Unveiling kiwifruit TCP genes: evolution, functions, and expression insights

ABSTRACT

The TEOSINTE-BRANCHED1/CYCLOIDEA/PROLEFERATING-CELL-FACTORS (TCP) gene family is a plant-specific transcriptional factor family involved in leaf morphogenesis and senescence, lateral branching, hormone crosstalk, and stress responses. To date, a systematic study on the identification and characterization of the TCP gene family in kiwifruit has not been reported. Additionally, the function of kiwifruit TCPs in regulating kiwifruit responses to the ethylene treatment and bacterial canker disease pathogen (Pseudomonas syringae pv. actinidiae, Psa) has not been investigated. Here, we identified 40 and 26 TCP genes in Actinidia chinensis (Ac) and A. eriantha (Ae) genomes, respectively. The synteny analysis of AcTCPs illustrated that whole-genome duplication accounted for the expansion of the TCP family in Ac. Phylogenetic, conserved domain, and selection pressure analysis indicated that TCP family genes in Ac and Ae had undergone different evolutionary patterns after whole-genome duplication (WGD) events, causing differences in TCP gene number and distribution. Our results also suggested that protein structure and cis-element architecture in promoter regions of TCP genes have driven the function divergence of duplicated gene pairs. Three and four AcTCP genes significantly affected kiwifruit responses to the ethylene treatment and Psa invasion, respectively. Our results provided insight into general characters, evolutionary patterns, and functional diversity of kiwifruit TCPs.

KEYWORDS:

• Kiwifruit
• TCP gene family
• Actinidia chinensis
• Actinidia eriantha
• whole-genome duplication
• expression profiles
• conserved motif

1. Introduction

The TEOSINTE BRANCHED 1, CYCLOIDEA, and PROLIFERATING CELL FACTORS (TCP) constitute a plant-specific transcription factor family that was initially reported in 1999, influencing various aspects of plant growth and development by modulating cell growth and proliferation.^1–3 TCP proteins are characterized by the presence of a conserved TCP domain located at the N-terminus.^4,5 This TCP domain is a 59-amino-acid basic helix – loop – helix (bHLH) motif, playing roles in DNA binding, protein-protein interaction, and nuclear targeting.^4,5 Based on sequence similarity and diversification of the TCP domain, the TCP family can be categorized into two classes: class I (PCF class) and class II (TCP-C class).^2,4,6–8 In contrast to the TCP-C class proteins, PCF class proteins feature a four-amino-acid deletion in the TCP domain.^4,5 The TCP-C class can be further subdivided into two subclasses, CIN and CYC.^2–4 Besides the TCP domain, certain TCP-C class members contain an additional 18–20-residue conserved domain (R domain), forming a hydrophilic α-helix or a coiled-coil structure that facilitates protein-protein interactions.³ Most CYC subclass members possess a conserved ECE motif (a glutamic acid-cysteine-glutamic acid stretch) with an uncharacterized function.²

Previously, researchers have reported the regulatory role of TCPs in various aspects of plant growth and development, such as flower development,^9,10 seed germination,^11,12 and responses to stress.¹³ The class I TCP members usually promote cell differentiation and plant growth.¹⁴ In Arabidopsis, AtTCP14, and AtTCP15 from class I can regulate embryonic growth during seed germination by activating the gibberellin-dependent cell cycle.¹¹ AtTCP15 can also promote flowering by increasing the expression level of SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1).¹⁵ RNA-interference of the Arabidopsis AtTCP16 gene results in abortion of early pollen development.¹⁶ Overexpression of the AtTCP16 gene has been shown to induce the formation of ectopic meristem.¹⁷ Members of the class I TCP identified in other species have exhibited diverse functions.^18–20 In peach, virus-induced silencing of PpTCP.A2 increased ethylene production and induced fruit riping.¹⁹ In tomato, three TCP genes (SlTCP12, SlTCP15, and SlTCP18) were specifically expressed in the fruit, indicating their involvement in the tomato fruit ripening.²⁰ Another study demonstrated that the strawberry FaTCP11 gene affected ripening-related processes and flavan-3-ols synthesis.¹⁸

In contrast, most class II TCP genes inhibit cell differentiation and plant growth.^21,22 Five CIN-subclass AtTCP genes (AtTCP2, AtTCP3, AtTCP4, AtTCP10, and AtTCP24) targeted by microRNA miR319 were involved in regulating petal growth and development in Arabidopsis.⁹ Another CIN-subclass AtTCP3 gene increased flavonoid biosynthesis and has regulated silique development in Arabidopsis.^23,24 Two CYC-subclass AtTCP genes (AtTCP12 and AtTCP18) in Arabidopsis suppressed bud outgrowth.²⁵ The CYC-subclass AtTCP1 gene regulated plant growth and development by interfering with the expression level of the DWARF4 gene to affect brassinosteroid (BR) biosynthesis.²⁶ Functions of class II TCP members were also investigated in other species.^27,28 The strawberry FvTCP9 significantly affected the expression pattern of a series of genes associated with fruit development and ripening.²⁸ Two tomato CYC-subclass TCP genes (SlTCP7 and SlTCP9) suppressed axillary bud initiation and outgrowth.²⁷

Kiwifruit has gained a global popularity due to its high vitamin C content and abundance of minerals.^29,30 Belongs to the Actinidia genus, kiwifruit comprises 54 species and 75 taxa.³¹ The whole genome of the A. chinensis (Ac) and A. eriantha (Ae) have been previously reported,^22,23 revealing distinct flowering times and cell development patterns in both species.^31,32 While the TCP gene family is well-documented for its crucial role in plant growth and cell development,^1,2 a systematic investigation and functional analyzes of the TCP gene family in kiwifruit have not been reported to date.

In the present study, we conducted a comprehensive identification of the TCP gene family members from the genome of A. chinensis and A. eriantha. Our work represents the first report on the gene structure, motif compositions, chromosomal distributions of the TCP gene family for both kiwifruit species. Further, we analyzed and compared the phylogenetic relationships and evolution patterns of the TCP gene family. Cis-elements were examined, and expression patterns in different tissues and under different stress conditions were investigated. The results obtained from our study provide crucial information regarding the structural characteristics, evolutionary patterns, and potential functions of the TCP genes in the two kiwifruit species.

2. Materials and methods

2.1. Gene identification and analysis

We retrieved the genome and protein sequences of two kiwifruit species (A. chinensis and A. eriantha) from the Kiwifruit Genome Database (KIR) (http://kiwifruitgenome.org/). All the AtTCP protein sequences were obtained from the TAIR website (https://www.arabidopsis.org/). The candidate genes from AcTCPs and AeTCPs were identified using the software HMMER 3.0 based on the Hidden Markov Model (HMM) of the TCP domain profile (PF03634). Further, we employed the Conserved Domain Database (CDD) (https://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.html) and the simple modular architecture research tool (SMART) (http://smart.embl.de/) to confirm the conserved TCP domain of the candidate TCP proteins and the candidate TCP proteins containing the TCP domain were obtained and used for further analysis.

2.2. Analysis of kiwifruit TCP protein structure

The protein length, theoretical isoelectric point (pI), grand average of hydropathicity (GRAVY), and molecular weight (MW) of the TCP gene family in the two kiwifruit species were computed using the ProtParam on ExPASy server (http://web.expasy.org/protparam/). The subcellular localization of kiwifruit TCP proteins was predicted using the online web software CELLO (v2.5, http://cello.life.nctu.edu.tw/).

2.3. Gene structure, motif analysis, and chromosomal distribution of kiwifruit TCPs

The genome sequences and coding sequences of the TCP genes in the two kiwifruit species were obtained using the TBtools.³³ The structures of TCP genes were investigated using The gene structure display server (GSDS 2.0, http://gsds.cbi.pku.edu.cn/). The conserved motifs of TCP protein were identified using Multiple expectation maximization for motif elicitation tool (MEME, http://meme-suite.org/tools/meme) with a maximum of 10 motifs.³⁴ The genome locations of TCP genes were extracted from the corresponding GFF file using an in-house Perl script, and the chromosomal distributions were rendered using MapGene2 Chrome (http://mg2c.iask.in/mg2c_v2.0/).

2.4. Multiple sequence alignments and phylogenetic analysis of kiwifruit TCPs

The multiple sequence alignments of TCP proteins from A. thaliana, A. chinensis, and A. eriantha were performed using ClustalX with default parameters.³⁵ The phylogenetic tree was constructed by MEGA X software using the neighbor-joining (NJ) method with a bootstrap value of 1000.³⁶

2.5. Duplications and syntenic analysis of kiwifruit TCPs

To identify gene duplication of kiwifruit TCPs, the whole gene sequences of A. chinensis and A. eriantha were aligned using BLASTP with an e-value of 1 × 10⁻¹⁰. The duplication patterns of kiwifruit TCPs were identified using the MCScanX software with default parameters.³⁷ The synonymous (Ks) and nonsynonymous (Ka) mutation rates of the duplicated TCP gene pairs were calculated using TBtools software.³³ The syntenic analysis of kiwifruit TCPs was conducted using the MCScanX software with default parameters to produce the collinearity blocks across the whole genome.³⁷ The collinearity gene pairs of kiwifruit TCPs were visualized using TBtools.³³

2.6. Cis-elements analysis in the promoter region of kiwifruit TCPs

To analyze cis-elements involved in regulating TCP genes, the 2000-bp promoter sequences upstream of AcTCP genes in kiwifruit were obtained using the TBtools software,³³ and cis-elements were predicted and obtained from the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).³⁸

2.7. Expression analysis of kiwifruit TCPs

To investigate the expression patterns in different tissues, developmental stages, or stress treatments, we collected five published RNA-seq data (PRJNA187369, PRJNA277383, PRJNA328414, PRJNA514180, and PRJNA691387) from NCBI (https://www.ncbi.nlm.nih.gov/). We further re-analyzed these transcriptome data using the genome of the ‘Red5’ cultivar as reference genome.^31,32 The reads alignment was performed using the HISAT2 software (v2.0.1),³⁹ and the transcripts were assembled and quantified using the STRINGTIE software (v2.1.5).⁴⁰

2.8. Protein structure prediction of kiwifruit TCP protein

We first retrieved full-length protein sequences of AcTCPs using TBtools. The three-dimensional models of AcTCP proteins were predicted by Phyre2 web (http://www.sbg.bio.ic.ac.uk/phyre2/html/page. cgi?id=index) with default parameters.⁴¹

3. Results

3.1. Identification of TCP gene family in kiwifruit

To identify TCP proteins in kiwifruit, we employed the software HMMER 3.0 to search TCP proteins from genomes of Ac and Ae based on the Hidden Markov Model (HMM) of the TCP-domain profile (PF03634).⁴² In total, we identified 40 and 26 putative TCP family members from Ac (referred to as AcTCP) and Ae (referred to as AeTCP) genomes, respectively (Figure 1 and Table S1). Further, we confirmed the presence of the TCP domain in AcTCP and AeTCP using PFAM and Conserved Domain Database (CDD).^43,44 Our results showed that all putative TCP proteins in Ac and Ae contained the conserved TCP domain (Figure 1). Besides the TCP domain, several TCP proteins in Ac and Ae possessed other conserved domains (Figure 1). The coding sequence (CDS) length of AcTCPs and AeTCPs ranged from 480 bp (AcTCP34) to 1383 bp (AcTCP24) and from 441 bp (AeTCP22) to 2103 bp (AeTCP11) (Table 1). The putative AcTCPs and AeTCPs encoded proteins ranged from 160 to 461 amino acid (aa) and 147 to 701 aa in length, respectively (Table 1). The predicted molecular weight of AcTCP and AeTCP proteins ranged from 17.27 to 48.31 kDa and 16.02 to 77.02 kDa, respectively (Table 1). Moreover, the theoretical isoelectric point (pI) for AcTCP proteins varied from 5.20 to 10.33, and for AeTCP proteins ranged from 4.45 to 11.83 (Table 1). The subcellular localization of kiwifruit TCP proteins was predicted, and all AcTCP and AeTCP proteins were localized in the nucleus of the plant cell (Table 1).

Figure 1. Conserved domain of kiwifruit TCP genes predicted by CDD (a) and SMART (b).

The Red and blue rectangle showed kiwifruit TCP genes identified in A. chinensis (Ac) and A. eriantha (Ae), respectively.

Table 1. Protein composition and physiochemical characteristics of kiwifruit TCPs.

Species	Name	Gene_id	CDS length	Peptide residue	MW (kD)	pI	GRAVY	Subcellular location
A. chinensis	AcTCP01	Acc01442.1	1011	337	35.47	7.03	−0.30	Nuclear
	AcTCP02	Acc02302.1	957	319	36.27	6.61	−0.77	Nuclear
	AcTCP03	Acc02509.1	825	275	29.28	9.21	−0.64	Nuclear
	AcTCP04	Acc02707.1	1029	343	35.75	5.20	−0.29	Nuclear
	AcTCP05	Acc03187.1	528	176	18.63	10.33	−0.24	Nuclear
	AcTCP06	Acc04055.1	954	318	35.89	6.15	−0.75	Nuclear
	AcTCP07	Acc05178.1	1071	357	37.82	7.00	−0.66	Nuclear
	AcTCP08	Acc05559.1	810	270	28.81	9.13	−0.56	Nuclear
	AcTCP09	Acc05857.1	813	271	30.32	9.47	−0.61	Nuclear
	AcTCP10	Acc06329.1	909	303	31.89	7.88	−0.43	Nuclear
	AcTCP11	Acc06470.1	1275	425	46.11	7.76	−0.73	Nuclear
	AcTCP12	Acc06697.1	1365	455	47.97	6.92	−0.62	Nuclear
	AcTCP13	Acc09084.1	939	313	33.47	9.54	−0.43	Nuclear
	AcTCP14	Acc13889.1	669	223	24.24	5.81	−0.55	Nuclear
	AcTCP15	Acc14596.1	966	322	34.40	9.64	−0.39	Nuclear
	AcTCP16	Acc15732.1	1044	348	36.56	8.53	−0.24	Nuclear
	AcTCP17	Acc18253.1	1137	379	40.26	7.91	−0.57	Nuclear
	AcTCP18	Acc18647.1	972	324	36.89	8.73	−1.01	Nuclear
	AcTCP19	Acc19119.1	1062	354	38.95	6.24	−0.72	Nuclear
	AcTCP20	Acc19964.1	1044	348	36.73	6.54	−0.28	Nuclear
	AcTCP21	Acc20454.1	597	199	21.11	6.06	−0.20	Nuclear
	AcTCP22	Acc20688.1	888	296	33.14	5.86	−0.66	Nuclear
	AcTCP23	Acc20776.1	1257	419	44.98	7.82	−0.75	Nuclear
	AcTCP24	Acc21269.1	1383	461	48.31	6.57	−0.60	Nuclear
	AcTCP25	Acc21859.1	987	329	36.98	6.14	−1.07	Nuclear
	AcTCP26	Acc21895.1	1140	380	42.53	8.79	−0.69	Nuclear
	AcTCP27	Acc25137.1	879	293	32.73	5.82	−0.80	Nuclear
	AcTCP28	Acc26440.1	1110	370	40.07	5.91	−0.72	Nuclear
	AcTCP29	Acc26616.1	975	325	36.93	8.33	−1.09	Nuclear
	AcTCP30	Acc26652.1	1101	367	40.77	9.14	−0.62	Nuclear
	AcTCP31	Acc28314.1	738	246	26.12	9.71	−0.37	Nuclear
	AcTCP32	Acc28550.1	1086	362	40.59	8.97	−0.93	Nuclear
	AcTCP33	Acc29857.1	1140	380	40.20	7.31	−0.52	Nuclear
	AcTCP34	Acc30140.1	480	160	17.27	8.82	−0.26	Nuclear
	AcTCP35	Acc30510.1	675	225	23.54	6.69	−0.45	Nuclear
	AcTCP36	Acc30781.1	879	293	33.27	9.16	−0.79	Nuclear
	AcTCP37	Acc30834.1	981	327	36.49	9.34	−0.59	Nuclear
	AcTCP38	Acc31269.1	1044	348	39.21	6.05	−0.95	Nuclear
	AcTCP39	Acc31364.1	762	254	26.73	9.72	−0.32	Nuclear
	AcTCP40	Acc32197.1	489	163	17.83	10.17	−0.85	Nuclear
A. eriantha	AeTCP01	DTZ79_02g14520	849	283	29.65	6.00	−0.22	Nuclear
	AeTCP02	DTZ79_03g01820	816	272	28.26	9.86	−0.22	Nuclear
	AeTCP03	DTZ79_03g07410	471	157	16.95	10.68	−0.62	Nuclear
	AeTCP04	DTZ79_03g07430	522	174	18.40	9.18	−0.32	Nuclear
	AeTCP05	DTZ79_05g08630	1011	337	37.98	8.68	−0.76	Nuclear
	AeTCP06	DTZ79_06g03830	912	304	32.11	7.88	−0.43	Nuclear
	AeTCP07	DTZ79_06g05430	1275	425	46.03	8.21	−0.73	Nuclear
	AeTCP08	DTZ79_06g09390	1782	594	63.13	9.12	−0.58	Nuclear
	AeTCP09	DTZ79_08g09290	918	306	32.93	10.13	−0.45	Nuclear
	AeTCP10	DTZ79_12g11690	693	231	25.39	8.73	−0.65	Nuclear
	AeTCP11	DTZ79_14g05220	2103	701	77.02	9.10	−0.22	Nuclear
	AeTCP12	DTZ79_16g08010	1149	383	40.58	9.83	−0.50	Nuclear
	AeTCP13	DTZ79_16g09430	828	276	31.21	9.40	−0.70	Nuclear
	AeTCP14	DTZ79_16g13170	1215	405	43.47	9.33	−0.75	Nuclear
	AeTCP15	DTZ79_18g14200	471	157	16.02	4.45	0.18	Nuclear
	AeTCP16	DTZ79_19g04170	1518	506	53.55	9.13	−0.53	Nuclear
	AeTCP17	DTZ79_19g10800	984	328	36.76	6.68	−1.03	Nuclear
	AeTCP18	DTZ79_19g11290	1077	359	40.43	8.73	−0.55	Nuclear
	AeTCP19	DTZ79_19g16560	936	312	35.01	9.67	−0.79	Nuclear
	AeTCP20	DTZ79_23g13830	1113	371	41.25	7.17	−0.67	Nuclear
	AeTCP21	DTZ79_23g19240	1017	339	36.68	6.46	−0.81	Nuclear
	AeTCP22	DTZ79_25g00720	441	147	16.13	11.87	−0.69	Nuclear
	AeTCP23	DTZ79_26g00970	963	321	36.81	9.07	−0.98	Nuclear
	AeTCP24	DTZ79_26g06930	1200	400	42.79	8.66	−0.44	Nuclear
	AeTCP25	DTZ79_27g03390	939	313	35.40	8.36	−0.72	Nuclear
	AeTCP26	DTZ79_27g04050	834	278	31.05	9.73	−0.70	Nuclear

3.2. Phylogenetic analysis of kiwifruit TCP gene family

To explore the phylogenetic relationship and evolutionary pattern of TCP genes in kiwifruit, the phylogenetic tree was constructed by neighbor-joining (NJ) method for the full-length protein sequences of the identified 40 AcTCPs, 26 AeTCPs, and 24 AtTCPs in Arabidopsis. Consistent with previous reports in Arabidopsis and other species,^45–48 both AcTCPs and AeTCPs were classified into two classes (class I and II) (Figure 2). 21 out of 40 AcTCPs and 14 out of 26 AeTCPs were assigned in class I (Figure 2). Similarly, class II of AcTCPs and AeTCPs were grouped into two subclasses (the CIN and CYC subclass) (Figure 2) as reported previously in other plant species.¹ The CIN subclass contained 12 AcTCPs and 6 AeTCPs, while the CYC subclass included 7 AcTCPs and 6 AeTCPs (Figure 2). In comparison, AcTCPs and AeTCPs grouped with different TCP genes in Arabidopsis indicated that both AcTCPs and AeTCPs probably have diversified functions similar to TCP genes in Arabidopsis (Figure 2).

Figure 2. Phylogenetic tree of TCP proteins.

The full-length TCP protein sequences from Arabidopsis (At, black gene name and circles), A. chinensis (Ac, red gene name and circles), and A. eriantha (Ae, blue gene name, and circles) were aligned using ClustalX 2.0 with default parameters. Then, the unrooted phylogenetic tree was constructed using MEGA X and the Neighbor-Joining method. Class I and II were highlighted using red and blue sectors, respectively. Light green and light red clades indicated the CIN and CYC subclass, respectively.

3.3. Chromosomal localization of kiwifruit TCPs

The 40 AcTCP genes were randomly distributed on 19 chromosomes of Ac (Figure 3(a)). Our results showed that chromosome 23 and 27 included the most number of TCP genes (five AcTCP genes on each chromosome), followed by chromosome 18 containing four AcTCP genes, and chromosomes 2, 3 and 6 contained three AcTCP genes (Figure 3(a)). Chromosome number four and nine contained two and one AcTCP genes, respectively (Figure 3(a)). Similarly, the 26 AeTCP genes were unevenly distributed on 14 chromosomes. Chromosome 19 contained the highest number of AeTCP genes (Figure 3(b)). The chromosome number three, six and 16 had three AeTCP genes in each (Figure 3(b)). Three chromosomes (chromosome 23, 26, and 27) contained two AeTCP genes each, and the rest of the seven chromosomes contained one AeTCP gene (Figure 3(b)).

Figure 3. Distribution of TCP genes in Ac (a) and Ae (b) chromosomes.

The length of the column represents the size of the chromosome. The numbers in the heatmap represent the number of TCP genes.

3.4. Gene structure and conserved motifs analysis of kiwifruit TCPs

The exon-intron structure is a primary evolutionary feature of a gene family and provides a clue for functional diversification and classification.⁴⁹ The exon number of AcTCPs and AeTCPs varied from one to 10 (Figures 4 and S1). However, the exon numbers of AeTCP genes were higher than that of AcTCP genes (Figures 4(a) and S1(a)). For AcTCP genes, both class I and CIN subclass had relatively similar exon numbers, while the exon number of CYC subclass genes varied from one to four (Figures 4(a) and S1(a)). However, exon numbers for both class I and II varied, indicating that AcTCPs and AeTCPs had discrepant gene structures (Figures 4(a) and S1(a)). The intron number of a gene notably regulates gene function by alternative splicing of transcripts.^50,51 Most AcTCP genes (34 out of 40 AcTCPs) were intronless, while a lower proportion of AeTCPs (7 out of 26 AeTCPs) was intronless, suggesting that AeTCPs potentially produced functional diversification of genes by alternative splicing (Figures 4(a) and S1(b)).

Figure 4. Gene structure and conserved motif architecture of TCP family in two kiwifruit species.

(a) Gene structures of TCP genes in two kiwifruit species. The left panel indicated the phylogenetic tree containing AcTCP and AeTCP proteins; the middle panel showed the ranges of kiwifruit TCP classification; the right panel showed gene structures of kiwifruit TCP genes. The green rectangle shows exons, the yellow rectangle shows UTRs, and the regular line represents introns. (b) Conserved domain arrangements of kiwifruit TCP proteins. Rectangles with different colors represented different conserved domains. (c) Conserved motif architecture of kiwifruit TCP proteins. Rectangles with different colors represented different conserved motifs.

The conserved domains and motif architectures of kiwifruit TCPs were predicted by Pfam and MEME, respectively (Figure 4(b,c)).³⁴ In total, ten conserved motifs (motif 1 to motif 10) were identified for the kiwifruit TCPs (Figure S2). The conserved motif number in each kiwifruit TCP gene varied from two to seven (Figure 4(c)). The motif number of kiwifruit TCP genes within class I ranged from two to seven, while it varied from two to five in class II (Figure 4(c)). All TCP genes in both kiwifruit species contained motif 1 and 2, and we confirmed that motif 1 and 2 constituted the conserved TCP domain with Pfam and CDD databases (Figure 4(b,c)). Furthermore, we identified several class-specific motifs in kiwifruit TCPs (Figure 4(c)). Our results showed that motif 3, 6, 7, and 10 were present in class I, while motif 4 and 5 were specific for class II (Figure 4(c)). Also, Motif 8 and 9 were specifically found in the CIN and CYC subclass, respectively (Figure 4(c)). The class-exclusive motifs potentially affected the functional diversification of kiwifruit TCPs. Similar to results of exon-intron structure, phylogenetically related TCP genes showed conserved motif structures, including motif number and organization, which indicated their similar functions (Figure 4(c)).

3.5. Synteny and duplicated gene analysis of kiwifruit TCPs

Gene duplication and loss are the main evolutionary forces driving the expansion or contraction of gene families, and duplicated genes can result in gene redundancy or new functionalization.⁵² To visualize the synteny relationships among homologous TCP genes and infer gene duplication events, we conducted a collinearity analysis using MCScanX.³⁷ Our results determined that gene pairs underwent five types of gene duplication (singleton duplication [SD], dispersed duplication [DD], proximal duplication [PD], tandem duplication [TD], and whole-genome duplication [WGD]). We identified 43 and 12 pairs of genes that resulted from duplication in Ac and Ae kiwifruit, respectively (Figure 5 and Table 2). All duplicated gene pairs were present in both classes (Figure 5 and Table 2). The production of all duplicated gene pairs by the whole-genome duplication illustrated that the WGD accounted for the expansion of kiwifruit TCP families (Table 2).

Figure 5. Chromosome distribution and synteny relationship of TCP genes in two kiwifruit species.

The blue and red bars indicated chromosomes for Ac and Ae, respectively. The syntenic gene pairs were connected by lines with different colors.

Table 2. TCP duplication events identified in kiwifruit.

Duplicated genes		Duplication types		Ka	Ks	Ka/Ks
Gene 1	Gene 2	Gene 1	Gene 2
AcTCP01	AcTCP04	WGD	WGD	0.04	0.24	0.16
AcTCP02	AcTCP30	WGD	WGD	0.40	2.74	0.15
AcTCP02	AcTCP26	WGD	WGD	0.38	2.88	0.13
AcTCP02	AcTCP36	WGD	WGD	0.23	0.76	0.30
AcTCP02	AcTCP06	WGD	WGD	0.08	0.13	0.60
AcTCP03	AcTCP35	WGD	WGD	0.12	0.79	0.15
AcTCP03	AcTCP08	WGD	WGD	0.10	0.86	0.12
AcTCP14	AcTCP22	WGD	WGD	0.46	1.50	0.31
AcTCP14	AcTCP27	WGD	WGD	0.50	2.80	0.18
AcTCP14	AcTCP28	WGD	WGD	0.46	1.50	0.30
AcTCP15	AcTCP13	WGD	WGD	0.05	0.24	0.20
AcTCP16	AcTCP20	WGD	WGD	0.07	0.25	0.28
AcTCP16	AcTCP01	WGD	WGD	0.39	1.94	0.20
AcTCP16	AcTCP04	WGD	WGD	0.39	1.89	0.21
AcTCP17	AcTCP33	WGD	WGD	0.02	0.30	0.07
AcTCP17	AcTCP10	WGD	WGD	0.32	1.71	0.19
AcTCP18	AcTCP25	WGD	WGD	0.28	1.12	0.25
AcTCP19	AcTCP28	WGD	WGD	0.18	0.85	0.22
AcTCP20	AcTCP01	WGD	WGD	0.41	1.62	0.25
AcTCP20	AcTCP04	WGD	WGD	0.43	1.92	0.22
AcTCP20	AcTCP13	WGD	WGD	0.38	2.53	0.15
AcTCP21	AcTCP05	WGD	WGD	0.07	0.33	0.22
AcTCP22	AcTCP27	WGD	WGD	0.09	0.24	0.35
AcTCP24	AcTCP12	WGD	WGD	0.03	0.29	0.11
AcTCP25	AcTCP29	WGD	WGD	0.06	0.23	0.27
AcTCP25	AcTCP09	WGD	WGD	0.48	2.41	0.20
AcTCP26	AcTCP30	WGD	WGD	0.04	0.18	0.25
AcTCP26	AcTCP36	WGD	WGD	0.37	1.24	0.30
AcTCP26	AcTCP06	WGD	WGD	0.36	3.15	0.11
AcTCP29	AcTCP32	WGD	WGD	0.49	2.34	0.21
AcTCP29	AcTCP37	WGD	WGD	0.52	2.40	0.22
AcTCP30	AcTCP36	WGD	WGD	0.38	1.30	0.30
AcTCP30	AcTCP06	WGD	WGD	0.38	2.15	0.18
AcTCP32	AcTCP38	WGD	WGD	0.09	0.21	0.43
AcTCP33	AcTCP10	WGD	WGD	0.26	1.44	0.18
AcTCP34	AcTCP40	WGD	WGD	0.82	1.41	0.58
AcTCP34	AcTCP05	WGD	WGD	0.58	1.74	0.34
AcTCP35	AcTCP08	WGD	WGD	0.04	0.14	0.28
AcTCP35	AcTCP10	WGD	WGD	0.62	1.86	0.34
AcTCP36	AcTCP06	WGD	WGD	0.20	0.75	0.27
AcTCP36	AcTCP09	WGD	WGD	0.84	2.97	0.28
AcTCP37	AcTCP09	WGD	WGD	0.10	0.26	0.38
AcTCP40	AcTCP05	WGD	WGD	0.53	1.00	0.52
AeTCP02	AeTCP11	WGD	WGD	0.37	1.70	0.21
AeTCP05	AeTCP26	WGD	WGD	0.08	0.22	0.37
AeTCP06	AeTCP24	WGD	WGD	0.28	1.41	0.20
AeTCP08	AeTCP16	WGD	WGD	0.04	0.24	0.18
AeTCP09	AeTCP11	WGD	WGD	0.48	2.25	0.21
AeTCP12	AeTCP24	WGD	WGD	0.18	0.54	0.33
AeTCP13	AeTCP17	WGD	WGD	0.35	1.32	0.26
AeTCP13	AeTCP23	WGD	WGD	0.11	0.28	0.41
AeTCP17	AeTCP23	WGD	WGD	0.27	0.95	0.28
AeTCP18	AeTCP20	WGD	WGD	0.11	0.28	0.39
AeTCP18	AeTCP25	WGD	WGD	0.43	1.66	0.26
AeTCP20	AeTCP25	WGD	WGD	0.41	1.48	0.27

We calculated the Ka/Ks ratio to estimate the selection pressure that kiwifruit TCPs experienced after the gene duplication (Table 2). Generally, the Ka/Ks value reflects the selection pressure during evolution (Ka/Ks = 1: neutral selection; Ka/Ks > 1: positive selection; Ka/Ks < 1: purifying selection).⁵³ In the present study, the Ka/Ks ratio of AcTCPs ranged from 0.07 to 0.60 with an average of 0.26, and the fluctuation range for AeTCPs was from 0.18 to 0.41 with an average of 0.28 (Table 2 and Figure S3). The Ka/Ks values for all TCPs were less than one, suggesting that purifying selection was the primary evolutionary force on kiwifruit TCPs.

3.6. Cis-element analysis in promoter regions of kiwifruit TCPs

The 2,000-bp upstream regions of TCP genes from Ac and Ae were extracted and employed for the cis-element prediction. In GGGtotal, 1,263 cis-acting elements attributed to 25 responsive functions were detected in the promoter of TCP genes. The number of cis-element in all TCP genes ranged from 8 (AcTCP38) to 29 (AeTCP08) (Figure 6 and Table S2). Except for AcTCP31, cis-elements related to light-responsiveness were the most abundant in the promoter regions of TPC genes of two species, indicating the crucial functions that light plays in modulating TCP function throughout plant growth and development, with implications for fruit quality and yield. Five MeJA-responsiveness cis-elements were found in the promoter regions of both AcTCP24 and AeTCP16, the highest number found among all gene members. Furthermore, the promoter regions of AcTCP24 and AeTCP16 both exhibited cis-elements responsive to abscisic acid, salicylic acid, and auxin (Figure 6 and Table S2). The high consistency of cis-acting elements observed in these gene members with close phylogenetic relationships suggests a conserved mechanism of gene expression regulation. Genes such as AcTCP16 and AcTCP15, which displayed a high frequency of auxin-responsive elements, suggesting a significant role in growth and development processes mediated by auxin (Figure 6 and Table S2). The diversity of these cis-elements suggested a complex network of hormone signaling that regulates the diverse developmental stages of kiwifruit. Promoter regions of several TCP genes, like AcTCP31, are enriched with cis-elements associated with drought responsiveness such as the MBS motif (Figure 6 and Table S2). This suggested a potentially pivotal role for these genes in drought stress tolerance. Moreover, the cis-elements associated with response to environmental stress, such as anaerobic induction (ARE), zein metabolism regulation (O2-site) and low-temperature responsiveness (LTR), were detected from different TCP genes (Figure 6 and Table S2). The differential distribution of stress-responsive elements suggested a functional specialization within the TCP family in kiwifruit, where certain members may have more prominent roles in stress response pathways.

Figure 6. Cis-acting elements on promoters of TCP genes in two kiwifruit species.

Based on the functional annotation, number and color are represented the number of corresponding cis-elements from each responsive function.

3.7. Expression patterns of kiwifruit TCPs in different tissues

Firstly, we obtained two transcriptome data to investigate expression patterns of AcTCP genes in different tissues (Figure 7). The first transcriptome data compared expression profiles of three tissues (leaf, immature fruit, and ripe fruit) of the Ac-originated cultivar ‘Hongyang’ (HY) (Figure 7(a)). The second transcriptome data investigated expression profiles of TCPs in eight tissues of the Ac-originated cultivar ‘Hort16A’ (Figure 7(b)). Our results showed that AcTCP genes had highly tissue-specific expression patterns in the cultivar of HY and Hort16A (Figure 7). Two AcTCPs (AcTCP03 and AcTCP23) were highly expressed in all three tissues (Figure 7(a)). Similar to HY, AcTCP03 and AcTCP23 were highly expressed in the four tissues (Leaf-sink, leaf, Fruit-T1 and Fruit-T2) (Figure 7(b)). Furthermore, AcTCP23 was also highly expressed in flower (Flower-bud and Flower) and root tissues, but its expression was low in shoot of Hort16A. AcTCP03 was highly expressed in all eight tissues from Hort16A, as were AcTCP39, AcTCP24, AcTCP31 and AcTCP12 (Figure 7(b)). AcTCP28 was highly expressed in seven other tissues except shoot in the Hort16A. AcTCP14 had higher expression level only in Leaf and Leaf-sink from Hort16A (Figure 7(b)). AcTCP15 exhibited higher expression level in flower-bud. In both HY and Hort16A, AcTCP19 showed low expression in fruit tissues compared with other tissues (Figure 7). This extensive tissue-specific expression suggested functional diversification within the TCP gene family after whole-genome duplication. We also found that gene pairs with a closer phylogenetic relationship exhibited divergent expression patterns in the first transcriptome data, indicating the functional diversification of AcTCPs (Figure 7(a)). For example, AcTCP01 and AcTCP04 had a close phylogenetic relationship and AcTCP04 has specifically expressed in kiwifruit leaf (Figures 2 and 7(a)). In contrast, AcTCP01 had an extremely low expression level in all three tissues (Figure 7(a)). However, several gene pairs with a close phylogenetic relationship had similar expression profiles in the first transcriptome, implying the functional redundancy of AcTCPs (Figures 2 and 7(a)). For instance, AcTCP19 and AcTCP28 formed duplicated gene pairs, and both genes were highly expressed in kiwifruit leaf and immature fruit (Table 2 and Figure 7(a)).

Figure 7. Expression profiles of AcTCP genes in different tissues.

The heatmap indicated FPKM values after log2 transformation (fragments per kilobase of exon model per million mapped reads) of AcTCP genes. (a) Expression profiles of AcTCPs in three different tissues of ‘Hongyang’ (HY). (b) Expression profiles of AcTCPs in seven different tissues of ‘Hort16A’.

3.8. Expression patterns of kiwifruit TCPs with the ethylene treatment

To further confirm whether hormonal treatments influenced the expression of kiwifruit TCP genes, we re-analyzed both transcriptome data to estimate expression profiles of AcTCPs in different stages of fruit riping and ethylene treatment for cultivar ‘Hort16A’ (Figure 8). Our results indicated that both transcriptome data were highly coherent (Figure 8). The expression profile revealed about ten AcTCP genes that were highly expressed in at least one or several stages of fruit riping (Figure 8(a,b)). Expression profiles of seven AcTCPs were changed under the ethylene treatment, and expression levels of three out of seven AcTCP genes were significantly changed (Fig. S4). Expression levels of AcTCP03 and AcTCP23 were significantly depressed with the ethylene treatment, indicating that both AcTCP genes were related to the delay of kiwifruit fruit riping (Figure. S4). AcTCP03 was the homologous gene of AtTCP20, which was found to delay the cell and leaf senescence in Arabidopsis (Figure 2).⁵⁴ From the above-mentioned results, we inferred that AcTCP03 potentially played a repressor role in kiwifruit fruit riping (Figure. S4). AcTCP23 was the homologous gene of AtTCP02, which was also related to leaf senescence, and we assumed that AcTCP23 might have a similar function to AcTCP03 (Figure 2). On the contrary, the expression level of AcTCP35 was extremely increased upon ethylene treatment (Fig. S4). AcTCP35 had a close phylogenetic relationship to AcTCP03 and had an opposite expression trend with the ethylene treatment, indicating that both genes produced functional diversities after duplication (Figures 2 and S4). The three-dimensional protein structure of AcTCP03, AcTCP23, and AcTCP35 were predicted by Phyre2.⁴¹ The protein structures were successfully modeled with a confidence level of over 95%. Although AcTCP03 and AcTCP35 had a close phylogenetic relationship, the protein structure of AcTCP03 was more similar to the protein structure of AcTCP23, indicating a similar function of AcTCP03 and AcTCP23, and the protein structure potentially determined the functional divergence of AcTCP03 and AcTCP35 (Figure 8(c)). We also compared the upstream promoter regions of these three AcTCP genes (Figure 6 and Table S2). We found two ethylene-responsive elements (ERE) in promoter regions of AcTCP03 and three ERE in AcTCP23, while there was no ERE present in the promoter region of AcTCP35 (Figure 6 and Table S2). These results suggested that differences in promoter regions potentially affected gene functions of AcTCP03, AcTCP23, and AcTCP35 in responses to the ethylene treatment.

Figure 8. Expression patterns of AcTcps in different stages of fruit development and the ethylene treatment for the ‘Hort16A’ cultivar.

(a) Expression profiles of AcTCPs in ‘Hort16A’. DAFB, days after the full bloom of fruit; DAT, days after ethylene treatment. (b) Expression profiles of AcTCPs in ‘Hort16A’. DAA, days after anthesis; DAT, days after ethylene treatment. (c) Predicted models of AcTCP03, AcTCP23, and AcTCP35 proteins. Models were visualized by rainbow colour from N to C terminus.

3.9. Expression patterns of kiwifruit TCPs under the invasion of Psa

Based on two transcriptome datasets, we further investigated AcTCPs responses to the invasion of kiwifruit bacterial canker disease pathogen Psa (Figure S5). The first transcriptome data investigated transcriptional responses of the susceptible cultivars ‘hongyang’ (HY) to the invasion of Psa (Figure. S5A). The expression level of about 14 AcTCPs significantly increased under Psa infection, indicating that AcTCP potentially regulated kiwifruit responses to the Psa invasion (Figure S5(a)). The second transcriptome data compared different expression perturbations of two kiwifruit cultivars with varying resistance to Psa (HT, highly resistant to Psa infection; HY, highly susceptible to the Psa infection) (Figure S5(b)). Expression profiles of 16 AcTCP genes exhibited divergent patterns in both kiwifruit materials with disparate resistance to the Psa infection (Fig. S5B). Expression levels AcTCP02 and AcTCP09 were significantly increased in HY while decreased in HT, suggesting that both AcTCPs potentially enhanced kiwifruit susceptibility to the invasion of Psa (Figure 9(a)). However, the expression of AcTCP06 and AcTCP12 was high in HT and low in HY, indicating their role in the positive regulation of kiwifruit resistance to Psa invasion (Figure 9(b)). AcTCP02 and AcTCP06 had a close phylogenetic relationship and sequence similarity, yet their expression patterns were diametrically opposite in HT and HY (Figures 2 and 9(b)). We first predicted protein structures of AcTCP02, AcTCP09, AcTCP06, and AcTCP12, and the results showed that AcTCP02 and AcTCP06 had a similar protein structure, indicating that the functional divergence between AcTCP02 and AcTCP06 was not determined by protein structure (Figure. S6). Next, we compared differences of promoter regions of AcTCP02 and AcTCP06 (Figure 9(c) and Table S2). Our results illustrated that cis-elements arrangements of promoter regions of AcTCP02 and AcTCP06 had significant differences (Figure 9(c) and Table S2). We found that two immune-responses-related cis-elements (AT-rich and TC-rich elements) were particularly present in the promoter regions of AcTCP02 (Figure 9(c) and Table S2). AT-rich element mediated maximal elicitor-mediated activation, and TC-rich element is a cis-acting element was involved in defense and stress responsiveness, suggesting that AcTCP02 regulated kiwifruit responses to the Psa infection (Figure 9(c) and Table S2). The cis-acting element involved in salicylic acid responsiveness (TCA-element) was only identified in the promoter region of AcTCP06, indicating that AcTCP06 could enhance kiwifruit resistance to Psa (Figure 9(c) and Table S2). The above results illustrated that the divergence of the cis-element architecture for AcTCP02 and AcTCP06 determined their functional diversification by regulating their different expression patterns under Psa infection.

Figure 9. Expression levels of four AcTcps in two kiwifruit materials (HT and HY).

HT had a high resistance to Psa infection and HY was susceptible to the invasion of Psa. (a) Expression levels of AcTCP02 and AcTCP09. (b) Expression levels of AcTCP12 and AcTCP06.

4. Discussion

The plant-specific TCP gene family regulates a wide range of biological processes throughout the whole life span of plants, primarily regulating plant growth and development, hormonal pathways crosstalking, and plant immunity.^{1,5,11,13,15,22,24,26,51,54,55} Genome-wide identification of the TCP gene family had been accomplished in several plants.^8,14,56–61 However, genome-wide characterization of the TCP gene family had not been conducted in kiwifruit. Here, we conducted the comprehensive genome-wide identification and characterization of the TCP gene family in two kiwifruit species (Ac and Ae), comparing their characters and evolutionary patterns. In addition, we investigated their expression dynamics and roles in response to hormonal treatment and pathogen infection. This study advances our understanding of the kiwifruit TCP gene family and provides genomic data for molecular breeding in improvement of quality and resistance.

4.1. TCP genes family members are widely distributed in kiwifruit

In total, 40 and 26 TCP genes were identified in Ac and Ae, respectively (Figure 1 and Table 1). Compared to the TCP family in Arabidopsis (24 AtTCP genes), the number of AcTCP genes was significantly higher (Figure 1 and Table 1). In contrast, the number of AeTCP genes was consistent with AtTCP genes (Figure 1 and Table 1). Based on the results of collinearity analysis, we identified 43 and 12 duplicated gene pairs that were entirely owing to WGD in Ac and Ae (Figure 5 and Table 2). Genomic analyzes verified that both Ac and Ae genomes experienced three ancient WGD events.^31,32 However, the difference in TCP gene number between Ac and Ae indicated that the TCP gene family in Ac and Ae had undergone inconsistent evolutionary patterns.^31,32 We inferred that translocation, gene retention, and gene loss post-whole-genome duplications were accounted for the expansion of AcTCPs in Ac and variations of the TCP gene number in Ac and Ae genomes.

4.2. Multiple factors affecting functional differentiation of TCP gene family in kiwifruit

Consistent with the results reported in other species, AcTCPs and AeTCPs were divided into two classes and class II was further grouped into two subclasses (CIN and CYC subclass) (Figure 2) [2,5,21]. Sequence similarity determined the classification of TCP genes, and TCP genes belonging to different groups had divergent functions, indicating that classification illustrated primary functional diversification of kiwifruit TCPs.^1,2 Conserved domain and motif analysis illustrated that the TCP domain was present in all AcTCPs and AeTCPs, and presence of class-specific motifs decided functional specializations of kiwifruit TCPs (Figure 4). Gene structure analysis showed that AeTCPs possessed more complex exon-intron structures than AcTCPs, suggesting that AcTCPs and AeTCPs experienced divergent evolutionary patterns and functional diversification (Figures 4(b) and S1). Our results confirmed that three-dimensional protein structure and cis-element architecture in the promoter region affected the function of kiwifruit TCPs assigned as sister clades in phylogenetic analysis (Figures 8 and 9). All in all, gene structure, motif organization, three-dimensional protein structure, and cis-element arrangement precisely controlled the functional divergence of kiwifruit TCPs.

4.3. AcTCPs regulated kiwifruit responses to the ethylene treatment and Psa invasion

Based on transcriptome datasets, we found that AcTCPs showed highly tissue-specific expression patterns, and tissue-specific expression could directly affect AcTCPs function.⁹ The differential expression of AcTCP genes under ethylene treatment revealed a complex regulatory network where AcTCP03 and AcTCP23 act as repressors of fruit ripening, likely through ethylene-responsive element in their promoter regions (Figure 8 and S4). This suggested that these genes, through their interaction with ethylene, play crucial roles in delaying the ripening process. Interestingly, AcTCP35, which lacks the ethylene-responsive element in its promoter region, exhibited an opposite expression pattern. AtTCPs were involved in the biosynthesis and signaling of salicylic acid (SA), jasmonic acid (JA), ethylene, abscisic acid (ABA), and auxin by interacting with relative proteins.¹ Additionaly, the promoter region of AcTCP03 contained gibberellin-responsiveness and MeJA-responsiveness cis-acting element (Figure 6). This indicated that AcTCP03 was likely involved in the regulation of fruit development processes mediated by gibberellin and jasmonic acid signaling pathways.

Previous studies verified that TCP proteins could be employed as plant pathogen effector targets.¹ Kiwifruit bacterial canker disease caused by Psa is a disaster for the worldwide kiwifruit industry.⁶² We utilized two different transcriptome datasets to investigate the effects of AcTCPs on kiwifruit resistance to the invasion of Psa (Figures S5 and 8). By comparing expression profiles of AcTCPs in two kiwifruit materials with varying resistance to Psa, we identified four AcTCPs (AcTCP02/09 and AcTCP06/12) that had discrepant responses to the invasion of Psa (Figure 8). The AcTCP02/09 had high expression levels in HT, while AcTCP06/12 highly expressed in HY, indicating that AcTCP02/09 and AcTCP06/12 antagonistically regulated kiwifruit resistance to the Psa infection (Figure 9). The divergence in expression patterns between closely genes with high sequence similarity and phylogenetic closeness, such as AcTCP02 and AcTCP06, were of particular interest. This suggested that the functional diversification between these genes is likely mediated at the level of gene regulation rather than structural variance. This hypothesis is further supported by the analysis of promoter regions, revealing distinct cis-element arrangements that correspond with their differential expression patterns. The presence of immune-response-related cis-elements in the promoter region of AcTCP02 and the identification of a salicylic acid-responsive cis-elements in the promoter of AcTCP06 illustrated the genetic basis for their functional diversification (Figure 9). These results suggested that AcTCP02 may be involved in a broad-spectrum defense response, potentially making the plant more susceptible to Psa by diverting resources from more targeted defense mechanisms, while AcTCP06 may specifically enhance resistance through pathways associated with salicylic acid, a key hormone in plant defense against biotrophic pathogens.^63,64

In the present study, we identified TCP genes from kiwifruit genomes. Furthermore, we identified the potential response of kiwifruit TCPs to hormonal treatments and biotic stress. Our research will lay a foundation for accelerating the genetic breeding of kiwifruits.

Author contributions

Zhicheng Wu designed the experiment and Donglin Li conducted the experiment. Haibo Li, Huimin Feng and Ping Qi contributed to transient expression assay and sample tissue collection. Donglin Li and Zhicheng Wu performed the data processing and drafted the manuscript. All authors have read and agreed the unpublished version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15592324.2024.2338985

Additional information

Funding

This study was supported by the Special Fund Project for Science and Technology Innovation Strategy of Guangdong Province [210901084530323] [230306166273758] and National Natural Science Foundation of China [No. 32302588].