Identification and Functional Analysis of the Flower Development-Related TCP Genes in Erycina pusilla

Abstract

Orchid flowers have evolved in concert with pollinators to form highly specialized structures resulting in zygomorphy. In dicotyledons, it is widely accepted that CYC-like genes are involved in the dorsoventral polarity establishment of flowers, which determines the development of zygomorphic flowers. However, the function of TCP transcription factors involved in orchid floral development is rarely known. Here, we found 15 unigenes with TCP domain (EpTCPs) from the previously reported Erycina pusilla unigene database. The expression patterns of EpTCPs in various tissues and different floral organs were successively detected by quantitative real-time PCR. The results revealed that the CYC-like gene (EpTCP25) and CIN-like genes (EpTCP11 and EpTCP26) were highly expressed in inflorescences but lowly expressed in leaves and roots. What is more, these three genes were expressed relatively high in the dorsal labellum, and EpTCP26 showed differential expression along the dorsoventral polarity of tepals, which was high in the dorsal and low in the ventral. Ectopic expression of EpTCP25 in Arabidopsis repressed primary root growth and delayed flowering. EpTCP26 overexpression in Arabidopsis promoted primary root growth and leaf growth. In contrast, EpTCP11 overexpression repressed primary root growth and changed the radially symmetric flower to a bilaterally symmetric flower by inhibiting the elongation of one or two adjacent petals. In addition, the homeotic transition of floral organs is generated when these genes are ectopically expressed in Arabidopsis, suggesting their roles in floral morphogenesis. Altogether, our results indicate that CIN-like genes would be associated with the unique flower pattern development of Erycina pusilla.

1. Introduction

Orchidaceae is one of the largest and most evolutionary families in the plant kingdom [1]. The flower of the orchid has evolved in concert with insects to form highly specialized structures, so the orchid is considered to be one of the ideal plants for the study of floral developmental biology in monocotyledons. The typical orchid flower is bilaterally symmetrical and includes three whorls of floral organs: the outer whorl consists of three outer tepals (named sepals); the second whorl consists of two ventral inner tepals (named petals) and a highly specified dorsal inner tepal (named labellum); and the third whorl is a column, which is the reproductive structure consisting of fused male stamen and female tissues.

Previous research on the genetic mechanism of orchid zygomorphy flower development mainly focuses on exploring how the orthologous MADS-box genes involved in the ABCDE model participate in the development and evolution of highly specialized floral organs in orchids [2,3,4,5,6,7,8,9]. The ‘Orchid Code’ and ‘Perianth Code’ models are used to demonstrate that AP3-like and AGL6-like MADS-box transcription factors play essential roles in different floral organ formations [8,10,11]. In Cymbidium sinense and Cymbidium ensifolium, recent studies indicate that regulation of peloric or pseudopeloric flower development is associated with duplication and subfunctional differentiation of the Perianth Code genes [9,12,13].

At the early developmental phase of floral organogenesis of orchid, dorsal floral organs primordia first initiate in the flower meristem, followed by ventral floral organs primordia, and, finally, column primordium initiates at the central region [9]. Therefore, there will be cis-acting elements on the Perianth Code genes that regulate the position effect in the flower primordium. Correspondingly, there will be trans-acting factors involved in the position effect regulation. TEOSINTE BRANCHED1/CYCLOIDEA/PCF (TCP) transcription factors are presumed to be associated with the determination of the dorsal–ventral position information on orchid flower primordium.

TCP transcription factors, with the conserved TCP domains, are divided into Class I (PCF or TCP-P class) and Class II (TCP-C class) [14]. Class II members are subdivided into the CIN clade and CYC/TB1 clade [15]. Previous studies only found that CYC-like genes control flower monosymmetry development in dicots. They contribute to establishing the dorsoventral polarity of flowers [16,17,18,19,20,21,22,23,24,25,26]. However, there are a few functional studies of orchid TCP transcription factors. In Orchis italica, 12 independent TCP gene fragments from the inflorescence transcriptome database are identified, but there are no members of the CYC/TB1 clade. Using the primers designed based on the CYC/TB1-like gene sequence of Phalaenopsis equestris, Oitatb1 is amplified from the genomic DNA of Orchis italica. However, the expression of Oitatb1 cannot be detected in inflorescences [27]. In Phalaenopsis, three CYC-like genes are isolated, and functional analysis shows that PhCYC2 and PhCYC3 contribute to the ventral petal identity and are associated with labellum development [28]. Using the genome database of Phalaenopsis equestris, 23 TCP genes including Class I and Class II members are characterized. Functional analysis indicates that PePCF10 and PeCIN8 play an important role in orchid ovule development by regulating cell proliferation [29].

Erycina pusilla can be used as a model plant to study flowering physiology, pollination behavior and gene function of orchids because of its short growth cycle, flowering in bottles and small genome [30]. However, there is no report on the characterization and function of TCP genes in Erycina pusilla. Therefore, we characterized the expression patterns of EpTCPs isolated from the reported Erycina pusilla transcriptome data [30]. Then, three flower development-related genes (the CYC/TB1 clade EpTCP25, the CIN clade EpTCP11 and EpTCP26), whose expression level was relatively high in the labellum, were ectopically expressed in Arabidopsis to analyze their functions in plant growth and flower development. Our results provide molecular evidence that besides CYC-like genes, CIN-like genes also played roles in the control of orchid unique flower-type development.

2. Materials and Methods

2.1. Plant Materials

The Erycina pusilla plantlets were kept in the tissue culture room at 25 ± 2 °C under 16 h light/8 h dark photoperiod. The Arabidopsis (Col-0) were grown in the growth chamber (22 °C, 16 h of light per day).

2.2. TCP Family Members Identification and Phylogenetic Analysis

The TCP protein sequences of Arabidopsis were obtained from the NCBI database and accession numbers were listed in Supplementary Table S1. These AtTCP amino acid sequences were used to search homologous unigenes from the reported Erycina pusilla unigene database [30] by the tblastn in BioEdit Sequence Alignment Editor software (version 7.1.9). The candidate unigenes were translated into amino acid sequences before analysis of the conserved TCP domain using the online tool CDD from NCBI (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) (accessed on 10 February 2017). The deduced amino acid sequences of Arabidopsis and Erycina pusilla TCP transcription factors were aligned using MEGA6, and the well-aligned sequences were further processed to generate Neighbor-joining trees under 1000 replicates of bootstrap iterations [31].

2.3. Relative Expression Analysis of EpTCPs

Total RNA from various tissue parts of Erycina pusilla was extracted from different tissues using the RNAprep Pure Plant Kit (Polysaccharides and Polyphenolics-rich) (Tiangen, Beijing, China). A 1 μg aliquot of total RNA was reverse transcribed according to the instruction manual of the PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa, Dalian, China). The qPCR assays were performed using the SYBR^® Green PCR Master Mix (Applied Biosystems, Waltham, MA, USA) on a Roche LightCycler 480 (Roche Applied Science, Penzberg, Upper Bavaria, Germany) according to its manual. Relative expression was calculated using the 2^−ΔΔCT method with ACTIN as the reference gene [32]. The ACTIN gene sequence was obtained from the reported Erycina pusilla unigene database [30] according to Nr annotation. Each measurement was carried out in triplicate with three biological replicates. Graphs were constructed by GraphPad Prism 9.0 (GraphPad Software). All the primer sequences used in qPCR are shown in Supplementary Table S2.

2.4. Arabidopsis Transformation

The coding sequences of EpTCP11 (Genbank accession number: PP515908), EpTCP25 (Genbank accession number: PP515909) and EpTCP26 (Genbank accession number: PP515910) were amplified, respectively, using the following primer pairs: 5′- ACGGGGGACTCTTGACCATGGAT-ATGGAAAGTGATGAGATGTATTCG -3′ and 5′- AAGTTCTTCTCCTTTACTAGTCA-GATCTATACCCTTACCATCTCCTTTCAGTT -3′, 5′- ACGGGGGACTCTTGACCATG-GATATGTATAACCCCCCATCCCCC -3′ and 5′- AAGTTCTTCTCCTTTACTAGTCAG-ATCTATAACCCTAACAACATTACCTTCCTC -3′, 5′- ACGGGGGACTCTTGACCATG-GATATGCCCATCTCCCACATTATTC -3′ and 5′- AAGTTCTTCTCCTTTACTAGTCA-GATCTATATGATATGGCATGCTGTCAAGC -3′. The amplified products were cloned into the pCAMBIA1302 vector double digested at the restriction sites NcoⅠ and BglII using a ClonExpress II One Step Cloning Kit (Vazyme, Nanjing, China). The recombinant plasmids were further transformed into the Agrobacterium tumefaciens EHA105 strains. The Agrobacterium-mediated plant transformation was performed using the floral dipping method [33]. Seeds of 35S:EpTCP11, 35S:EpTCP25 and 35S:EpTCP26 transgenic plants were selected on half-strength Murashige and Skoog (MS) medium supplemented with 20 mg L⁻¹ Hygromycin B (Roche, Shanghai, China). The hygromycin-resistant plants were identified by PCR. Three independent lines of 35S:EpTCP11, 35S:EpTCP25 and 35S:EpTCP26 transgenic Arabidopsis plants were selected randomly to carry out real-time PCR detection and phenotypic analysis.

2.5. Statistical Analysis

Data were analyzed in Microsoft Excel 2016. All data were expressed as means ± SD. Statistical analysis was performed by t-test. p < 0.05 and p < 0.01 are considered statistically significant.

3. Results

3.1. Phylogeny of TCP Transcript Factors from Erycina pusilla

Fifteen TCP genes were confirmed from the reported Erycina pusilla unigene database [30]. A phylogenetic tree was constructed using 15 Erycina pusilla TCP transcript factors (EpTCPs) and 21 Arabidopsis TCP transcript factors (AtTCPs) (Figure 1). Based on the AtTCP gene family grouping information [14], two types of TCP can be distinguished: Class I, which is only formed by the PCF clade, and Class II, which can be further subdivided into the CYC/TB1 clade and CIN clade. The PCF clade contained EpTCP1, EpTCP10, EpTCP12, EpTCP15, EpTCP17, EpTCP20, EpTCP22 and EpTCP24. The CYC/TB1 clade contained EpTCP25, whereas the CIN clade contained EpTCP2, EpTCP5, EpTCP11, EpTCP14, EpTCP21 and EpTCP26.

3.2. Expression Patterns of the EpTCP Genes in Different Tissues

Quantitative real-time PCR was performed to analyze the expression patterns of the fifteen EpTCP genes in various tissues of Erycina pusilla. The examination tissues including leaves, roots, inflorescences the length of 1~2 mm (I1), inflorescences the length of 3~4 mm (I2), floral buds the length of 3~4 mm (B1) and floral buds the length of 5~7 mm (B2) (Figure 2A). The B2 stage floral bud has completed labellum specialization. Although the morphology of the labellum has not yet been specialized, each round of floral organs in the B1 stage floral bud can be separated, whereas it is difficult to isolate floral organs from the floral bud whose developmental stage is earlier than B1.

The relative expression levels of EpTCPs are shown in Figure 2B. EpTCP25 showed a remarkable expression level in I2 stage inflorescences, while the expression of EpTCP14 was highest in roots and low in other tissues. The expression levels of EpTCP11 and EpTCP21 were relatively high in B2-stage floral buds and I1-stage inflorescences, moderate in I2-stage inflorescences and B1-stage floral buds, and low in leaves and roots. EpTCP12 and EpTCP26 were mainly expressed in B2-stage floral buds and I1-stage inflorescences; however, the expression of EpTCP12 in leaves and roots was higher than that in B1-stage floral buds and I2-stage inflorescences. EpTCP22 showed relatively high expression in inflorescences and B1-stage floral buds, while the expression level was higher in leaves and roots than in B2-stage floral buds. The relative expression of EpTCP24 was highest in I1-stage inflorescences and lowest in roots, whereas EpTCP1, EpTCP5 and EpTCP15 showed the highest expression in B2-stage floral buds. EpTCP2 and EpTCP10 showed much higher accumulation in leaves and roots than in inflorescences and floral buds. EpTCP17 and EpTCP20 were highly expressed in roots, which was close to the expression level in floral buds.

3.3. Expression Patterns of the EpTCP Genes in Different Floral Organs

According to the expression patterns of EpTCP genes in different tissues, we selected four genes (EpTCP11, EpTCP24, EpTCP25 and EpTCP26), which showed low expression in leaves and roots but high expression in inflorescences. What is more, the highest relative expression level was higher than 2. To further analyze the expression patterns in various floral organs, B1-stage floral buds were dissected into five parts (Figure 3A), including two dorsal outer tepals (OT-D), one ventral outer tepal (OT-V), one dorsal labellum (L-D), two ventral inner tepals (IT-V) and a column (C).

The qPCR assay of the four EpTCP genes is shown in Figure 3B. The results showed that the four EpTCP genes were highly expressed in the column tissues. And the expression pattern of these four EpTCPs in different tepals was as follows: EpTCP24 is mainly expressed in outer tepals, while the expression of EpTCP25 was relatively high in ventral outer tepal, dorsal labellum and ventral inner tepals when compared with dorsal outer tepals. EpTCP11 and EpTCP26 showed the highest expression level in the dorsal labellum. Furthermore, EpTCP26 showed higher expression in dorsal outer tepals than in ventral outer tepals. All the results indicated that EpTCP26 showed differential expression along the dorsoventral polarity of tepals, which was high in the dorsal and low in the ventral.

3.4. Phenotypes of EpTCP11, EpTCP25 or EpTCP26 Overexpression in Arabidopsis

Based on the results of expression patterns in different floral organs, we selected EpTCP11, EpTCP25 and EpTCP26, whose expression levels were relatively high in the labellum, to verify the function by ectopically expressing them in Arabidopsis, respectively. The presence of the transgene in transgenic lines was confirmed by PCR (Supplementary Figure S1). Three independent lines of 35S:EpTCP11, 35S:EpTCP25 and 35S:EpTCP26 transgenic Arabidopsis plants were selected randomly to carry out a phenotypic analysis. The results of real-time PCR showed that exogenous EpTCP was expressed in all the selected transgenic lines (Supplementary Figure S2).

3.4.1. Effects of EpTCP11, EpTCP25 or EpTCP26 Overexpression on the Primary Root Growth of Arabidopsis

The seeds of 35S:EpTCP11, 35S:EpTCP25 and 35S:EpTCP26 transgenic Arabidopsis were germinated on MS medium for 7 days, and then the primary root lengths were measured. As shown in Figure 4, overexpression of EpTCP11 or EpTCP25 significantly inhibited the elongation of Arabidopsis primary roots. EpTCP11 overexpression inhibited primary root growth most severely, and the root lengths shortened by 40%~50%. In contrast, overexpression of EpTCP26 significantly promoted the elongation of primary roots in Arabidopsis.

3.4.2. Effects of EpTCP11, EpTCP25 or EpTCP26 Overexpression on the Plant Growth and Development of Arabidopsis

The wild-type and transgenic Arabidopsis plants were transplanted into nutrient soil after growing on MS medium for 7 days. About 15 days after transplantation, the wild-type and transgenic Arabidopsis plants began to bolt in succession. The leaf size and rosette leaf number of 35S:EpTCP11 transgenic plants (Figure 5B–D,K) were almost the same as the Col wild-type Arabidopsis (Figure 5A,K), while in 35S:EpTCP26 transgenic plants, leaves were larger than those of the Col wild-type Arabidopsis (Figure 5H–J), but the number of rosette leaves was not significantly different from that of the wild-type Arabidopsis (Figure 5K). Compared with the Col wild-type Arabidopsis, the rosette leaf number of 35S:EpTCP25 transgenic plants was significantly increased, which indicates that overexpression of EpTCP25 delayed the flowering time (Figure 5E–G,K).

3.4.3. Effects of EpTCP11, EpTCP25 or EpTCP26 Overexpression on the Floral Organs Development of Arabidopsis

The flowers of 35S:EpTCP11, 35S:EpTCP25 and 35S:EpTCP26 transgenic Arabidopsis were observed for a week from the beginning of flowering. There were four phenotypic variations of flowers in the transgenic Arabidopsis. Flowers with phenotypic variation 1 changed from radial symmetry to bilateral symmetry with one petal shortened (Figure 6B), while flowers with phenotypic variation 2 changed from radial symmetry to bilateral symmetry with two adjacent petals shortened (Figure 6C). Flowers with phenotypic variation 3 exhibited a petaloid structure at the top of one stamen (Figure 6F). And flowers with phenotypic variation 4 showed stamenoid structure at the petal margin (Figure 6G).

Flowers of 35S:EpTCP11 transgenic Arabidopsis had the highest probability of phenotypic variation 1 and phenotypic variation 2. The variation rates of these two phenotypes reached 26–42%, which was much higher than the variation rate of wild-type Arabidopsis at less than 10%. Moreover, the average length of long petals and short petals of the flowers with phenotypic variation 1 and 2 in 35S:EpTCP11 transgenic Arabidopsis were compared with the average length of the petals of wild-type flowers, respectively. The results showed that the length of short petals was significantly different from that of wild-type flower petals, while the length of long petals was almost the same as that of wild-type flower petals (Figure 6D). All these suggested that EpTCP11 may change the flower symmetry.

Phenotypic variations of petaloid stamen and stamenoid petal occurred in the flowers of 35S:EpTCP11, 35S:EpTCP25 and 35S:EpTCP26 transgenic Arabidopsis. Although the variation rate was only 2–8%, flowers with these two phenotypic variations were not found in wild-type Arabidopsis. The result suggested that EpTCP11, EpTCP25 and EpTCP26 may cause the homeotic conversion of floral organs.

4. Discussion

The coevolution of orchid flowers and pollinators leads to the highly specialized morphology of the orchid flower, which makes the orchid plant one of the ideal materials for studying flower-type development of monocotyledons. Orchid has a typical zygomorphic flower. And CYC-like TCP gene is the determinant of flower symmetry development in dicotyledons. In the present study, based on the reported Erycina pusilla transcriptome data [30], by expression pattern characterization and ectopically overexpression in Arabidopsis, we found that CIN-like TCP genes would be involved in flower symmetry development of Erycina pusilla.

Erycina pusilla, an orchid model species, has typically bilaterally symmetrical flowers. But identification and functional studies of TCP transcription factors have not been reported.

It has been reported that TCP transcription factors play roles in flower development from the early stage [17]. Therefore, by detecting the relative expression of EpTCPs in different tissues of Erycina pusilla, we screened out four genes that were lowly expressed in leaves and roots but highly expressed in inflorescences. Then, the expression of these four genes was further analyzed in various flower organs using small floral buds at the B1 stage, whose labellum has not yet been specialized.

The flowers of Erycina pusilla are typically bilaterally symmetrical, and each whorl of tepal establishes dorsoventral asymmetry. The outer perianths can be classified into two dorsal narrow sepals and one ventral petaloid sepal, whereas the inner perianths can be classified into one dorsal well-differentiated labellum and two ventral normal petals. The qPCR assay showed that EpTCP11 was highly expressed in the dorsal labellum, EpTCP26 was highly expressed in the dorsal tepals including the dorsal labellum and the dorsal outer sepals, whereas EpTCP25 was lowly expressed in the dorsal outer sepals. Therefore, we inferred that EpTCP11, EpTCP25 and EpTCP26 could be correlated with the regulation of zygomorphic flower development in Erycina pusilla. Phylogenetic tree analysis showed that EpTCP25 belongs to the CYC/TB1 clade while EpTCP11 and EpTCP26 belong to the CIN clade.

It is reported that members of the CYC/TB1 clade inhibit plant growth and proliferation [16,34]. Overexpression of Antirrhinum CYC in Arabidopsis represses the growth of plants and leaves [35]. Similarly, overexpression of the CYC-like gene EpTCP25 in Arabidopsis inhibited primary root elongation and delayed flowering. However, accumulated evidence demonstrates that CYC-like genes play key roles in the zygomorphic flower development of dicotyledons [16,17,18,19,20,21,22,23,24,25,26]. In monocotyledons Commelinaceae and Zingiberales, CYC-like genes are predominantly expressed in the ventral petals [36,37]. In contrast, CYC-like genes are recruited to specialize in the development of dorsal petals in dicotyledons [38]. In Phalaenopsis, PhCYC1, PhCYC2 and PhCYC3 with high homology are isolated. PhCYC2 and PhCYC3 are highly expressed in the inner tepals (two petals and one labellum) of wild-type Phalaenopsis equestris, while they are lowly expressed in the labelloid petals and labellum of peloric flower (three labellums mutants). With the heterologous overexpression of PhCYC2 and PhCYC3 in Torenia furnieri, the results indicate that PhCYC2 and PhCYC3 contribute to the ventral petal identity and are associated with labellum development [28]. In this study, overexpression of EpTCP25 in Arabidopsis did not change floral symmetry. This finding may be related to the fact that there was no difference in the expression of EpTCP25 between the dorsal labellum and ventral petals. And it was possible that other TCP transcription factors were involved in the bilaterally symmetric flower development of Erycina pusilla.

Overexpression of the CIN-like gene EpTCP11 repressed the primary root growth of Arabidopsis and changed the radially symmetric flower to a bilaterally symmetric flower by inhibiting the elongation of one or two adjacent petals. In contrast, overexpression of the CIN-like gene EpTCP26 in Arabidopsis promoted primary root growth and leaf growth but did not affect petal growth. In previous studies on Phalaenopsis equestris, it was found that overexpression of PeCIN8 represses plant growth of Arabidopsis [29]. All these indicated that different members in the same clade of TCP gene family have functional divergence, and even play opposite roles in the same developmental process.

Phylogenetic tree analysis showed that EpTCP11 possessed high homology with AtTCP24 and AtTCP2 of Arabidopsis. It has been reported that AtTCP24 associates with ABAP1 and negatively regulates the transcription of AtCDT1a and AtCDT1b, which may limit DNA replication and consequently block cell cycle progression at G1 to S [39]. AtTCP2 can upregulate NGA3 gene expression by activating its promoter. NAG3 overexpression represses plant growth and makes leaves narrow and short. So, part of the CIN-like TCP role in leaf development may be mediated by NGA3 upregulation [40]. Further experiments are needed to explore whether the molecular mechanism of EpTCP11 in regulating primary roots and individual petals growth of Arabidopsis is similar to that of AtTCP24 and AtTCP2.

EpTCP26 possessed high homology with AtTCP10, AtTCP3 and AtTCP4 of Arabidopsis. In Arabidopsis, the normal expression of the CIN-like TCP gene is critical for leaf morphogenesis [41,42,43,44,45]. Expression of a chimeric repressor from AtTCP3 (AtTCP3SRDX) resulted in the inhibition of main root elongation and the formation of larger leaves with wavey surfaces and serrated margins because of excessive growth in the marginal region. AtTCP10 and AtTCP4 have functions similar to that of AtTCP3 [46]. In Cyclamen persicum, CpTCP1, which is clustered in the same group as AtTCP3 and AtTCP4, has a function similar to that of AtTCP3. Expression of CpTCP1SRDX generates larger leaves and smaller flowers with undulating edges [47]. These CIN-like TCP genes may regulate the morphogenesis of vegetative organs and flowers by suppressing boundary-specific gene expression. In this study, we observed that overexpression of EpTCP26 in Arabidopsis promoted root elongation and leaf growth, but did not affect petal size. We inferred that different downstream genes could be involved in the regulatory pathway of EpTCP26.

Unexpectedly, overexpression of EpTCP11, EpTCP25 or EpTCP26 generated a homeotic transition of floral organs. Whether they affect the floral organ identity by regulating B and C function MADS-box genes remained to be further studied.

5. Conclusions

In this study, we screened three flower development-related TCP genes (the CYC/TB1 clade EpTCP25, the CIN clade EpTCP11 and EpTCP26) from the reported Erycina pusilla transcriptome database [30]. They all expressed relatively high levels in the dorsal labellum. Ectopic expression of these genes in Arabidopsis led to the homeotic transition of floral organs. What is more, CIN-like gene EpTCP11 overexpression changed the radially symmetric flower to a bilaterally symmetric flower. Our study offered new insight for further exploring the genetic regulation of orchid floral zygomorphy.

In previous research, ‘Orchid Code’ and ‘Perianth Code’ indicate that the Perianth Code MADS-box gene duplications followed by sub-functionalization events contribute to the evolutionary origin of morphological novelties in orchids [8,9,13]. Together with our studies, we assumed that different TCP genes, including CYC-like and CIN-like genes, were probably recruited to interact with various MADS-box genes for distinguishing different types of floral organs during orchid evolution.