabidopsis ap2-6 mutant. We also thank Dr. Rebecca Njeri Damaris from Pwani University for her valuable assistance with English language editing. Abbreviations APETALA2 Bimolecular fluorescence complementation Coding sequence PISTILLATA SEPALLATA3 Yeast two-hybrid Authors’ contributions X.L., Y.Q., and Z.L. designed the experiments and wrote the manuscript; X.L. and Y.Q. conducted the experiments and edited figures and tables; X.L., Y.Q., D.C. and J.D. analyzed the data; S.C., P.Y., and Z.L. revised the manuscript. All authors have read and approved the final manuscript. Funding This work was partially supported by the National Natural Science Foundation of China (Grant numbers 32102422 and 32302564), Minjiang University (MJY22037). Data availability The datasets generated and/or analysed during the current study are available in the NCBI repository. Accession numbers of the proteins used in this study were as follows: NnAP2a (LOC104606822), NnAP2b (LOC104601729), NnPI (XP_010269077.1), NnAP3a (XP_010263856.1), NnAP3b (XP_010271509.1), NnAP3c (XP_010240899.1), NnSEP3 (XP_010250667.1), and β-actin (LOC104586251). Declarations Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Liu Xuelian and Qin Yu contributed equally to this work. Contributor Information Yang Pingfang, Email: yangpf@hubu.edu.cn. Lin Zhongyuan, Email: lzy2019@mju.edu.cn. References Associated Data Supplementary Materials Additional file 1: Fig. S1. Phylogenetic tree of the AP2 family in lotus and Arabidopsis. The phylogenetic tree was constructed by MEGA11 software according to the Neighbor-Joining method. Fig. S2. The phenotypes of flowering time in the wild-type (WT) and 35S::NnAP2s (NnAP2a and NnAP2b) transgenic lines of Arabidopsis. Bars = 10 mm. Additional file 2: Fig. S3. The expression levels of NnAP2s and AtAP2 in wild-type and transgenic plants as determined by semi-quantitative RT-PCR. Atactin gene was used as the internal control. Additional file 3: Fig. S4. The expressions of AtAP3, AtPI, AtAG, and AtSEP3 endogenous genes related to floral development were examined with qRT-PCR in wild-type (WT), 35S::NnAP2a, and 35S::NnAP2b transgenic Arabidopsis. Significant difference was assessed by One-Way ANOVA and indicated by asterisks; (*) p ≤ 0.05, (**) p ≤ 0.01. Data are expressed as the mean of three biological replicates, with error bars indicating the SD values. Fig. S5. Predicted functional partners with NnAP2a (A) and NnAP2b (B) in Nelumbo nucifera, respectively. AP3 also named NnAP3a; LOC104606822: NnAP2a; LOC104601729: NnAP2b; LOC104607542: NnAP3b, LOC104585644: NnAP3c. PI also named as NnPI. Fig. S6. Protein interactions between NnAP2a and NnAP3a, NnAP3b, NnAP3c, NnPI, NnSEP3 by the BiFC system. Note: YFP is Yellow Fluorescent Signal Protein; DAPI: fluorescent dye; BF: bright field; Nu is the nucleus of the cell. Scale bars = 50 μm. Fig. S7. Protein interactions between NnAP2b and NnAP3a, NnSEP3 by the BiFC system. Note: YFP is Yellow Fluorescent Signal Protein; DAPI: fluorescent dye; BF: bright field; Nu is the nucleus of the cell. Scale bars = 50 μm. Fig. S8. Protein interactions between empty vector YN and NnAP3a, NnAP3b, NnAP3c, NnPI, NnSEP3 by the BiFC system as negative controll. Note: YFP is Yellow Fluorescent Signal Protein; DAPI: fluorescent dye; BF: bright field; Nu is the nucleus of the cell. Scale bars = 50 μm. Additional file 4: Table S1 Identity matrix of the AP2 for the full-length protein sequences in different species. Table S2 The number of floral organs after over-expression NnAP2s in Arabidopsis. Table S3 Primers used in this study. Data Availability Statement The datasets generated and/or analysed during the current study are available in the NCBI repository. Accession numbers of the proteins used in this study were as follows: NnAP2a (LOC104606822), NnAP2b (LOC104601729), NnPI (XP_010269077.1), NnAP3a (XP_010263856.1), NnAP3b (XP_010271509.1), NnAP3c (XP_010240899.1), NnSEP3 (XP_010250667.1), and β-actin (LOC104586251).