Wild plants of C. goeringii were cultivated and collected from the Forest Orchid Garden greenhouse at Fujian Agriculture and Forestry University (Fuzhou, Fujian Province, China). The temperature of the greenhouse culture was approximately 25°C. The roots, stems, leaves and flowers (sepal, petal, lip and column) were used in this study. All samples were gathered, put in tubes, and frozen before being stored in an ultralow temperature refrigerator at −80°C.

CgTCP (C. goeringii TCP gene) genes were identified in the C. goeringii genome (Sun et al., 2021), and five species (Arabidopsis thaliana, Oryza sativa, Ananas comosus, Phalaenopsis equestris, and Dendrobium catenatum) TCP proteins served as the queries. The genome and transcriptome data of C. goeringii were downloaded from the NCBI database¹, and the Arabidopsis thaliana TCP proteins were downloaded from TAIR², the rice TCP proteins download from Rice Genome Annotation Project (http://rice.uga.edu/), the Ananas comosus, Phalaenopsis equestris, and Dendrobium catenatum TCP proteins download from previous studies (Lin et al., 2016; Zhang et al., 2021a; Wang et al., 2022). BLAST and HMMER were used to identify the TCP genes. A BLAST table (E-value < 0.05) and sequence were acquired from TBtools (Chen et al., 2020b). The hidden Markov model (HMM) of the TCP domain (PF03634) was downloaded from the Pfam protein family database³, and the HMM profile was used to identify the TCP protein sequences through the Simple HMM Search in TBtools (Chen et al., 2020b). Combining the results of BLAST and Hmmsearch, all TCP genes were analyzed by NCBI Batch-CDD tools⁴, and the genes containing the entire TCP domain were retained. The ExPASy online tool⁵ (Artimo et al., 2012) used to predict physicochemical properties of the TCP genes, and the Plant-mPloc⁶ (Chou and Shen, 2010) used to predict subcellular localizations.

According to the gene location of the TCP genes in the C. goeringii genome, complete gene sequences were extracted. Complete gene sequences and coding sequences (CDSs) were prepared for the gene structure analysis in GSDS⁷ (Hu et al., 2015). The MEME⁸ online tool (Bailey et al., 2009) used the default parameters to examine the motifs of TCP genes, and the visualization of the results of the gene structure and motifs were combined using TBtools.

The TCP genes of C. goeringii were collected and identified, and the sequences were aligned by MUSCLE (Edgar, 2004) using MEGA 7 (Kumar et al., 2016). Alignment was also conducted in the TCP genes of C. goeringii, Arabidopsis thaliana, Phalaenopsis equestris (Lin et al., 2016) and Dendrobium catenatum (Zhang et al., 2021a). The phylogenetic analysis was performed by the maximum likelihood (ML) method using RA×ML (RAxML-HPC2 on XSEDE) in the CIPRES Science Gateway web ^server9 (Miller et al., 2010) with the Protein CAT model and GTR matrix with 1,000 bootstrap iterations. The Evolview (He et al., 2016) used to polish the output phylogenetic tree.

The protein sequences (PEPs), CDSs and annotation files (gff files) of the C. goeringii, D. catenatum and P. equestris genomes were prepared for a collinearity analysis and gene mapping. The TCP gene location information in the C. goeringii genome chromosomes was acquired in the gff file, and visualization was conducted by the R package ‘RIdeogram’ (Hao et al., 2020). JCVI v1.2.10 (https://pypi.org/project/jcvi/) was used in the collinearity analysis of C. goeringii, P. equestris and D. catenatum. First, the PEP, CDS and gff files were formatted. Second, the PEP files were aligned and analyzed to acquire the collinearity file. The anchor file was used in the visualization, and the TCP genes were highlighted in the collinearity map.

In total, 2000 bp upstream sequences of TCP genes were retrieved to investigate the regulatory functions in plant growth and development. The sequences were extracted by TBtools according to the gene locations in the annotation file. The cis-acting elements of the TCP genes were identified and annotated in the promoter regions by the online tool PlantCARE¹⁰ (Lescot et al., 2002). TBtools (Chen et al., 2020b) was used to display the findings of the cis-acting element counts and annotation.

RNA-Seq by Expectation Maximization (RSEM) (Li & Dewey, 2011) was employed for the transcript quantification and calculation of the fragments per kilobase per million mapped reads (FPKM) value of each gene in the transcriptomic analysis. The FPKM matrix and heatmaps were generated by TBtools (Chen et al., 2020b).

Roots, stems, leaves, and mature flowers (sepal, petal, lip, and column) from C. goeringii grown at the Forest Orchid Garden of Fujian Agriculture and Forestry University were taken for a quantitative real-time PCR (qRT-PCR) analysis to confirm the expression pattern of the TCP genes. Three replicates of each type of tissue were sampled in the analysis. The total RNA was extracted from the tissues by RNA Simple Plant Kit. The RNA concentration in each tissue was in the range of 34.3–398.9 ng/µl, with A260/280 values ranging from 2.01 to 2.14, indicating that the extracted RNA was of high quality. The primer tool in Geneious (Kearse et al., 2012) was employed to design specific PCR primers. In Supplementary Table 1 , the gene-specific primers for the three candidate genes are given together with the associated internal reference genes. The cDNA synthesis and qPCR were carried out using the Vazyme/R223 and Yeasen/11202ES03 kits, respectively. RT−qPCR was carried out to confirm the precise expression of Class II genes in the roots, stems, leaves and flowers using CIN genes in C. goeringii (GL10643, GL16029 and GL28896). Three biological replicates and three technical replicates were used in each experiment. The relative gene expression was calculated using the 2^−ΔΔCT method.

The protein file of the C. goeringii genome was used to search against the eggNOG 5.0 database using EggNOG-mapper v2¹¹ (Huerta-Cepas et al., 2019) for Gene Ontology (GO) functional annotation. Sequence similarity was used to predict function, sequence alignment was used to predict orthology, E-values and bit scores were used to filter out low-quality orthology alignments, and GO annotation terms associated with proteins involved in well-known biological processes were used to classify functions.

In total, 14 full-length TCP genes were identified from the C. goeringii genome. The Supplementary Data Sheet contained the complete protein sequences for TCP. InterproScan 5 (Jones et al., 2014) was used to confirm that all potential TCP genes encode the conserved TCP domain. The 14 TCP genes had an average length of 323 aa (range 169–572 aa), which was located on eight chromosomes. The average molecular weight (MW) was 35321.21 kDa (range 19216.05–63038.55 kDa). The grand average of hydrophilic (GRAVY) values were all negative in C. goeringii TCP proteins, suggesting genes with strong hydrophilicity. The average isoelectric point (pI) of the TCP proteins was 8.5 (range 6.60–10.55), indicating relatively strong acidity. The results of the subcellular location predictions revealed that all TCP proteins were found in the nucleus, except for one gene (GL08210), suggesting that they mostly function as TFs in the nucleus ( Table 1 ).

TCP genes of four species (Arabidopsis thaliana, Phalaenopsis equestris, Dendrobium catenatum and C. goeringii) were used to conduct the phylogenetic analysis. In total, 87 genes in four species were classified into two clades ( Figure 1 ). In total, 14 genes of C. goeringii were divided into two clades, Class I (PCF genes) and Class II (CIN genes and CYC/TB1 genes). Similar to other orchids, three types of genes exist in C. goeringii as follows: four PCF genes, nine CIN genes and one CYC gene. We also performed the TCP proteins phylogenetic analysis of 18 different plant phylostrata groups species, results showed Class I and Class II (Class II-A, B, C) were divided ( Supplementary Figure 1 ; Supplementary Table 2 ). The TCP protein of algae only found in Class I and orchids TCP proteins were scattered in different clades.

We also analyzed the sequence alignment of CgTCP genes. The TCP domain was present in all CgTCP genes, and the domain can be divided into four parts ( Figure 2A , Supplementary Figure 2 ). The basic region was the most conserved, followed by two helix regions, and the loop region was more varied than the other regions. The TCP domain of the CgTCP proteins was similar to that in P. equestris and D. catenatum (Lin et al., 2016; Zhang et al., 2021a), indicating that the TCP domain is highly conserved in orchids. One CgTCP protein (CgCYC1) exited an R domain compared with CgCIN9 ( Figure 2B ), and the R domain was only present in the CYC gene, which is consistent with the phylogenetic results.

The gene location on the chromosome of C. goeringii was examined, and 14 TCP genes were located on eight chromosomes ( Figure 3A ). In addition, a collinearity analysis was conducted among the TCP genes of C. goeringii, P. equestris, and D. catenatum. The collinear relationship among the TCP genes was examined to identify potential duplication events during TCP gene evolution in orchids. The results demonstrated a nearly 1:1 correspondence between all TCP genes in the three orchids, indicating minimal genomic rearrangements and TCP orthologs reshuffling after the lineages of Dendrobium and Cymbidium diverged ( Figure 3B ). The results showed that gene pseudogenization and loss might lead to a decrease in TCP in C. goeringii.

To explore the gene structure of TCP genes in orchids, the intron–exon and up/downstream regulatory element distributions of C. goeringii, P. equestris and D. catenatum were analyzed ( Figure 4A ). The results revealed that the TCP family of C. goeringii had 1–5 exons and 0–4 introns, P. equestris had 1–3 exons and 0–2 introns, D. catenatum had 1–2 exons and 0–1 intron. Although similarity in the gene structure was found in each subclade, the TCP genes of C. goeringii have a high degree of variance in the intron length and exon numbers in comparison with A. thaliana, P. equestris and D. catenatum. In general, most Class I genes exhibited fewer exons than the Class II genes. Notably, in Class I, CgPCF1, CgPCF2 and CgPCF4 have no exons, whereas in Class II, CgCIN2 and CgCIN4 have no exons ( Figure 4A ).

The motifs of the TCP proteins in C. goeringii, P. equestris and D. catenatum were analyzed by the MEME, and 10 motifs were set as upper bounds ( Figure 4B ). The number of TCP motifs both ranged from three to six in C. goeringii and D. catenatum, and ranged from two to six in P. equestris. Motif 1 encoded the TCP domains ( Figure 4C ). The PCF proteins all contain motif 1, motif 2 and motif 4, CIN protein all contain motif 1 and motif 3, the PCF proteins were more conserved compare with CIN and CYC proteins by motif analysis.

The 2,000 bp upstream regions of C. goeringii, P. equestris and D. catenatum TCP genes were extracted for the identification of putative cis-elements and used to explore the promotor region functions. In total, 1644, 2250, 2440 cis-acting elements attributed to 26, 26, 29 types and 9, 9, 10 responsive functions were identified in C. goeringii, P. equestris and D. catenatum, respectively ( Figure 5A and Supplementary Table 3 ). TATA-box made up the majority of these components (67.76%, 52.18%, 46.31%), followed by CAAT-box (11.25%, 12.80%, 15.49%) in C. goeringii, P. equestris and D. catenatum, respectively ( Supplementary Table 4 ). The cis-element functions included phytohormone responsiveness to auxin, abscisic acid (ABA), gibberellin (GA), methyl jasmonate (MeJA) and salicylic acid; stress responsiveness, such as anoxic, anaerobic, drought and low-temperature stress; and growth and development elements, such as light response, cell cycle regulation, endosperm expression, meristem expression and circadian control ( Figure 5A ). Light responsiveness was the most prevalent element function in each TCP gene, suggesting the crucial functions that light plays in modulating TCP function throughout plant growth and development ( Figure 5A ). The second and third most prevalent types were MeJA-responsive and ABA-responsive elements, which were likewise broadly distributed in the majority of orchid TCPs ( Supplementary Table 5 ). These results suggest that these elements may have roles in controlling these two phytohormones. We also discovered other components involved in meristem expression and meristem-specific activation, which is in line with the crucial functions played by TCP in the preservation of meristem homeostasis.

A GO analysis was performed to annotate the gene functional classifications of the TCP genes of C. goeringii and investigate the important biological processes ( Figure 5B ). As a result, the GO terms “cellular developmental process”, “intracellular anatomical structure” and “DNA-binding transcription factor activity” were the most related to plant growth and development in the GO ontologies “Biological Process,” “Cellular Component,” and “Molecular Function,”, respectively ( Figure 5B and Supplementary Table 6 ). The findings are consistent with the fact that TCP is a premier regulator involved with a number of downstream transcriptional networks and functions mostly acts in the nucleus ( Table 1 ). Several other terms, such as “circadian rhythm” and “seed germination”, are consistent with TCP’s function.

Based on transcriptome data from C. goeringii tissues, including flowers, leaves, stems, and roots, an expression analysis was carried out. The expression profile revealed that Class I TCP genes were expressed at lower levels in differentiating tissues and mature organs, while the CgPCF2 gene was expressed at moderate levels in various tissues ( Figure 6A ). Among the Class II genes, three CIN genes (CgCIN2, CgCIN5 and CgCIN6) exhibited high expression levels in flower organs ( Figure 6A , Supplementary Table 7 ). In addition, the CgCYC1 gene was only expressed in the stem, indicating that the Class II TCP genes function in plant development.

Flowers showed strong expression of class II genes such CgCIN2, CgCIN5, and CgCIN6 ( Figure 6A ), indicating high expression and their potential significance in driving tissue differentiation. The gene expression of CIN homologs in C. goeringii was studied further in order to elucidate the precise functions of the Class II TCP genes by analyzing in the leaves, flowers, and stems using RT−qPCR ( Figure 6B ; Supplementary Table 8 ). The CgCIN2 showed high expression in the sepals and lips, the CgCIN5 showed high expression in the lips and columns, the CgCIN6 showed extremely high expression in columns, of three examined CIN genes were barely detected in the stems and roots ( Figure 6B ). To understand the underlying functions of the TCP gene in C. goeringii, a functional study of these two classes will be a crucial next step.