Comprehensive analysis of the LiCPKs in Lagerstroemia indica for insights into evolution and regulatory roles in flower maturation and bud-to-branch development

Abstract

Calcium ions (Ca²⁺) serve as crucial second messengers in a wide range of developmental and stress-related processes in plants. Calcium-dependent protein kinases (CPKs) play essential roles in Ca²⁺ signal transduction, regulating essential stress responses. In this study, we identified 38 LiCPKs from the Lagerstroemia indica L. genome. Phylogenetic tree classified these CPKs into five distinct groups-CPKI, CPKII, CPKIII, CPKIV, and CRK, indicating a closer genetic relationship between crape myrtle and Arabidopsis, compared to rice. Gene duplication analysis revealed that segmental duplication is the primary mechanism driving the expansion of LiCPKs. Weighted gene co-expression network analysis (WGCNA) highlighted the involvement of LiCPKs in developmental processes, including flower pigmentation, particularly during red flower maturation. Additionally, WGCNA revealed that several LiCPKs are involved in regulating developmental transitions from bud initiation to branch formation. These findings offer meaningful insights into the evolutionary history, structural diversity, and functional significance of LiCPKs, contributing to a deeper understanding of their roles in plant development and adaptation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08647-y.

Introduction

Plants that thrive in fluctuating environments contend with a range of challenges, encompassing both abiotic and biotic stresses [1]. To cope with these ever-changing conditions plants have evolved an array of adaptive strategies such as the accumulation of osmo-protectants, activation of antioxidant systems and modulation of gene expression involved in stress tolerance [2–5]. Within the complex signaling networks that plants utilize to environmental stresses, calcium ions (Ca²⁺) emerge as key cellular messengers, along with cyclic AMP and inositol, with crucial roles in the detection and response to environmental stresses [6]. The dynamics of cytoplasmic Ca²⁺ levels characterized by their intensity, duration, and frequency are critical in determining the strength and persistence of Ca²⁺ signals. To interpret these Ca²⁺ signals, plants rely on an array of calcium-sensing proteins, including calmodulins (CaMs), calmodulin-like proteins (CMLs), calcineurin B-like proteins (CBLs), and the particularly noteworthy class of Ca²⁺-dependent protein kinases (CPKs) [7, 8].

So far, the CPK gene family has been studied in several model plant species. For instance, in Arabidopsis, 34 CPK genes are discovered [9], rice had 31 [10], maize contained 40 [11], poplar had 30 [12], grapevine featured 19 [13], tomato possessed 29 [14], Gossypium barbadense encompassed 84 [15], and bamboo included 30 [16]. CPKs are distinguished by their ability to directly translate upstream Ca²⁺ signals into protein phosphorylation events, a key regulatory mechanism in the plant’s stress response [17]. CPKs comprise of four key domains: the N-terminal region, containing sites for myristylation or palmitoylation, the autoinhibitory junction domain that suppresses kinase activity, and the CML domain (CaMLD) at the C-terminal end, which contains EF-hand motifs for Ca²⁺ binding. These domains allow CPKs to relay Ca²⁺ signals within the cytoplasm and initiate responses to environmental shifts [18]. Fluctuations in intracellular Ca²⁺ levels prompt swift reactions from CPKs, with the ability to regulate a wide range of substrates through phosphorylation. This process is crucial in orchestrating plant physiological responses, guiding growth and development and mediating adaptation to various environmental stresses.

CPKs target a variety of substrates, and this includes transcriptional regulators, ion channels, transporters, and proteins involved in signal transduction pathways [19]. In Arabidopsis, AtCPK12 elicits phosphorylation of ABA-responsive element-binding factors ABF1 and ABF4, as well as protein phosphatase 2 C (PP2C) under Ca²⁺ signaling conditions [20]. Other CPKs, such as AtCPK2 and AtCPK20, influence pollen tube elongation by activating the anion channel SLAH3 [21]. In addition, AtCPK6 and AtCPK33 have been implicated in the floral signaling of Arabidopsis through phosphorylation of FLOWERING LOCUS T (FT) [22]. AtCPK3 and AtCPK4 have been involved in Arabidopsis root development via phosphorylation of PLAIVA and PLAIVB, key components of the auxin signaling pathway [23]. AtCPK8 is also recognized for its interaction with and phosphorylation of the tetrameric iron porphyrin protein CATALASE3 (CAT3) in response to drought stress [24]. In rice, OsCPK24 enhances cold resistance by elevating proline and glutathione levels and by inhibiting OsGrx10 through phosphorylation, thereby sustaining higher glutathione levels [25].

While significant progress has been made in characterizing the functions of CPKs, most research has focused on model plants. In contrast, the roles of CPKs in L. indica L. (crape myrtle), a member of the Lythraceae family widely used in urban landscaping for its elegant architecture and diverse floral colors and patterns [26], remain poorly understood. Further investigation is needed to clarify the biological functions of CPKs in this ornamental species. In L. indica, both flower coloration and branching patterns are crucial determinants of ornamental quality. The pigment composition and spatial distribution within petals define flower color, which is regulated by genetic and environmental factors. Similarly, the development of complex branching patterns arises from axillary buds and is controlled by a coordinated network of genetic, hormonal, and environmental cues [27]. The transformation of a dormant or axillary bud into an actively growing branch represents a fundamental morphogenetic process shaping plant architecture. Understanding the molecular mechanisms underlying this transition is fundamental to revealing how plants integrate developmental and environmental signals to shape their growth. However, the role of LiCPKs in regulating bud activation and branch formation in L. indica remains largely unexplored. This study therefore aims to conduct a comprehensive investigation of CPK genes in L. indica (LiCPKs). The objectives are to characterize their evolutionary relationships, gene structures, conserved motifs, regulatory elements, potential protein interactions, and functional annotations. In addition, the expression patterns of LiCPKs related to branching development, flower color, and red flower maturation are also analyzed. Together, this analysis provides new insights into the evolutionary and functional characteristics of LiCPKs and lays a foundation for future studies on their potential roles in controlling branching and floral development in L. indica.

Materials and methods

Genome-wide identification of CPKs in L. indica

The complete genome sequences for L. indica were retrieved from the China National GeneBank DataBase (CNGBdb) with the accession number CNP0003018. Additionally, the CPK protein sequences from Arabidopsis thaliana and Oryza sativa were sourced from the Phytozome database. A BLASTP search was performed on the L. indica genome, utilizing the protein sequences from A. thaliana and O. sativa CPKs as reference points (Supplemental Table 1). The search criteria included a significant e-value of ≤ 1.0 × 10^− 10 and a minimum identity of over 40%. The domains of CPK-specific Hidden Markov Models (HMMs), namely Pkinase (PF00069) and EF-hand_7 (PF13499), were obtained from the Pfam database [7, 28]. These HMMs were then used to query and identify the corresponding CPKs within the L. indica genome, applying an e-value threshold of 1.0 × 10^− 5. Additionally, the identified LiCPKs were renamed based on their chromosomal locations, and the fundamental physicochemical properties of these LiCPKs were determined using the ExPASy online tool. Moreover, the Cell-PLoc 2.0 online tool was employed to predict the subcellular localization of these proteins.

Phylogenetic tree, conserved motif, and gene structure analysis of LiCPKs

The ClustalW software was utilized for aligning the full-length CPKs from Arabidopsis, rice, poplar, and crape myrtle (Supplemental Table 1). Subsequently, the construction of the phylogenetic tree was carried out using MEGA7 software, employing the neighbor-joining (NJ) method with 1, 000 bootstrap replications to ensure robustness of the tree. The tree was then visualized and refined using the interactive tree of life (iTOL) platform. In addition, the MEME program was employed to identify conserved motifs within each LiCPK. The search was conducted with a maximum of 10 motifs, using the default settings for all other parameters. The LiCPK structures were elucidated using the general feature format version 3 (gff3) annotation file specific to L. indica. The visualization of both the phylogenetic tree and gene structures was accomplished with the aid of TBtools software [29].

Syntenic gene pairs and Ka/Ks ratios

The chromosomal positions of LiCPKs were identified by consulting the gff3 annotation file and visualized using TBtools. To assess the gene collinearity within the LiCPKs, the Multiple Collinearity Scan toolkit (MCScanX) was applied, and the outcomes were graphically represented using the Circos tool within the TBtools. Additionally, in an effort to delve deeper into the evolutionary relationships of orthologous CPKs across different species, including L. indica, A. thaliana, O. sativa, Eucalyptus grandis, Vitis vinifera, Populus trichocarpa, and Salix purpurea, the MCScanX program was employed once more. The results of this analysis were visualized using the Dual Synteny Plot feature in TBtools software, offering a comprehensive view of the syntenic relationships among these species. Moreover, the coding sequences (CDSs) of LiCPKs were extracted, and the non-synonymous substitution rate (Ka), synonymous substitution rate (Ks), and Ka/Ks ratio were calculated using TBtools.

Analysis of cis-acting regulatory elements, protein interaction networks, and functional annotation

The 2, 000-base pair (bp) sequences upstream from the translation start site of all LiCPKs were extracted for the identification of potential cis-acting regulatory elements using the PlantCARE database. The findings from this analysis were then visualized using TBtools software. In addition, the putative protein interaction networks on AtCPKs were achieved from the STRING database. Utilizing homologous sequence alignment, the interaction networks for LiCPKs were constructed with the aid of Cytoscape software. Additionally, for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotation assessments, the accession numbers of LiCPKs were submitted to TBtools software. TBtools was also employed to carry out enrichment evaluations for both GO and KEGG, providing insights into the biological processes (BP), cellular components (CC), molecular functions (MF), and metabolic pathways associated with LiCPKs.

Transcriptome data analysis of LiCPKs response to salt treatment

Stem cuttings of crape myrtle, both salt-sensitive and salt-tolerant varieties, measuring 8 to 10 cm in length and 4 to 5 mm in thickness, were cultured in soil mixed with sterilized peat and perlite (2:1) under controlled greenhouse conditions for 3 to 4 weeks. These conditions included a 16-h light cycle, a temperature of 25 °C, and a humidity level of 60%-70%. Subsequently, the crape myrtle plants were subjected to a 200 mmol/L sodium chloride (NaCl) treatment. After a 4-h treatment period, samples were collected for RNA sequencing (RNA-seq). Three biological replicates were used for each group, with each replicate consisting of pooled stem cuttings from multiple plants to ensure representation. Additionally, three technical replicates were performed for each biological sample to ensure consistency in the RNA sequencing process. In the salt-tolerant group, the control was labeled N-CK, and the treatment was labeled N-T. For the salt-sensitive group, the control was designated M-CK, and the treatment was designated M-T. The quality of RNA samples was carefully monitored and assessed. Sequencing libraries were then prepared for the Illumina platform, with precise quantitative quality control ensuring effective concentrations greater than 2 nanomolar (nM) using the quantitative polymerase chain reaction (qPCR) method. Ultimately, the libraries were sequenced on the Illumina platform system by Baimake Cloud Technology Co., Ltd, in China. The RNA-seq data used in this study were originally deposited into the CNGBdb with the accession number CNP0003991, as reported in previous studies [30]. These data were reanalyzed specifically to investigate the expression patterns and functions of LiCPKs in response to salt stress.

Weighted gene co-expression network analysis (WGCNA) of LiCPK gene expression specific to flower colors, red flower maturation, and bud-to-branch development

The expression analysis of LiCPKs was conducted across four distinct flower color categories: red, purple, pink, and white, designated as BB, PB, Pu, and WB, respectively. The foundational raw RNA-seq data for this study was housed in the CNGBdb under the accession number CNP0001693. Additionally, RNA-seq data corresponding to the early developmental stage of red flowers (R-bud) and the mature stage (BB) were obtained from the CNGBdb under the accession number CNP0001693. To explore the potential mechanism behind the flower maturation, raw RNA-seq data and qRT-PCR (with primers listed in Supplemental Table 2) were also analyzed for further investigation of this mechanism. LiActin (evm.model.Chr10.488), considered as the internal control, was used to calculate the relative expression levels of the gene based on the 2^−ΔΔCT method [31, 32]. Moreover, to explore the molecular mechanism underlying bud-to-branch development, RNA-seq was performed on samples collected from buds (I), upper tissues at the bending sites (UTBS, A), and lower tissues at the bending sites (LTBS, B), with data available under accession number CNP0003991. Transcript levels were estimated using the method of fragments per kilobase of transcript per million mapped fragments (FPKM). A comparative differential expression analysis was performed between these groups, each with triplicate biological samples, employing the DESeq2 tool. The p-values were adjusted to calculate the false discovery rate (FDR) for statistical significance. Genes that showed an adjusted fold change of at least 1.5 and an FDR threshold below 0.05 were identified as differentially expressed genes (DEGs). Moreover, the co-expression networks among the LiCPKs were investigated using the Pearson correlation coefficient to measure the strength of these interactions. The genes co-expressed with LiCPKs underwent comprehensive GO and KEGG enrichment analysis to elucidate their biological roles and the pathways in which they participate.

Results

Genome-wide identification of LiCPKs in L. indica

Based on BLASTP and Pfam analysis, a total of 38 CPKs were identified within the L. indica genome (Supplemental Table 3). These LiCPKs were sequentially named LiCPK1-LiCPK35 and LiCRK1-LiCRK3 according to their chromosomal order. The 24 chromosomes of L. indica genome were designated as Chr1-Chr24, and 38 LiCPKs were found to be unevenly distributed across 21 of these chromosomes, indicating potential functional divergence. Specifically, Chr19 had the highest number of LiCPKs, with four, while Chr2, Chr13, and Chr16 each contained three. Chr3, Chr6, Chr10, Chr17, Chr20, Chr21, Chr22, and Chr24 each had two LiCPKs. The fewest LiCPKs were found on Chr1, Chr4, Chr5, Chr7, Chr8, Chr9, Chr15, Chr18, and Chr23, with only one each, which might indicate these regions have a more specialized function (Supplemental Fig. 1). The longest LiCPKs, encoded by LiCRK1 and LiCRK2, consisted of 593 amino acids (aa), whereas the shortest, LiCRK3, encoded only 414 aa. The CDS lengths of LiCPKs varied from 1,245 to 1,782 bp. The predicted molecular weights (MWs) of LiCPKs ranged from 47.54 kDa for LiCRK3 to 66.71 kDa for LiCRK2, with isoelectric points (pIs) spanning from 5.20 for LiCPK27 to 9.22 for LiCPK35, suggesting variability in their charge properties, which could influence their interactions with other cellular components (Supplemental Table 3). The instability index of LiCPKs ranged from 34.08 to 53.28, indicating that 13 of the proteins are stable, while 25 are unstable, which suggests that the more unstable proteins may be involved in highly dynamic processes like signal transduction under stress conditions. The grand average of hydropathicity for all LiCPKs was negative suggesting that they are hydrophilic proteins. Subcellular localization predictions indicate that all LiCPKs are localized to the nucleus (Supplemental Table 3) implying a significant role for LiCPKs in nuclear processes, potentially including gene expression regulation.

Identification and phylogenetic relationship of CPKs among Arabidopsis, rice, poplar, and crape myrtle

The phylogenetic analysis of CPKs across four plant species, including A. thaliana, O. sativa, P. trichocarpa, and L. indica, revealed distinct evolutionary relationships within the CPK family. A total of 42 CPKs from Arabidopsis, 34 from rice, 30 from poplar, and 38 from crape myrtle were identified and classified. The CPK family was divided into five major subgroups: CPKI, CPKII, CPKIII, CPKIV, and CRK, based on sequence homology and phylogenetic clustering (Fig. 1). The CPKI and CPKIII groups exhibited significant variation in the number of CPKs across the species. L. indica displayed a similar number of CPKs in these groups when compared to P. trichocarpa, A. thaliana, and O. sativa. Specifically, the CPKI group contained 10, 11, 11, and 13 CPKs in A. thaliana, O. sativa, P. trichocarpa, and L. indica, respectively. Similarly, the CPKIII group included 8, 8, 9, and 13 CPKs in A. thaliana, O. sativa, P. trichocarpa, and L. indica, respectively. The CPKII group, in contrast, had 13, 8, 8, and 7 CPKs in A. thaliana, O. sativa, P. trichocarpa, and L. indica, respectively. Additionally, each subgroup displayed significant species-specific expansions, with particular emphasis on the CPKIII group, which showed notable diversification in woody species such as P. trichocarpa and L. indica. In contrast, O. sativa and L. indica exhibited a more conserved distribution of CPKs across the subgroups. The clustering of L. indica CPKs with those of O. sativa and P. trichocarpa further supports the notion of shared evolutionary paths among dicot and monocot species, while also highlighting the unique evolutionary trajectory of CPKs in L. indica.

Fig. 1.

Phylogenetic analysis of CPKs across Arabidopsis thaliana (At), Oryza sativa (Os), Populus trichocarpa (Pt), and Lagerstroemia indica (Li). A phylogenetic tree illustrating the evolutionary relationships among these CPKs. The tree was constructed using the neighbor-joining (NJ) method as implemented in MEGA7 software and visualized with iTOL for enhanced clarity. CPKs were grouped into five distinct categories, color-coded for visual distinction: CPKI, CPKII, CPKIII, CPKIV, and CRK

Gene structure and conserved motif analysis

To gain a deeper understanding of the potential structural evolution and functional diversification of LiCPKs, a comparative analysis of the exon-intron organization was performed. (Supplemental Fig. 2). Almost all LiCPKs grouped into CPK I contained seven exons and six introns. The LiCPKs in CPK II displayed a variety of exon numbers: LiCPK14, LiCPK17, LiCPK18, and LiCPK34 each had 8 exons and 7 introns, while LiCPK19, LiCPK22, and LiCPK26 each had 9 exons and 8 introns. In CPK III group, the majority of LiCPKs featured 8 exons and 7 introns, with the exceptions of LiCPK6, LiCPK7, LiCPK11, and LiCPK33, which had 7 exons and 6 introns. The three LiCPKs clustered into CPKIV both contained 12 exons and 11 introns, while LiCRK1 and LiCRK2 each had 11 exons and 10 introns, and LiCRK3 had 10 exons and 9 introns. This variation suggests potential functional specialization, particularly in the CRK subfamily.

The exploration into conserved motifs conducted via MEME analysis unveiled 10 distinct motifs across LiCPKs, which helped an understanding their functional blueprint (Supplemental Fig. 3). Motifs 1–4 were closely associated with the protein kinase domain, while motifs 5, 6, and 8 were linked to the EF-hand domain. Additionally, motif 7 was identified as containing the ATP binding site (Supplemental Table 4). LiCPKs clustered into CPKI, CPKII, and CPKIII harbored motifs 1–10, whereas LiCPKs grouped into CPKIV contained motifs 1–5, 7, 8, and 10. LiCRK1, LiCRK2, and LiCRK3 all contained conserved motifs 1–5. Additionally, LiCRK1 included motif 8 in addition to motifs 1–3, LiCRK2 contained both motifs 8 and 9 alongside motifs 1–3, and LiCRK3 contained motif 9 in addition to motifs 1–3.

Syntenic analysis of LiCPKs

Gene duplication serves as a key mechanism through which organisms acquire new genes and functions. Segmental and tandem duplication events analyzed to understand the evolutionary process of LiCPKs (Fig. 2) showed that 31 gene pairs were located within segmental duplication regions, whereas no tandem duplication was detected. This indicates that segmental duplication played a significant role in the expansion of the LiCPK gene family. A higher degree of sequence similarity within gene pairs located in the same group was observed, suggesting that these duplications might have contributed to functional diversification of LiCPKs within L. indica. The Ka, Ks, and Ka/Ks ratios calculated to assess the evolutionary pressures on LiCPKs (Fig. 3 and Supplemental Table 5) showed that except for the pair LiCPK27/LiCPK2 reflecting a high sequence divergence value, the Ka/Ks ratios of collinear gene pairs were seen consistently less than one. This suggests that LiCPKs have likely been subjected to purifying selection throughout their evolution, maintaining their functional integrity while minimizing nonsynonymous mutations.

Fig. 2.

Chromosomal collinearity of LiCPKs in L. indica genome. The outermost blocks represented the individual chromosomes of L. indica. The middle section utilized a line format to depict the density of LiCPKs along the chromosomes. The inner section complemented this with a heatmap, where color intensity corresponded to gene density, further emphasizing regions of high gene concentration. The red lines in the middle section were used to visually connect genes that exhibit collinearity, indicating conserved gene orders that may result from segmental duplication events

Fig. 3.

Lollipop chart representation of Ka, Ks, and Ka/Ks ratios in LiCPK gene pairs. The Ka and Ks values for LiCPKs were calculated using the Ka_Ks calculator in TBtools, a suite of bioinformatics tools. Each “lollipop” on the chart symbolized a syntenic gene pair, with the length of the stick representing the Ka, Ks, and Ka/Ks values for that gene pair at a specific site. The Ka/Ks ratio was a critical measure for inferring selection pressures, where a Ka/Ks value significantly less than 1 suggested purifying selection, a value near 1 indicated neutral evolution, and a value greater than 1 implied positive selection

Collinear map constructed using orthologous genes as shown in Fig. 4 presented the synteny among CPKs across various species. A total of 55 orthologous CPKs were identified between L. indica and A. thaliana, while the same had been 10 between L. indica and O. sativa, 47 between L. indica and E. grandis, 69 between L. indica and P. trichocarpa, 72 between L. indica and S. purpurea, and 31 between L. indica and V. vinifera (Supplemental Table 6). When comparing the number of orthologous CPKs, the LiCPK-SpCPK and LiCPK-PtCPK pairs showed a higher count compared to the pairs LiCPK-AtCPK, LiCPK-VvCPK, and LiCPK-OsCPK. This suggests that L. indica exhibits higher sequence homology with A. thaliana compared to O. sativa. Additionally, the collinearity analysis also reveals that the synteny between L. indica and P. trichocarpa as well as S. purpurea is greater than that observed with A. thaliana.

Fig. 4.

Comparative syntenic analysis of CPKs across diverse plant species, including L. indica, A. thaliana, O. sativa, Eucalyptus grandis, Vitis vinifera, Populus trichocarpa, and Salix purpurea. The grey lines connected collinear gene pairs, indicating regions of conserved gene order between species. The red lines specifically highlighted syntenic CPKs, drawing attention to the gene pairs that were orthologous across the species examined. Chromosomes from each species were represented by distinct colored rectangles

The Ka, Ks, and Ka/Ks ratios for CPK gene pairs across multiple plant species were illustrated in Supplemental Fig. 4 and Supplemental Table 7. It could be noticed that some Ka/Ks values between L. indica and A. thaliana, for the gene pairs LiCPK32/AtCPK18, LiCPK21/AtCPK6, and LiCPK21/AtCPK5 exhibited high sequence divergence with pS values greater than 0.75, hinting substantial evolutionary differences between these species. Similarly, the Ka/Ks values for the gene pairs between L. indica and O. sativa also demonstrated notable sequence divergence, particularly for the pair LiCPK32/OsCPK7, with a pS value exceeding 0.75. In contrast, the Ka/Ks ratios for CPK gene pairs between L. indica and E. grandis, L. indica and P. trichocarpa, and L. indica and S. purpurea were mostly less than 1 indicating that these genes are under purifying selection. These findings suggest that the CPKs from these species have been highly conserved throughout their evolutionary history with limited non-synonymous mutations. The majority of the gene pairs between L. indica and E. grandis, L. indica and P. trichocarpa, and L. indica and S. purpurea did not show significant sequence divergence, this suggests that these genes have evolved more slowly and are under selective pressure to maintain their functional stability.

Cis-acting element prediction in LiCPK promoters

To explore the regulatory mechanisms underlying the expression of LiCPKs during plant development and in response to various stresses, we performed an analysis of the putative cis-acting regulatory elements within a 2, 000-bp DNA sequence from the promoter regions of LiCPKs using the PlantCARE database. The results revealed the involvement of multiple phytohormone-responsive elements, including factors that respond to abscisic acid (ABA), auxin (IAA), methyl jasmonate (MeJA), gibberellin (GA), and salicylic acid (SA). This suggests that the expression of LiCPKs is modulated by a variety of phytohormones, which are crucial for regulating growth, development, and stress responses in plants. This finding is consistent with our hypothesis that LiCPKs play a central role in integrating hormonal signals, which could influence both developmental transitions and stress adaptations in L. indica. Additionally, several light-responsive elements were found, indicating that LiCPKs may also be regulated by environmental factors such as light conditions. This can suggest a role for LiCPKs in responding to changes in the plant’s external environment. The promoters of some LiCPKs also contained stress-responsive elements such as those involved in anaerobic induction, defense and stress response, and low-temperature response, reinforcing the idea that LiCPKs are involved in the plant’s ability to adapt to various stress conditions. Moreover, various growth and development-related elements were identified encompassing circadian rhythm elements, seed-specific regulatory elements, zein metabolism regulation elements, meristem expression elements, endosperm expression elements, and elements related to the differentiation of palisade mesophyll cells. These findings support the notion that LiCPKs are integral to multiple developmental processes and may be involved in regulating key transitions such as seed development and tissue differentiation.

Protein interaction network analysis

To better understand the potential functional roles of LiCPKs in L. indica, protein interaction networks were constructed using the STRING database. The AtCPK interaction networks were identified using the STRING database (Supplemental Fig. 6), which facilitated the construction of homologous LiCPK protein-protein interaction networks. Additionally, the corresponding annotations for the interacting proteins are provided in Supplemental Table 8. Networks presented in Fig. 5. Information deduced, for instance, LiCPK3 was found to interact with nine proteins in crape myrtle, including evm.model.Chr6.761, was associated with regulating NAD(P)H oxidase H₂O₂-forming activity and Ca²⁺ binding had fetched interesting details. Another notable interaction was with evm.model.Chr3.277, an ABA-responsive element binding protein 2, which had its own implications on mediating ABA-dependent stress responses. LiCPK2 was predicted to interact with 15 functional proteins, such as evm.model.Chr4.1399, a homolog of AT1G12480, which encoded a slow anion channel-associated protein 1 and was involved in cellular ion homeostasis and S-type anion currents. Another protein, evm.model.Chr3.770, encoded a plasma membrane proton ATPase is implicated in drought response mechanisms by regulating the function of stomata. LiCPK18’s interactions included evm.model.Chr22.243, encoding a respiratory burst oxidase homologue C, involved in ROS production and stimulated Ca²⁺ influx. In addition, CRK3 was predicted to interact with 13 proteins, including evm.model.Chr14.360, which encoded a CML protein involved in nitrate-induced brassinosteroid (BR) signaling and hypoxia response, and evm.model.Chr19.106, which encoded salt overly sensitive 3, crucial for K⁺ nutrition, K⁺/Na⁺ selectivity, and salt tolerance.

Fig. 5.

Protein-protein interaction network of LiCPKs. The nodes representing LiCPKs and their putative interaction partners are depicted as circles. The size and color of the circles correspond to the strength of the interaction score, with larger circles indicating higher confidence in the protein-protein interactions

Functional annotation analysis of LiCPKs

To gain further insights into the functional roles of LiCPKs in L. indica, we performed GO enrichment analysis. The results revealed that LiCPKs were annotated with 26 significantly enriched terms,, which were classified into several MF and BP. Among these, key MF terms included adenyl nucleotide binding, anion binding, ribonucleotide binding, protein kinase activity, catalytic activity, Ca²⁺ binding, ATP binding, and ion binding. Additionally, terms related to transferase activity, nucleotide binding, phosphotransferase activity, metal ion binding, cation binding, and organic cyclic compound binding were also identified (Supplemental Fig. 7A). These results suggested that LiCPKs are involved in various enzymatic functions, including binding and catalytic activities, which are crucial for their biological roles. Additionally, KEGG pathway enrichment analysis revealed six primary pathways in which LiCPKs participate, highlighting their involvement in plant-pathogen interactions, protein kinase signaling, environmental adaptation, organismal systems, metabolism, and brite hierarchies (Supplemental Fig. 7B). The identified pathways emphasized the importance of LiCPKs in essential cellular processes such as signal transduction, stress responses, and metabolic regulation.

Expression analysis of LiCPKs under salt stress

Transcriptome analysis employed to examine the expression patterns of LiCPKs in both salt-tolerant and salt-sensitive crape myrtle varieties under salt condition led to certain interesting findings. The analysis revealed 13 LiCPKs that showed differential expression under salt stress, which were selected for further investigation (Fig. 6). Notably, LiCPK4 and LiCPK19 exhibited reduced expression in both cultivars upon salt stress. Conversely, LiCPK24 displayed the lowest expression in salt-tolerant varieties under the same conditions. LiCPK30 showed an increase in expression in salt-sensitive varieties but a decrease in salt-tolerant ones when subjected to salt stress. LiCPK16 and LiCPK28 maintained relatively high expression levels in both cultivars, with LiCPK16 showing the highest expression in salt-sensitive varieties. WGCNA analysis revealed that LiCPK16 is significantly associated with the black module. GO and KEGG enrichment analysis conducted on co-expressed genes to elucidate the BP potentially involving LiCPK16 is displayed in the Supplemental Fig. 8. A total of 113 genes within the black module were enriched for GO terms related to UDP-glycosyltransferase, catalytic, and glycosyltransferase activities, membrane component, as well as sulfur compound metabolism (Supplemental Fig. 8A). It is interesting to note that these factors were found to be involved in KEGG pathways related to environmental information processing, signal transduction, the MAPK signaling pathway, protein processing in the endoplasmic reticulum, plant hormone signal transduction, transporters, and glycosyltransferases, and find themselves work as elements of complex regulatory networks (Supplemental Fig. 8B).

Fig. 6.

Heatmap of LiCPK expression patterns under salt stress condition. The experiment was designed to assess the expression profiles of these genes in both salt-sensitive (M-CK) and salt-tolerant (N-CK) varieties under normal conditions, as well as under salt stress (M-T and N-T, respectively). The heatmap utilized a color gradient to represent relative expression levels, with purple shades indicating higher expression levels and red shades signifying lower expression levels. The graphs located to the right and above the heatmap displayed the distribution of gene expression levels across all samples, with the bar plots on the right showing the total expression levels for each gene, and the bar plots at the top illustrating the overall expression distribution for each condition (M-CK, M-T, N-CK, N-T)

Variation in the expression of LiCPKs across distinct flower color phenotypes

The potential functions of LiCPKs were explored in detail from the expression patterns of 33 differentially expressed LiCPKs across four different flower colors and gets illustrated in Fig. 7A. Therefore, these 33 LiCPKs were selected for further analysis due to their significant differential expression across the flower color categories. The genes categorized based on their expression profiles fell into four distinctive groups. Cluster I included LiCPK2, 23, 25, 28, 32, and LiCRK1, Cluster II comprised LiCPK1, 4, 7, 11–13, 15, 17, 19, 20, 24, 26, 27, 31, 33–35, and LiCRK2, Cluster III contained LiCPK9, and Cluster IV consisted of LiCPK3, 6, 10, 14, 16, and 18 (Fig. 7B). Using WGCNA, it was observed that LiCPK31 and LiCRK2 were significantly linked to the blue module, LiCPK29 was associated with the brown module, and LiCPK11, LiCPK15, LiCPK17, LiCPK19, and LiCPK20 stood grouped as turquoise module.

Fig. 7.

Weighted gene co-expression network analysis (WGCNA) of genes associated with flower color variation in crape myrtle. A The heatmap displays the differentially expressed genes (DEGs) related to LiCPKs across various flower colors, including red (BB), purple (PB), pink (Pu), and white (WB), as identified through RNA-sequencing (RNA-seq) analysis. B The clustered dendrogram illustrated the hierarchical clustering of DEGs in LiCPKs based on their expression patterns. C The Gene Ontology (GO) enrichment analysis section highlighted the biological processes (BP), molecular functions (MF), and cellular components (CC) that were overrepresented among the co-expressed genes. D The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis identified the crucial metabolic and regulatory pathways in which the co-expressed genes are implicated

Significantly, LiCPK31 and LiCRK2, associated with the blue module displayed elevated expression in the Pu and WB, in stark contrast to their reduced expression in BB and PB. This pattern suggests a potential role for these genes in the regulation of flower color, particularly in the synthesis or modification of pigments. To further explore the molecular mechanisms underlying flower color development, we conducted comprehensive GO and KEGG enrichment analysis for LiCPK31, LiCRK2, and their co-expressed counterparts. The GO enrichment analysis indicated significant enrichment for terms related to binding, transferase activity, and oxidoreductase activity (Fig. 7C). These findings suggest that LiCPK31, LiCRK2, and their co-expressed genes may be involved in enzymatic processes critical for pigment biosynthesis and other cellular functions related to flower color. Moreover, these genes were implicated in several KEGG pathways, including the biosynthesis of secondary metabolites, plant hormone signal transduction, environmental adaptation, and carbohydrate metabolism. Notably, the blue module, which included LiCPK31 and LiCRK2, showed a correlation with higher expression in certain flower colors, suggesting a potential role in color determination (Fig. 7D). These pathways are essential for regulating plant development and response to environmental cues, highlighting the multifunctional roles of LiCPKs. The GO and KEGG enrichment analysis further highlighted the involvement of these genes in critical biological processes, such as binding, enzymatic activities, and pathways related to secondary metabolite biosynthesis, plant hormone signaling, environmental adaptation, and carbohydrate metabolism. These results collectively support the hypothesis that LiCPKs play a key role in flower color regulation and provide a molecular basis for understanding determination of flower color in L. indica.

Expression profiles of LiCPKs during the maturation process of red flowers

To investigate the underlying mechanisms of red flower maturation, we conducted a detailed analysis of the expression patterns of 8 differentially expressed LiCPKs during the maturation process. Unlike the early stages, LiCPKs throughout the maturation process revealed that LiCPK4, LiCPK13, LiCPK17, and LiCPK20 have higher expression levels. In contrast, LiCPK1, LiCPK11, LiCPK15, and LiCPK19 exhibited higher expression levels during the initial process and lower levels during maturation (Fig. 8A and B). WGCNA analysis revealed that LiCPK11, LiCPK15, and LiCPK19 were grouped into the turquoise module. The GO and KEGG enrichment analysis undertaken to explore the potential mechanisms underlying the maturation process of red flowers presented a data with enhanced role for LiCPK11, LiCPK15, and LiCPK19 when compared with their co-expressed counter parts. The GO analysis revealed significant enrichment in hydrolase, catalytic, transferase, and acyltransferase activities (Fig. 8C). Similarly, the KEGG analysis showed comparable results, with genes involved in carbohydrate and lipid metabolism, biosynthesis of secondary metabolites, plant hormone signal transduction, glycosyltransferase activity, and starch and sucrose metabolism (Fig. 8D). The expression profile analysis and subsequent enrichment findings suggests that LiCPKs play pivotal roles in the maturation process of red flowers, with distinct expression patterns in the initial and mature stages. These findings provide evidence that LiCPKs are integral to flower maturation, particularly in the regulation of metabolic pathways and stress responses.

Fig. 8.

Expression profiles and functional analysis of LiCPKs during the mature process of red flowers. A Heatmap depicting the expression profiles of LiCPKs from the RNA sequencing dataset. B qRT-PCR analysis of LiCPKs during the initial and mature stages of red flower development. C Gene Ontology (GO) enrichment analysis of LiCPKs and their co-expressed genes. D Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of LiCPKs and their corresponding co-expressed genes

Expression patterns of LiCPKs during the process of bud-to-branch development

To investigate the underlying mechanisms of bud-to-branch development in L. indica, WGCNA was employed. The WGCNA results indicated that LiCPK19 and LiCPK29 can be grouped into the turquoise and blue modules, respectively. This module grouping suggests that these LiCPKs are co-expressed with genes that are involved in similar regulatory networks during bud-to-branch development. LiCPK19 exhibited reduced expression in the bud stage but increased expression in the lower tissues of the branch, indicating its potential role in the transition from bud dormancy to active branch growth. This pattern suggests that LiCPK19 may function in the early stages of branch formation and could be essential for regulating the onset of active growth in the developing branch. Co-expression analysis revealed 413 genes with expression patterns correlated with LiCPK19 (Fig. 9A), indicating a co-regulated gene network that likely plays a role in bud-to-branch development. Subsequent GO and enrichment analysis of these co-expressed genes highlighted their involvement in BP, CC, and MF, with significant enrichment in binding, transporter activity, and oxidoreductase functions that impact biosynthesis occurring in the cell membrane and cell wall organization (Fig. 9B). These findings suggest that LiCPK19 regulates key cellular processes involved in the structural changes required for branch development. Additionally, KEGG enrichment analysis indicated that these genes are predominantly implicated in carbohydrate metabolism, flavonoid biosynthesis, and biosynthesis of other secondary metabolites (Fig. 9C). These pathways are essential for cellular growth and adaptation during bud-to-branch development, reinforcing the idea that LiCPK19 plays a role in regulating metabolic and biosynthetic pathways during branch formation.

Fig. 9.

Co-expression analysis of LiCPK19 and associated genes throughout bud to branch development in crape myrtle. A The co-expression trends of LiCPK19 with other genes were depicted, illustrating how the expression of these genes correlated across various stages of development. B The GO enrichment analysis section presented the BP, CC, and MF that are significantly overrepresented among the co-expressed genes. C A Sankey diagram was utilized to visually represented the KEGG pathway enrichment analysis of co-expressed genes

Conversely, LiCPK29 demonstrated higher expression levels in the bud and decreased expression in the branch, suggesting a role in the early stages of bud initiation or dormancy. This expression pattern indicates that LiCPK29 may be involved in regulating the initial transition phases of bud-to-branch development. Co-expression analysis identified 980 genes with similar expression profiles to LiCPK29 (Fig. 10A), which suggests a regulatory network that is co-expressed with LiCPK29 during early bud formation. To understand the biological significance of these co-expressed genes, GO enrichment analysis was performed, which revealed roles in binding, organelle localization, nuclear function, and metabolic processes related to organic substances, macromolecules, and nitrogen compounds (Fig. 10B). These roles suggest that LiCPK29 is involved in cellular structure, metabolic regulation, and growth-related processes during the early stages of bud formation. Additionally, KEGG enrichment analysis suggests that these genes were primarily associated with transporters, signaling and cellular processes, and membrane trafficking (Fig. 10C). This indicates that LiCPK29 may regulate cellular communication and trafficking processes that are essential for proper bud-to-branch development.

Fig. 10.

Co-expression analysis of LiCPK29 and associated genes during bud to branch development in crape myrtle. A The co-expression trends of LiCPK29 were illustrated alongside other genes, revealing how their expression levels change in concert. B A bubble chart visualization was employed for GO enrichment analysis, depicting the distribution of co-expressed genes across various BP, CC, and MF. C The KEGG pathway enrichment analysis was presented through a grouped point bar chart, which visually represented the involvement of co-expressed genes in various biological pathways as annotated in the KEGG

Discussion

CPKs are widely recognized for their crucial roles in plant growth and development, as well as their essential functions in mediating responses to both biotic and abiotic stresses [13, 33]. Since the CPK gene family in crape myrtle has not been comprehensively investigated in past, and merge it with succeeding statement by modifying it as the principal finding of this research that of the identified total of 38 LiCPKs in crape myrtle genome, comprising 35 LiCPKs and 3 LiCRKs gets noteworthy attention. The number of CPKs in crape myrtle was comparable to that found in Arabidopsis (34) [9], rice (33) [10], maize (40) [11], poplar (30) [12], and bamboo (30) [16] However, the number of LiCRKs in crape myrtle was significantly lower compared to other species such as A. thaliana (8 AtCRKs), O. sativa (5 OsCRKs), and P. trichocarpa (9 PtCRKs). The reduction in the number of LiCRKs might be attributed to the loss of a subset of LiCDPKs during the evolutionary process, potentially due to functional redundancy within the CPK and CRK gene families [12]. In addition, akin to CPKs observed in other plant species like Arabidopsis and rice, LiCPKs were categorized into five distinct groups. The distribution of LiCPKs across these groups showed a higher frequency in groups CPK I-III, while group CPKIV contained a comparatively smaller number of LiCPKs. Moreover, crape myrtle exhibited a similar count of CPKs in CPK III and IV, but the numbers of CPKs clustered into CPK I and II varied significantly when compared to those in other species. This variation supported the hypothesis that the disparity in CPK counts among different plant species perhaps is predominantly influenced by the number of CPKs clustered into CPK I or CPK II, serving the idea of independent evolution of gene stocks across various organisms [34]. The phylogenetic classification mirrored patterns previously observed in Arabidopsis and other species [2, 35]. The distribution of CPKs within each group varied among the three species, with crape myrtle and Arabidopsis CPKs being more frequently clustered together compared to rice. This closer genetic relationship suggested that the CPK gene family in crape myrtle shares more evolutionary similarities with dicotyledonous species, such as Arabidopsis, than with monocotyledonous species like rice.

It is important to note that introns are integral to the genetic structure adding to the functional security and the complexity of genomes [36]. In the present survey of LiCPKs, the majority were observed to have a consistent number of introns, numbering six or seven. This pattern is in line with findings from other plant species such as potato and wheat, where a similar intron count was reported [7]. However, there were exceptions within the LiCPKs, such as LiCPK19, LiCPK22, and LiCPK26, each of which contained eight introns indicating a potential for variability even within closely related genes. Notably, the LiCPKs that are part of the CPKIV group stood out with an increased number of introns, totaling eleven. This was a higher count compared to the majority of LiCPKs and suggested a possible expansion or diversification of intronic regions within this specific subgroup. These observations hint at an evolutionary trend where LiCPKs may have experienced an increase in intron number over time. The increase in intron number has several implications for gene function and regulation [37]). Introns can contain regulatory elements that influence gene expression, such as enhancers and silencers, and their presence can affect mRNA splicing, leading to the production of different protein isoforms [38]. Previous studies have shown that variations in intron-exon structures are linked to functional differentiation and adaptive responses in plants [39, 40]. The variation in intron number among LiCPKs might therefore reflect an adaptation to different regulatory demands or environmental pressures contributing to the functional specialization and diversification of these genes in crape myrtle.

Exploring the mechanisms of plant evolution and gene family expansion reveals that tandem and segmental duplications are key drivers of genomic diversification. Among these, segmental duplication stands out as the predominant mode in the evolution of angiosperms [41]. In our investigation, we conducted an intra-genome collinearity analysis of CPKs, uncovering 31 pairs of duplicated LiCPKs. This analysis revealed a high degree of inter-genome collinearity in L. indica, P. trichocarpa, and S. purpurea, all of which were attributed to segmental duplication events. This finding underscores the significant role that segmental duplication has played in the expansion of LiCPK gene family. The gene duplication and collinearity analysis suggested that LiCPKs have been subjected to strong negative selection. This type of selection resulted in the conservation of gene function, potentially limiting the degree of functional differentiation that can occur among duplicated genes. The implication was that while gene duplication provides the raw material for evolutionary innovation, the pressure of negative selection may constrain the extent to which these duplicated genes can diverge in function. Our findings aligned with studies on CPKs in other plant species such as wheat and Medicago truncatula [7, 42]. These studies reported a pattern of strong negative selection acting on CPKs suggesting a conserved evolutionary pathway is in place among different plant lineages. This conservation implies that while the specific functions of CPKs may vary to some extent among species, the underlying evolutionary forces shaping these genes are similar. Comparative analysis with other plant species helps not only to elucidate the general principles governing gene family evolution in plants but alludes to infer upon specific adaptations that may have occurred in crape myrtle and related species.

Flower color, a key factor in the visual appeal of ornamental plants, is influenced by genetic and environmental factors. Understanding the regulatory mechanisms behind flower color development is crucial for breeding new ornamental varieties, as pigments like carotenoids, flavonoids, and alkaloids are the primary determinants of color [43, 44]. In our study, the differential expression analysis of 33 LiCPKs across various flower colors revealed distinct gene clustering patterns, suggesting potential functional overlap or specialization among LiCPKs. This clustering indicated that these genes may be involved in different stages of flower coloration and regulatory networks that govern pigment production. Specifically, LiCPK31 and LiCRK2 which were linked to the blue module in our analysis exhibited elevated expression in certain flower colors, indicating their likely involvement in regulating pigment biosynthesis or modulation. These findings were further supported by GO and KEGG enrichment analysis, which highlighted key biological processes and pathways associated with the differentially expressed LiCPKs. The significant enrichment for GO terms related to binding, transferase, and oxidoreductase activities suggested that these genes play an important role in the synthesis and regulation of pigments, potentially through molecular conversions or binding mechanisms. Additionally, the involvement of these genes in KEGG pathways such as secondary metabolite biosynthesis, plant hormone signaling, environmental adaptation, and carbohydrate metabolism further emphasized their critical roles in flower coloration.

Previous studies have highlighted the critical role of CPKs in flower development with several investigations pointing to their involvement in various stages of flower morphogenesis and reproductive processes. For example, in Pharbitis nil, the transcript levels of PnCDPK1 transiently increased following the conversion of a leaf bud to a flower bud, suggesting a developmental regulatory role in floral morphogenesis [45]. Similarly, in Arabidopsis, AtCPK33 regulated flowering time and flower structure through direct phosphorylation of the flowering regulator FD at the shoot apical meristem [22, 46]. Additionally, In Lycopersicon esculentum, the transcript levels of LeCRK1 accumulated during fruit ripening and floral tissues, further supporting the involvement of CPK family members in reproductive organ development [47]. Our results revealed that LiCPKs in L. indica show clear developmental and tissue-specific expression patterns, underscoring their regulatory roles in flower maturation and bud-to-branch development. Among the 11 LiCPKs identified with stage-dependent expression, several genes such as LiCPK4, LiCPK13, LiCPK17, and LiCPK20 were strongly expressed during late flower maturation, suggesting their involvement in pigment accumulation and petal differentiation. Conversely, LiCPK1, LiCPK11, LiCPK15, and LiCPK19 showed higher expression at early floral stages, indicating potential functions in early morphogenetic signaling and cell differentiation. WGCNA further linked these LiCPKs to developmental modules associated with flower coloration and tissue maturation. In particular, LiCPK11, LiCPK15, and LiCPK19 clustered within a co-expression module enriched for GO and KEGG terms related to transferase activity and secondary metabolite biosynthesis, processes directly connected to pigment metabolism and petal development. These findings suggests that LiCPKs may act as central regulators integrating Ca²⁺ signaling with biochemical pathways underlying petal coloration and maturation.

Recent studies have highlighted the essential roles of CPKs in orchestrating plant growth and development. Munemasa et al. [48] demonstrated the involvement of CPKs in a variety of plant developmental processes. In M. truncatula, the suppression of MtCDPK1 resulted in a pronounced reduction in both root hair and root cell elongation [49]. Similarly, ectopic expression of BnaCPK2 in Brassica napus was linked to ROS accumulation and cell death [50]. In A. thaliana, AtCPK17 and AtCPK34 were implicated in the transduction of Ca²⁺ signal transduction, which is important for pollen tube tip growth [51]. In addition, the preferential accumulation of PnCDPK1 in petals and sepals suggested its role in flower morphogenesis [45]. Moreover, AtCPK28 in Arabidopsis was identified as a key regulatory factor in stem elongation and vascular tissue development [52]. Our results clearly consistent previous studies, provide new insights into the multifaceted roles of CPKs in plant morphogenesis. Specifically, LiCPK19, which exhibited down-regulation in the bud stage followed by up-regulation in the branch’s lower tissues, appeared to act as a modulator during early developmental phases. Co-expression analysis identified 413 genes co-regulated with LiCPK19 suggesting their collective involvement in diverse BP. The significant enrichment of genes associated with binding, transporter activity, and oxidoreductase functions, along with biosynthetic processes in the cell membrane and cell wall, points towards their coordinated role in the maturation of branch structures.

In contrast, LiCPK29, with its higher expression in the bud stage and subsequent downregulation in the branch, posits to play a positive regulatory role in the initial stages of development. The co-expression analysis identified 980 genes with analogous expression patterns presents an extensive regulatory network. GO and KEGG enrichment analysis were performed to further investigate the biological function and pathways associated with the co-expressed LiCPKs. The GO enrichment analysis revealed that the co-expressed genes were significantly enriched in binding activities including protein binding and ion binding, which are crucial for their interaction with substrates involved in cellular signaling and metabolic processes. Additionally, genes associated with organelle localization, particularly those linked to the endoplasmic reticulum and plasma membrane, were identified, suggesting that LiCPKs may regulate protein trafficking and cellular communication. Nuclear function-related terms such as DNA binding and transcription factor activity, were also enriched, supporting the role of LiCPKs in regulating gene expression and cellular organization. KEGG pathway analysis highlights the involvement of these co-expressed genes in critical pathways such as metabolic processes, signal transduction, and cell wall biosynthesis. These findings collectively suggests that LiCPKs are integral to the regulation of structural and functional maturation in plant development, particularly in bud and branch formation.

Overall, these findings provide a comprehensive framework for understanding the functional and regulatory diversity of LiCPKs in L. indica. Future studies focusing on functional genomics approaches, such as tissue-specific expression validation, gene knockout or overexpression assays, and protein-protein interaction studies, will be essential to elucidate their precise biological roles. These insights may ultimately support molecular breeding strategies aimed at improving ornamental traits, particularly flower coloration and branch architecture, in L. indica and related species.

Conclusions

In this study, we identified 38 LiCPKs from the L. indica genome and analyzed their phylogenetic relationships, conserved motifs, and gene structures, revealing important insights into their evolutionary history. Collinearity analysis showed that gene duplications contributed to the expansion and diversification of LiCPKs, while promoter analysis identified key regulatory elements involved in their expression in response to environmental factors. Expression profiling showed that LiCPKs exhibit distinct tissue- and stage-specific expression patterns, particularly across different flower color phenotypes and during bud-to-branch development suggesting their key roles in regulating pigment biosynthesis, flower maturation, and structural differentiation. Co-expression and pathway enrichment analysis further demonstrated that LiCPKs are integrated into complex regulatory networks involving secondary metabolite biosynthesis, hormone signaling, and cell wall remodeling. These findings enhance our understanding of the roles of LiCPKs in plant development and flower morphogenesis. The identification of LiCPKs involved in flower pigmentation and maturation provides a basis for developing genetic tools to improve ornamental traits in L. indica and related species.

Authors’ contributions

Hui Wei, Project administration, Writing - original draft, Writing - review & editing; Yi Cao, Supervision, Validation, Investigation; Hewenyan Pan, Formal analysis; Xiaoxi Zhou, Conceptualization; Guoyuan Liu, Data curation, Resources; Bolin Lian, Software; Fei Zhong, Supervision; Jian Shi, Supervision; Lei Zhang, Writing - review & editing; Jian Zhang, Funding acquisition, Writing - review & editing.

Funding

This work was funded by Basic Research Program of Jiangsu (BK20250951), Basic Research Project of Nantong (JC2023104), Jiangsu Province College Students’ Innovation and Entrepreneurship Training Program (202410304106Y), and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

Data availability

The datasets generated and/or analyzed during the current study are available in Phytozome and the China National GeneBank DataBase (CNGBdb) repository, with the following accession numbers: the datasets on reference genome sequences and annotations for A. thaliana, O. sativa, E. grandis, V. vinifera, P. trichocarpa, and S. purpurea were obtained from the Phytozome repository (https://phytozome-next.jgi.doe.gov/); the datasets generated and analyzed during the current study on L. indica are available under accession number CNP0003018; the raw RNA-seq data on salt treatment for this study are available under accession number CNP0003991; the raw RNA-seq data on the developmental stage of flowers for this study are available under accession number CNP0001693; and the raw RNA-seq data on bud-to-branch development for this study are available under accession number CNP0003991. The datasets and materials used during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.