Genome-wide identification and expression profiling of the CoVQ gene family in Camellia Oleifera

Abstract

Background

Valine-glutamine (VQ) proteins are key regulators of plant growth, development, and stress responses. Camellia oleifera Abel, a major woody oil crop in southern China, is frequently challenged by abiotic stresses (including low phosphorus, drought, chilling, and potassium deficiency) that severely limit its productivity. Understanding the evolutionary and functional diversification of the VQ gene family in this species could offer valuable insights for molecular breeding.

Results

Using the diploid C. oleifera reference genome (CON v2.0, 2.95 Gb), we identified 39 CoVQ genes through HMMER-based screening and domain validation. Phylogenetic analysis grouped them into eight subclades, with family expansion primarily driven by 12 segmental duplication events (66.7% of duplicated pairs), all under strong purifying selection (Ka/Ks < 1; mean = 0.27). Notably, 84.6% of CoVQ genes are intronless, a feature potentially linked to LTR retrotransposon-mediated chromosomal rearrangements (LTR content: 49.76%). Tissue-specific expression revealed high expression of CoVQ1/6/13 in sepals (5 ~ 120 FPKM), while CoVQ16/22/31 were preferentially expressed in stamens. Under abiotic stresses, CoVQ1 was upregulated 278.2-fold after 12 days of drought, CoVQ18 showed rapid induction (7.96-fold) during early potassium deficiency, and CoVQ27/28 exhibited coordinated responses across low phosphorus, drought, and chilling. Promoter analysis revealed that 94.9% of CoVQ genes contain stress-responsive (e.g., ARE, DRE) and hormone-related (e.g., ABRE, ERE) cis-elements. Notably, CoVQ27/28 harbor nine stress-related elements (nearly twice the family average of 4.3), suggesting their potential as multifunctional regulatory hubs.

Conclusions

This study presents a systematic analysis of the VQ gene family in C. oleifera, integrating evolutionary, structural, and expression data to prioritize promising candidates for future functional validation and potential use in molecular breeding for improved stress resilience.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08227-0.

Introduction

Climate change has led to an increasing frequency and intensity of extreme weather events, such as droughts and floods, posing significant threats to global agriculture and forestry. Understanding the molecular mechanisms underlying plant stress responses, particularly those mediated by transcription factors (TFs), is therefore essential for enhancing the resilience of staple crops (e.g., rice, wheat), economically important forest trees (e.g., poplar, birch), and ecologically and economically valuable woody oil species such as Camellia oleifera and Vernicia fordii [1]. Such knowledge is critical for ensuring global food and oil security and promoting sustainable forest management.

Valine-glutamine (VQ) proteins are a family of plant regulatory factors defined by a conserved VQ motif with the consensus sequence FxxhVQxhTG (where x represents any amino acid and h a hydrophobic residue). They were first identified as σ factor-binding protein 1 (SIB1) in Arabidopsis, where they play a key role in abiotic stress responses [2, 3]. Since then, VQ genes have been found across the plant kingdom, from bryophytes to angiosperms, and homologous sequences have also been detected in fungi and basal metazoans, suggesting that the VQ domain is an evolutionarily ancient module co-opted for regulatory functions in plants [4].

To date, comprehensive genome-wide analyses have identified and characterized VQ gene families in multiple plant species, including 34 members in Arabidopsis, 40 in rice, 74 in soybean, 61 in maize, 23 in grape, and 31 in coix [5–10]. These proteins are classified into subtypes based on variations in the C-terminal residues of their VQ motifs-such as VTG, LTS, and LTG-reflecting functional diversification across lineages. Notably, many VQ gene promoters contain W-box cis-elements (TTGAC[C/T]), enabling interactions with WRKY transcription factors to form WRKY-VQ regulatory modules. These modules play crucial roles in integrating developmental processes, abiotic stress responses, and hormone signaling pathways [11–13].

Functionally, VQ proteins regulate diverse aspects of plant biology. For example, AtVQ14 modulates seed size by influencing endosperm development [14], while heterologous expression of AtVQ21 (also known as MKS1) delays flowering in petunia [15]. Loss-of-function mutations in AtVQ8 result in pale-green leaves due to impaired chloroplast development [16]. In stress responses, overexpression of TaVQ14 or PeVQ28 enhances drought tolerance and salt resistance in Arabidopsis, respectively [17, 18], whereas AtVQ9 acts as a negative regulator of salt stress by suppressing the activity of WRKY8 [19]. In wheat, TaVQ27-5 A interacts directly with TaWRKY51L-3 A to repress its activation of TaNOX genes (TaNOX-3 A, TaNOX-3B, and TaNOX-3D), thereby modulating reactive oxygen species (ROS) accumulation and improving thermotolerance [20]. Furthermore, VQ proteins are involved in defense signaling: AtVQ12/VQ29 regulates jasmonic acid (JA) signaling through protein dimerization [21], while JAV1 (VQ22) fine-tunes plant immunity by mediating ubiquitin-dependent degradation of key defense components [22].

C. oleifera is a perennial woody oil tree widely cultivated in southern China. Its seed oil, rich in oleic acid and bioactive compounds, is highly valued for its nutritional and industrial applications and is often referred to as “Eastern Olive Oil” [23]. However, the productivity and oil quality of C. oleifera are frequently compromised by its high sensitivity to abiotic stresses, including drought, cold, and salinity [24]. Given the well-documented roles of VQ genes in stress adaptation, the recent availability of a high-quality C. oleifera genome [25] presents a timely opportunity to systematically characterize the CoVQ gene family.

In this study, we conducted a comprehensive genome-wide identification of VQ genes in C. oleifera. We performed phylogenetic, gene structure, conserved motif, and cis-regulatory element analyses to explore the evolutionary features of the CoVQ family. Furthermore, we integrated transcriptomic data to examine their expression profiles across different tissues and under multiple abiotic stress conditions. Our findings provide a solid foundation for future functional characterization of VQ genes in C. oleifera and their potential application in breeding stress-resilient varieties.

Materials and methods

Plant materials

The high-yield C. oleifera cultivar ‘Changlin23’ (National Certified Clone, No. S-SC-CO-009-2008), a widely cultivated tea-oil tree in southern China, was used in this study. Planting materials were obtained from the Jiangxi Provincial Center for Forestry Technology Promotion and Public Education, the official agency responsible for forestry extension in Jiangxi Province. The center maintains a germplasm bank of improved tea-oil camellia varieties, ensuring genetic authenticity and traceability. Living plants of ‘Changlin23’ are preserved at the Nanchang University Biological Garden (Nanchang, China), and genomic DNA samples are archived in our laboratory for reference. Roots, stems, mature leaves, floral organs (sepals, petals, stamens, pistils), and seeds at the mid-development stage were collected in late October 2024. All samples were immediately flash-frozen in liquid nitrogen and stored at − 80℃ until use.

Identification of CoVQ gene family

The diploid C. oleifera reference genome (CON v2.0, scaffold N50 = 180.0 Mb) [26] was downloaded from the GitHub repository (https://github.com/Hengfu-Yin/CON_genome_data). Initial screening for VQ domain-containing proteins was performed using HMMER 3.3.1 (hmmsearch) with the Pfam VQ domain model (PF05678; E-value threshold: 1 × 10⁻⁵). Candidate sequences were further validated using NCBI-CDD (v3.20; https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi), Pfam (v35.0; https://pfam.xfam.org/), and SMART (v9.0; http://smart.embl-heidelberg.de/) to confirm domain integrity. Sequences containing transmembrane domains or signal peptides were excluded. Protein isoelectric points (pI) and molecular weights (MW) were predicted using ExPASy ProtParam (https://web.expasy.org/protparam/), and transmembrane regions were analyzed with TMHMM 2.0 (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0).

Phylogenetic analysis

VQ protein sequences from Arabidopsis thaliana (TAIR), Oryza sativa (Phytozome), and other plant species (Ensembl Plants) were aligned using ClustalW2. A maximum-likelihood phylogenetic tree was constructed in MEGA7.0 with 1,000 bootstrap replicates and visualized using EVOLVIEW.

Gene structure, protein domain, and conserved motif analysis

Gene structures were inferred using the Gene Structure Display Server (GSDS 2.0; http://gsds.gao-lab.org/) based on GFF annotations. Conserved motifs within CoVQ proteins were predicted using the MEME suite (http://meme-suite.org/;defaultparameters).

Chromosomal localization and gene duplication analysis

Chromosomal positions of CoVQ genes were extracted from GFF3 files and visualized using TBtools. Tandem duplications were defined as gene pairs located within ≤ 200 kb on the same chromosome with > 70% sequence identity (BLASTp, E-value ≤ 1 × 10⁻¹⁰). Segmental duplications were identified using MCScanX (default parameters). Ka/Ks values for duplicated pairs were calculated with KaKs_Calculator 2.0. Collinear blocks between C. oleifera and four species (A. thaliana, O. sativa, C. sinensis, Populus trichocarpa) were visualized using Circos.

Promoter Cis-element analysis

Promoter regions (1,500 bp upstream of transcription start sites) were extracted and analyzed using PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/). Identified cis-elements were classified and graphically represented using TBtools (v1.068).

RNA-seq data analysis

Transcriptomic datasets for C. oleifera ‘Changlin23’ were retrieved from the NCBI SRA database under accession numbers: PRJNA973966 (tissue-specific expression), PRJNA915196 (chilling stress), PRJNA875963 (drought stress), PRJNA792896 (low phosphorus), and PRJNA623066 (low potassium). Gene expression levels were quantified as FPKM values, log₂-transformed as log₂(FPKM + 1), and visualized as heatmaps using the R heatmap package.

RT-qPCR validation

Nine CoVQ genes (CoVQ1, CoVQ6, CoVQ11, CoVQ14, CoVQ16, CoVQ18, CoVQ29, CoVQ31, and CoVQ39) exhibiting distinct expression patterns were selected for RT-qPCR validation. Total RNA extraction and qPCR procedures were performed according to the protocol described by Zhu et al. (2024), with CoGAPDH used as the reference gene for expression normalization [27]. Gene-specific primers (amplicon size: 80–150 bp; annealing temperature: 60 ± 1 °C) were designed using Primer3 (v0.4.0; Table S1).

Results

Identification and basic features of CoVQ genes

Using the diploid C. oleifera reference genome, 39 CoVQ genes were identified through HMMER screening combined with Pfam (PF05678), SMART, and CDD domain validation (Figure S1, Table S2). Multiple sequence alignment revealed conserved VQ motifs: 31 CoVQ proteins contained FxxxVQxLTG, five harbored FxxxVQxFTG (CoVQ6/8/9/13/23), one carried FxxxVQxVTG (CoVQ15), and two possessed FxxxVQxLTA (CoVQ24/25) (Fig. 1), consistent with VQ motif diversity in cucumber [28]. The CoVQ proteins exhibited considerable physicochemical variation: amino acid lengths ranged from 94 (CoVQ11) to 793 (CoVQ32), molecular weights from 10.2 kDa (CoVQ12) to 88.68 kDa (CoVQ32), and pI values from 4.69 (CoVQ14) to 10.77 (CoVQ39), with 56.4% (22/39) being alkaline (pI > 7). Despite the genome’s high heterozygosity (2.52%) and repetitive content (76.1%), stringent filtering using TMHMM and SignalP ensured reliable gene identification.

Fig. 1.

Multiple sequence alignment of the VQ domain in CoVQs. The FxxxVQxxTx motif is highly conserved. Based on the “xTx” residues of this conserved motif, CoVQ proteins were classified into four types: LTG, FTG, VTG, and LTA

Phylogenetic analysis of CoVQ proteins

A maximum-likelihood phylogenetic tree was constructed using MEGA7.0, incorporating VQs from C. oleifera (CoVQs), C. sinensis (CsVQs), Arabidopsis thaliana (AtVQs), and Oryza sativa (OsVQs) (Fig. 2). The CoVQ proteins clustered into eight groups (I ~ VIII), with Group VII containing the largest number (9), while Groups III ~ VI each had three members. This distribution contrasts sharply with that in tea (C. sinensis), where 50% of VQ genes (13/26) belong to Group VII, and Groups II, IV, and V are absent.

Fig. 2.

Phylogenetic tree of the VQ protein family from C. oleifera, A. thaliana, and O. sativa. Co: C. oleifera; At: A. thaliana; Os: O. sativa

Conserved motifs and gene structure of CoVQs

MEME analysis identified 20 conserved motifs, with each CoVQ protein containing 1–7 motifs (Fig. 3, Figure S2). All members contained the core Motif 1 (VQ domain), characterized by the consensus sequence FxxxVQxhTG (x: any amino acid; h: hydrophobic residue). Subgroup-specific motifs were observed: Motifs 2 and 4 were exclusive to Group VII, Motif 12 to Group VI, and Motif 20 to Groups III and V. Motif 14 was present only in partial Group I members, while Motifs 5/15/16 (Group VII), Motif 10 (Group II), and Motifs 7/11 (Group VIII) displayed subgroup-specific distributions. Members within the same phylogenetic group shared similar motif compositions, supporting functional conservation within subgroups.

Fig. 3.

Gene structure of CoVQs

Gene structure analysis revealed that 84.6% (33/39) of CoVQ genes lacked introns. Genes in Groups IV, VI, and VII were entirely intronless, whereas other groups contained 1 ~ 2 intron-containing members. Specifically, CoVQ13, CoVQ20, and CoVQ39 each had one intron, CoVQ37 had two, and CoVQ32 contained six (Fig. 3), suggesting widespread intron loss during evolution.

Collinearity and gene duplication analysis

CoVQ genes were distributed across 14 chromosomes (excluding Chr8), with Chr4, 9, and 11 each harboring five genes and Chr13 containing four (Figure S1). Chromosome length did not correlate with gene density (e.g., the shortest chromosome, Chr9, carried five genes), and genes on the same chromosome often belonged to different phylogenetic groups (e.g., four genes on Chr13 spanned Groups I, II, and V), indicating functional divergence independent of chromosomal location.

Gene duplication events-tandem and segmental-were analyzed using MCScanX. A total of 12 duplicated pairs were identified: eight segmental duplications (66.7%; e.g., CoVQ3/CoVQ30, CoVQ6/CoVQ13, CoVQ6/CoVQ23, CoVQ10/CoVQ27, CoVQ13/CoVQ23, CoVQ16/CoVQ22, CoVQ22/CoVQ33, CoVQ31/CoVQ33) and four tandem duplications (33.3%; e.g., CoVQ3/CoVQ4, CoVQ11/CoVQ12, CoVQ27/CoVQ28, CoVQ35/CoVQ36) (Fig. 4). Segmental duplication appears to be a major contributor to the expansion of the CoVQ gene family. Given that oil tea has undergone ancestral whole-genome duplication (WGD) events, some of these segmental duplicates may originate from such large-scale genomic duplications.

Fig. 4.

Synteny analysis of VQ genes. (A) Synteny analysis of interchromosomal relationships of CoVQ genes; (B) Synteny analysis of VQ genes between C. oleifera and others plant species. Red lines indicate collinearity of VQ genes, while gray lines represent collinearity of other genes in the genome

Ka/Ks analysis revealed strong purifying selection across all duplicated pairs (Ka/Ks < 1; Table 1). The lowest Ka/Ks value (0.09) was observed for CoVQ35/CoVQ36, while CoVQ13/CoVQ23 showed the highest (0.53). The mean Ka/Ks was 0.24 ± 0.09 (P < 0.01), indicating strong functional constraints on the duplicated CoVQ genes. While purifying selection appears to be the primary force maintaining their conservation, the retention of multiple paralogs over evolutionary time may also reflect incipient functional diversification—such as subfunctionalization or neofunctionalization—that could be explored through comparative expression and functional assays.

Table 1.

Ka, Ks and Ka/Ks values of duplicated CoVQ gene pairs

Gene 1	Distribution of gene 1	Gene 2	Distribution of gene 2	Ka	Ks	Ka_Ks
CoVQ30	chr11	CoVQ3	chr2	0.06638	0.54653	0.12145
CoVQ13	chr5	CoVQ6	chr3	NaN	NaN	NaN
CoVQ23	chr10	CoVQ6	chr3	0.42749	1.58401	0.26988
CoVQ27	chr11	CoVQ10	chr4	0.31502	1.72074	0.18307
CoVQ23	chr10	CoVQ13	chr5	1.02712	1.92371	0.53392
CoVQ16	chr7	CoVQ22	chr9	0.22299	0.87067	0.25611
CoVQ33	chr13	CoVQ22	chr9	0.53652	1.84173	0.29131
CoVQ27	chr11	CoVQ28	chr11	0.00000	0.00000	NaN
CoVQ33	chr13	CoVQ31	chr12	0.33427	0.63521	0.52623
CoVQ35	chr13	CoVQ36	chr13	0.00262	0.02774	0.09458
CoVQ3	chr2	CoVQ4	chr2	0.00387	0.02977	0.12988
CoVQ11	chr4	CoVQ12	chr4	0.00925	0.03171	0.29168

Note: ‘NaN’ entries indicate that the calculation of Ka/Ks ratios was not possible due to insufficient sequence divergence (i.e., Ks = 0), leading to division by zero

Cis-acting element analysis of CoVQ gene promoters​

Analysis of 1,500-bp promoter regions upstream of the CDS revealed diverse cis-regulatory elements (Fig. 5, Table S3). A total of 61.5% (24/39) of CoVQ promoters contained W-box elements ((T)(T)TGAC(C/T)), with CoVQ6 and CoVQ13 each harboring three, suggesting potential regulation by WRKY transcription factors. Notably, 94.9% (37/39) of genes contained stress-responsive elements, including ARE (antioxidant response, 32 genes), MBS (MYB-binding site, 28 genes), and DRE core (drought response, 24 genes). Strikingly, CoVQ27 and CoVQ28 each contained nine stress-related elements (e.g., LTR for low-temperature response and WUN-motif for wounding), far exceeding the family average (4.3 per gene).

Fig. 5.

Functional classification of cis-acting elements in CoVQ gene promoters. The presence and abundance of regulatory motifs related to abiotic/biotic stress, phytohormones, plant growth, and other functions are shown. The bar chart on the right indicates the total number of elements per gene

Hormone-responsive elements were present in 92.3% (36/39) of promoters. Ethylene-responsive elements (ERE) were most abundant (48 elements in 26 genes), followed by MeJA-responsive motifs (CGTCA/TGACG, 48 elements in 25 genes). CoVQ38 exhibited the highest regulatory complexity, with 10 hormone-responsive elements. Developmental and growth-related elements were dominated by light-responsive motifs: Box 4 (70 elements), GT1-motif (24 elements), and G-box (18 elements). Broadly expressed genes (CoVQ3/4/34/39) were highly enriched in these elements (mean: 8.2 per gene vs. family average: 4.7), suggesting light signaling as a key regulator of their pan-tissue expression.

Expression patterns of CoVQs in different tissues​

CoVQ genes exhibited distinct spatial expression patterns across seven C. oleifera tissues, classified as tissue-specific or ubiquitous (Fig. 6). Tissue-specific genes showed strong organ preference:

Fig. 6.

Tissue-specific expression profiles of CoVQ genes in C. oleifera. (A) Heatmap showing the expression levels of all CoVQ genes across seven tissues based on RNA-seq data (log₂FPKM). Hierarchical clustering was performed using Euclidean distance and average linkage. Red and blue represent high and low expression, respectively. (B) Validation of RNA-seq results by RT-qPCR for four representative genes (CoVQ1, CoVQ14, CoVQ16, CoVQ39). Expression values were normalized to reference genes and presented as relative fold changes. The correlation between RNA-seq and RT-qPCR data is indicated by R² and P-values (P < 0.05), showing strong consistency. Error bars represent standard deviation from three biological replicates

• Sepals: CoVQ1/6/7/13/29 were upregulated 5- to 120-fold;

• Stamens: CoVQ2/16/21/22/31 showed exclusive high expression, with CoVQ16 exhibiting 90-fold higher FPKM than other organs;

• Pistils: CoVQ27/28/37 were 3- to 7-fold elevated;

• Leaves: CoVQ14/15 showed 8- to 170-fold higher FPKM, suggesting roles in leaf development.

Ubiquitously expressed genes (CoVQ3/4/34/39) were stably detected in 5~-7 tissues and enriched in growth-related cis-elements (mean: 8.2 vs. 4.7), indicating roles in general developmental processes. RT-qPCR validation confirmed high consistency between RNA-seq and experimental data for CoVQ1/14/16/39 (Pearson’s R > 0.8, P < 0.05), supporting the reliability of the transcriptome analysis.

Expression dynamics of CoVQs under abiotic stresses​

The CoVQ gene family exhibited distinct temporal expression patterns in response to multiple abiotic stresses, revealing both stress-specific and cross-responsive regulatory roles.

In response to drought stress, five genes (CoVQ1/6/13/23/38) were significantly upregulated (> 10-fold). CoVQ1 showed a striking 278.2-fold induction by day 12, while CoVQ6 and CoVQ13 increased 26- to 58-fold, suggesting involvement in long-term drought adaptation. CoVQ23 and CoVQ27/28 peaked at days 12 and 8, respectively, indicating stage-specific regulation during drought progression.

Cold stress elicited more moderate transcriptional changes (maximum 4-fold), but CoVQ1 and CoVQ11 were consistently upregulated (2.17–2.38-fold) starting at the second timepoint and remained elevated throughout the treatment, highlighting their roles in early and sustained cold responsiveness.

By contrast, potassium deficiency led to widespread downregulation, with most CoVQs (e.g., CoVQ31/33) suppressed to 12~-21% of baseline levels by PS24. However, CoVQ18 stood out with a 7.96-fold induction during early deficiency and sustained expression, suggesting it may function as a key sensor or regulator in potassium signaling.

Under phosphorus limitation, dynamic expression patterns were observed: CoVQ29 exhibited a transient spike (9.26-fold on day 1), whereas CoVQ11 showed biphasic upregulation (2.75-fold on day 3, followed by a 1.88-fold recovery), reflecting complex temporal regulation in phosphate stress adaptation.

Notably, CoVQ27 and CoVQ28 displayed coordinated induction under drought, low phosphorus, and cold stress, suggesting a potential role as integrators of multiple abiotic stress signals (Fig. 7). RT-qPCR validation confirmed the reliability of RNA-seq data for representative stress-responsive genes, including drought-induced (CoVQ1/6), cold-responsive (CoVQ11/18), and nutrient-stress-associated (CoVQ18/31/29) members (Pearson’s R > 0.8, P < 0.05; Figure S3).

Fig. 7.

Expression profiles of CoVQ genes under abiotic stress. (A) Drought stress (Dr): samples collected at 4, 8, and 12 days post-treatment. (B) Cold stress (Co): samples collected at 3, 6, 16, and 32 h post-treatment. (C) Potassium deficiency (PS): samples collected at 12 and 24 days post-treatment. (D) Phosphorus deficiency (PHS): samples collected at 1, 3, 7, and 30 days post-treatment

Discussion

Gene family size and duplication dynamics of CoVQ genes

In this study, 39 CoVQ genes were identified in the diploid C. oleifera reference genome using a HMMER-based domain validation approach. Despite the large genome size of C. oleifera and two documented whole-genome duplication (WGD) events (the ancient γ event (~ 150 million years ago) and the more recent PPC-WGD (Polemonioids-Primuloids-Core Ericales WGD) event (4.46 ~ 5.99 Mya) that preceded the divergence of Camellia and Actinidia [25], the number of VQ genes (39) is only slightly higher than that in Arabidopsis (34) and lower than in rice (40) [16, 29].This contrasts sharply with the nonlinear scaling observed in soybean (74 VQ genes / 1,011.6 Mb) and maize (61 VQ genes / 2,192.4 Mb) [9, 26], suggesting that gene family expansion in C. oleifera has been constrained by lineage-specific evolutionary forces.

This constraint could be partly explained by the unique genomic architecture of C. oleifera: following the PPC-WGD, highly active long terminal repeat (LTR) retrotransposons-accounting for 49.76% of the genome-mediated extensive chromosomal breakage and recombination, resulting in the loss of approximately 75% of duplicated gene copies and significantly limiting gene family expansion [25]. The genomic instability driven by LTR retrotransposon bursts in woody perennials may thus represent a key mechanism for balancing gene redundancy and functional innovation under long-term selective pressures. For example, although Populus trichocarpa exhibits a higher retention rate of VQ gene duplicates (17 segmental duplication pairs among 51 genes, ~ 33.3%) compared to C. oleifera (8 segmental pairs among 39 genes, ~ 20.5%), both species predominantly rely on segmental duplications, indicating a conserved tendency among woody plants to maintain core regulatory genes through conservative duplication patterns under prolonged life cycle stresses [30].

Collinearity analysis identified 12 duplicated gene pairs in the CoVQ family-comprising 8 segmental and 4 tandem duplications-all of which evolved under strong purifying selection (Ka/Ks < 1, mean = 0.27). This “high duplication–high loss” dynamic appears closely linked to LTR-mediated chromosomal rearrangements: transposon activity can induce homologous recombination and chromosomal breakage, accelerating the elimination of redundant duplicates and ultimately shaping a moderately sized, evolutionarily stable gene family [31, 32]. In contrast, herbaceous species such as soybean retained ~ 75% of paralogous genes following two recent WGD events (14 and 42 Mya), reflecting their strategy of leveraging genetic redundancy to enhance short-term adaptability to environmental fluctuations [33].

Structural features of CoVQ genes

VQ genes in plants are typically intronless; however, in C. oleifera, 15.4% (6/39) of CoVQ genes retain introns, a proportion significantly higher than that observed in Arabidopsis (5.9%), rice (7.5%), and grape (11.1%) [16, 29, 30]. This elevated level of intron retention suggests that the CoVQ gene family has undergone lineage-specific evolutionary events in C. oleifera. Phylogenetic analysis revealed that intron-containing CoVQ37 and CoVQ39 cluster closely with homologous sequences across species, whereas CoVQ13 and CoVQ32 are dispersed across divergent clades, implying subfamily-specific intron retention patterns [34].

Cross-species comparisons further highlight distinct evolutionary trajectories: for example, in Populus trichocarpa, 14 intron-bearing VQ genes are distributed across six subgroups, with PtVQ2/3/15/22/40 forming a distinct cluster suggestive of intron loss [30]; in contrast, seven intron-containing VQ genes in maize (Zea mays), such as ZmVQ1/9/20, exhibit non-conserved distributions relative to monocot-dicot orthologs [26], indicating convergent intron acquisition. These findings collectively support a dual model of intron evolution [35]: closely related lineages (e.g., within the Salicaceae) tend to conserve ancestral intron structures and maintain phylogenetic clustering [30], while distantly related taxa achieve functional innovation through de novo intron gain [36]. Such divergence may arise from lineage-specific selective pressures-recently duplicated genes are more prone to intron loss due to genomic rearrangements (e.g., LTR retrotransposon-mediated chromosomal reorganization), whereas long-term evolution between major plant lineages shapes functional diversity via dynamic intron gain [37]. We further speculate that the unique biology of C. oleifera as a woody perennial—with its long life cycle and recurring exposure to environmental stresses—may have favored the retention of certain introns to facilitate complex regulation, thereby promoting functional diversification while preserving core regulatory functions. This represents an intriguing hypothesis for future comparative genomics studies in woody perennials [38].

Evolutionary trajectory of VQ proteins

Phylogenetic analysis of VQ proteins from C. oleifera (eudicot), Arabidopsis thaliana (eudicot), and Oryza sativa (monocot) classified them into eight subgroups, among which Groups III and V contain only eudicot members, while the remaining six subgroups (75%) include representatives from both monocots and dicots [30]. A cross-lineage phylogeny including rice, grape, Arabidopsis, and poplar further confirms that seven of these subgroups are widely conserved across monocots and eudicots, suggesting that the VQ domain was already highly conserved prior to the monocot-eudicot divergence [39]. The structural features of Groups III and V likely originated from subsequent eudicot-specific radiation events [40].

Notably, recent pan-genomic surveys have identified 88 and 7 VQ homologs in the moss Physcomitrella patens and the lycophyte Selaginella moellendorffii, respectively, and VQ-like sequences are also present in fungi and basal metazoans, supporting the hypothesis that the VQ protein family originated early in eukaryotic evolution [13, 41]. Within each subgroup, VQ proteins from C. oleifera, rice, and Arabidopsis form species-specific clusters, reflecting structural and functional divergence during long-term evolution [16, 29]. Intriguingly, all CoVQ proteins retain the canonical FxxhVQxhTG motif without substitutions at the valine (V→H/Q) or threonine positions-variants observed in Arabidopsis (e.g., LTS in AtVQ5) [16]-indicating stronger evolutionary conservation in the C. oleifera VQ family compared to those in Arabidopsis and rice.

Gene duplication-via segmental or tandem events-is a primary driver of gene family expansion [42, 43]. In the CoVQ family, 12 segmental and 4 tandem duplication pairs were identified, with segmental duplications accounting for 66.7% of expansion events, consistent with findings across diverse plant lineages [4]. All duplicated pairs exhibit Ka/Ks ratios < 1 (mean = 0.27), indicating strong purifying selection and suggesting that duplicated copies have been retained due to functional constraints, possibly involving subtle regulatory or expression divergence [44]. Thus, the CoVQ gene family not only demonstrates broad evolutionary conservation but also exhibits diversified adaptations in specific functions, providing valuable resources for studying plant evolution.

Tissue-Specific expression of CoVQ genes

VQ proteins are widely involved in the regulation of plant growth and development. For instance, the AtVQ8 mutant displays pale-green leaves and growth retardation, highlighting its critical role in these processes [45]. In this study, 94.9% (37/39) of CoVQ gene promoters contain at least one growth-related cis-regulatory element, and broadly expressed genes (e.g., CoVQ3/4/20/39) are enriched in such elements-consistent with findings in cucumber (CsVQ16-2 and CsVQ24 expressed in multiple organs) [28], soybean (24 VQ genes expressed across nine tissues) [9], and moso bamboo (14 VQ genes preferentially expressed in leaves, inflorescences, and roots) [8].

Tissue-specific expression profiles reveal distinct spatial patterns: sepal-enriched (CoVQ1/6/7/13/29/33), stamen-enriched (CoVQ2/16/21/22/31), leaf-enriched (CoVQ11/12/14/35/36), and seed-enriched (CoVQ30). Notably, collinear gene pairs-such as CoVQ6/13 (sepals), CoVQ16/22 (stamens), and CoVQ35/36 (leaves)-may act cooperatively in organ-specific development. Phylogenetic analysis further indicates that CoVQ11/12/35/36 are orthologs of AtVQ10, a key regulator of leaf development in Arabidopsis [46], and are highly expressed in oiltea leaves. Similarly, CoVQ16/21/22/31 are orthologs of AtVQ20, a stamen-specific regulator in Arabidopsis [47], and are enriched in stamens. These results suggest that CoVQ genes play conserved roles in regulating organ development in both monocots and eudicots [46].

Key roles of CoVQ27/28 and other genes in abiotic stress responses

CoVQ27 and CoVQ28 exhibited distinct temporal expression dynamics in response to drought, low-temperature, and low-phosphorus stresses. Under drought conditions, their transcript levels increased by 5.38- and 5.41-fold at days 4 and 8, respectively, and remained approximately 2-fold higher than baseline in later stages. These expression patterns were strongly correlated with the progression of stress (|r| > 0.85), indicating a coordinated regulatory role. The promoter regions of both genes contain nine stress-responsive cis-regulatory elements, including ARE (antioxidant response), TCCC-motif (involved in low-temperature response), and MBS (MYB-binding site associated with ABA signaling), suggesting their involvement in integrating multiple hormone-mediated stress signaling pathways. Notably, CoVQ27/28 are orthologous to AtVQ4, which modulates drought-responsive gene expression through interaction with WRKY transcription factors [13]. The conserved and highly similar expression profiles of CoVQ27 and CoVQ28 across tissues and stress conditions imply that their tandem duplication may confer regulatory robustness through functional redundancy or synergistic action.

In addition, CoVQ1 emerged as another key stress-responsive gene, showing a dramatic 278.2-fold upregulation at the late stage of drought and a 2.17-fold induction (13.70 TPM) during early cold exposure, reflecting rapid stress perception and sustained activation. Although its promoter harbors only two canonical stress-related elements (ARE and LTR), the presence of three ABREs (ABA-responsive elements) points to a potential role in ABA-dependent signaling pathways. Phylogenetic analysis indicates that CoVQ1 is orthologous to rice OsVQ7 (LOC_Os03g21650), a known positive regulator of drought tolerance via stomatal regulation [48]. Furthermore, the intronless structure of CoVQ1-commonly observed in rapidly inducible stress-responsive genes such as AtVQ9 [49]-may facilitate swift transcriptional activation under stress conditions.

Collectively, the CoVQ gene family potentially forms a multi-layered regulatory network underlying abiotic stress adaptation in C. oleifera, leveraging mechanisms such as gene duplication, cis-element diversification, and evolutionary conservation of function. Future functional characterization of CoVQ27/28 and CoVQ1 through transgenic approaches will help elucidate their molecular roles and provide valuable genetic targets for improving stress resilience in oil tea breeding programs.

Conclusions

This study presents a genome-wide identification of 39 CoVQ genes in C. oleifera, elucidating their evolutionary and functional dynamics in a woody perennial. The family expanded primarily through segmental duplications and evolved under strong purifying selection (Ka/Ks < 1, mean = 0.27), with widespread intron loss (84.6%) potentially linked to LTR retrotransposon-driven genomic reorganization. Expression analyses revealed tissue-specific patterns and key stress-responsive members: CoVQ1 was strongly induced by drought, CoVQ18 responded rapidly to potassium deficiency, and CoVQ27/28-enriched in stress-related cis-elements—exhibited coordinated induction across multiple abiotic stresses, implicating them as candidate multi-stress response regulators. Together, these results provide a preliminary integrated ‘evolution–structure–expression’ framework for VQ genes in C. oleifera, based on genomic, phylogenetic, and transcriptomic analyses.

It should be noted that this study is primarily based on bioinformatic predictions and expression correlation analyses. While the proposed roles of CoVQ genes in organ development and stress responses—as well as their potential interactions with WRKY transcription factors—are strongly supported by evolutionary conservation, cis-element enrichment, and co-expression patterns, functional validation through genetic approaches (e.g., CRISPR/Cas9-mediated knockout or overexpression in C. oleifera or heterologous systems) will be essential to confirm these hypotheses.

Authors’ contributions

Data curation, Z.F. and Z.L.; formal analysis, T.Z.; funding acquisition, H.L.; investigation, J.P.; methodology, Z.F.; project administration, H.L.; resources, Z.L.; software, Z.F., Z.L., J.P., T.Z. and R.H. ; supervision, B.Y.; validation, Z.F., Z.L., J.P., T.Z. and R.H.; visualization, Z.F., Z.L., J.P., T.Z. and R.H.; writing-original draft, Z.F., Z.L., J.P., T.Z., R.H., B.Y. and H.L.; writing-review and editing, Z.F. ,B.Y. and H.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (31360491).

Data availability

Data supporting the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

Not applicable. All methods were carried out according to relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.