Identification and functional evaluation of cyclin-dependent kinase genes reveals that CDKB1;1 and CDKB2;2 contribute to the balance of mitosis and endoreduplication in Medicago truncatula nodule

Abstract

Background

Cyclin-dependent kinases (CDKs) critically regulate plant cell cycle transitions, including mitosis-to-endoreduplication switches essential for growth and adaptation. In Medicago truncatula, nodules form through symbiotic nitrogen fixation with rhizobia. The terminal differentiation of bacteroids within nodule cells is critical for efficient nitrogen fixation. To maintain and optimize the functionality of these differentiated symbiosomes, host nodule cells undergo repeated rounds of endoreduplication. However, which CDKs are involved in regulating endoreduplication in nodule cells to support effective symbiotic nitrogen fixation remains largely unknown.

Results

We identified and characterized 29 CDK genes (15 CDKs and 14 CDKLs) classified into eight conserved subgroups. These genes displayed diverse exon/intron structures and protein motifs, with CDKA, CDKB, and CDKL subfamilies showing strong conservation with Arabidopsis thaliana. Expression analysis revealed specific downregulation of CDKB1;1, CDKB2;2, and CDKL13 in nodule infection to fixation zones. Protein–protein interaction (PPI) network and Gene ontology (GO) analyses demonstrated CDKB1;1 and CDKB2;2 involvement in cell cycle regulation. Overexpression of CDKB1;1 or CDKB2;2 disrupted endoreduplication and nitrogen fixation, with CDKB1;1 having the most pronounced effect, while CDKL13 appeared dispensable for symbiosis.

Conclusion

Our study presents the comprehensive genome-wide analysis of the CDK gene family in M. truncatula, demonstrating that the essential role of CDKB1;1 and CDKB2;2 downregulation in symbiotic nitrogen fixation and endoreduplication offers new insights into cell cycle regulation in nodules. It also identifies potential targets for improving nitrogen fixation efficiency in legumes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-025-07365-1.

Introduction

The cell cycle operates throughout the entire life cycle of a plant, playing a pivotal role in seed germination, vegetative growth, and reproduction, all of which depend on its precise regulation [1–3]. Mitosis, as the primary mode of cell division during the cell cycle, proceeds through four tightly regulated phases (G1, S, G2, and M) to ensure the accurate transmission of genetic information and cellular resources to daughter cells. This orderly division is crucial for plant growth and development, as well as for adapting to environmental conditions [4–6]. Endoreduplication, a specialized variant of the cell cycle, is widespread in plants [7, 8]. To adjust to particular ecological conditions, plants employ endoreduplication to increase cellular ploidy levels, facilitating cell expansion, enhancing metabolic activity, and improving nutrient uptake [9]. This process typically occurs in large, metabolically active, or highly specialized plant cells and tissues, such as those in the endosperm, trichomes, and fruits [10, 11]. Thus, the meticulous coordination between mitosis and endoreduplication is vital for sustaining normal plant growth and development and mounting effective responses to environmental challenges.

Mitosis and endoreduplication in plants share a common set of core regulatory components, including cyclins (CYCs), cyclin-dependent kinases (CDKs), and CDK inhibitors [12]. Among these, CDK, as a central regulator of the cell cycle, determines whether a cell enters mitosis or switches to the endoreduplication pathway through precise control of its activity [7, 11]. The plant CDK family consists of eight subfamilies, CDKA–CDKG and CDKL [13, 14], among which CDKA and CDKB play indispensable roles in regulating canonical mitosis [14, 15]. Overexpression of CDKA1;1 can lead to premature entry into the S phase, causing cell cycle disruption [16]. During the G1/S phase transition, the CDKA1/CYCD complex phosphorylates the retinoblastoma-related (RBR) protein, thereby releasing E2F transcription factors and activating the expression of S-phase-related genes [6, 17]. Once cells enter S phase, the CDKA1/CYCA complex orchestrates DNA replication [18]. At the G2 to M phase transition, the plant-specific CDKB subfamily, comprising CDKB1 and CDKB2, plays a critical regulatory role. CDKB1;1 expression is initiated in the G2 phase, and its complex with CYCB exhibits peak activity in mitosis [19, 20]. Genetic analyses indicate that CDKB1;1 loss-of-function mutants display a marked G2/M phase arrest, leading to a reduced root apical meristem and delayed leaf development [21]. In reverse, ectopic expression of CDKB1;1 (CDKB1;1. N161) causes precocious mitotic entry and abnormal cell division [22]. In mitosis, CDKB1/2–CYCB complexes phosphorylate multiple substrates to precisely control chromosome dynamics and spindle assembly [23, 24]. Exit from mitosis requires APC/C–CCS52-mediated ubiquitin-dependent degradation of CYCB [25]. In ccs52a mutants, failure to degrade CYCB results in sustained mitotic activity and the formation of multinucleated cells [26, 27]. In comparison, endoreduplication, a variant of the cell cycle, is defined by the suppression of mitotic CDKB1;1/CYCB activity and the sustained periodic activation of S phase CDKA/CYCA complexes [28, 29]. In Arabidopsis, the inactivation of the CDKB1;1/CYCA2;3 complex by the CCS52A1-APC/C serves as a key molecular switch governing the transition from mitosis to endoreduplication [22, 30]. Overexpression of CDKB1;1 suppresses endoreduplication and drives cells back into mitosis [22]. These genetic findings collectively highlight the central role of dynamic CDK activity balance in determining cell cycle mode transitions.

Symbiotic nitrogen fixation in legumes represents a highly specialized mutualistic interaction between plants and rhizobia [31]. This process entails intricate molecular interactions, signal transduction, and cellular reprogramming, ultimately leading to the formation of a specialized organ—nodule [32]. Based on the persistence of host cell proliferation, nodules are categorized as determinate or indeterminate [33]. While determinate nodules consist of a uniform population of symbiotic cells, the indeterminate nodules can be divided into five distinct zones [34]. Meristem zone, where G0 phase cells are reactivated to sustain continuous mitotic activity and generate new cells [35]. Infection zone, infection threads penetrate the nodule cell membranes, allowing rhizobia to be released into host cells and form symbiosomes [36]. Cells exiting the meristem transition from mitosis to endoreduplication, resulting in cell enlargement and the provision of intracellular space for rhizobial accommodation [32, 37]. Differentiation zone, symbiosomes undergo terminal differentiation, characterized by marked elongation and enlargement [33]. Fixation zone, where terminally differentiated symbiosomes actively carry out nitrogen fixation [38]. Senescence zone, both symbiotic cells and bacteroids undergo degeneration [39]. In M. truncatula, a model legume species, the terminal differentiation of bacteroids within nodule cells is a critical factor for effective nitrogen fixation [31, 33]. To support and maintain the function of differentiated symbiosomes, nodule cells undergo multiple rounds of endoreduplication [8, 40]. Compared to diploid cells in the meristem, nitrogen-fixing nodule cells can expand up to 100-fold in volume [41]. Their ploidy levels may reach 64 C, corresponding to a 32-fold increase in DNA content relative to diploid cells [42]. This extensive endoreduplication is essential for the high-level expression of symbiosis-related genes in nodule cells [40]. Therefore, elucidating the regulatory mechanisms underlying endoreduplication in nodule cells is crucial for understanding bacteroid-induced polyploidy in symbiotic development.

Nodule cells undergo endoreduplication during symbiotic nitrogen fixation, serving as a well-established model for studying this process [35]. This enables an in-depth investigation of how key regulatory factors, such as CDK, CCS52A, and the APC/C complex, are modulated to bypass mitosis (M phase). In M. truncatula, CCS52A, a substrate-specific activator of APC/C, is upregulated during nodule development and across different nodule zones, promoting the transition from mitosis to endoreduplication [43, 44]. Silencing of the CCS52A gene significantly reduces the ploidy level of nodule cells (from 64 C to below 16 C; C, C-value), accompanied by a marked decline in nitrogen fixation efficiency, confirming its central regulatory role in endoreduplication [43]. In Lotus japonicus, the VAG1 gene encodes a protein homologous to Arabidopsis ROOTHAIRLESS 1/HYPOCOTYL 7, a core component of plant DNA topoisomerase VI, and directly participates in the regulation of endoreduplication. Studies have shown that in vag1-1 mutants, the proportion of diploid nodule cells significantly increases, while the number of highly polyploid cells (>4 C) is markedly reduced [45]. Although the regulatory mechanisms of endoreduplication in nodule cells have been partially elucidated, it remains unclear how CDKs regulate the mitosis-to-endoreduplication switch and thereby affect nitrogen-fixing symbiosis in M. truncatula nodules.

This study aims to systematically characterize the CDK gene families in M. truncatula, addressing the critical knowledge gap in understanding which specific CDKs mediate nodule endoreduplication and sustain nitrogen-fixing symbiosis. Our study presents the comprehensive genome-wide analysis of the CDK gene family in M. truncatula, revealing lineage-specific expansion patterns and conserved structural features. The expression analysis showed that CDKB1;1, CDKB2;2, and CDKL13 are highly expressed in roots but significantly downregulated in nodules, particularly in the infection-to-fixation zones. Functional characterization demonstrates that ectopic expression of CDKB1;1 or CDKB2;2 impairs symbiotic nitrogen fixation and disrupts endoreduplication. These findings establish CDKB1;1 or CDKB2;2 as a crucial molecular switch controlling the mitosis-to-endoreduplication transition during nodule development, providing new insights into cell cycle regulation in symbiotic nitrogen fixation.

Materials and methods

Plant growth conditions, hairy root transformation, and rhizobia inoculation

The seeds of Medicago truncatula Jemalong A17 (wildtype) were provided by Professor Huairong Pan’s laboratory (College of Biology, Hunan University, China), where they are maintained under standard conditions. No additional permits were required for their use in this study. The plants were grown under controlled conditions with a 16 h light/8 h dark photoperiod at 22 °C. Hairy root transformation was performed using Agrobacterium rhizogenes strain ARqua1, following established protocols. Transgenic roots were selected based on their vector-conferred antibiotic resistance, using Fahraeus medium supplemented with 8 μg/mL kanamycin.

For rhizobial inoculation, we employed Sinorhizobium meliloti Rm1021 nifH::GUS and Sinorhizobium medicae ABS7 hemA::LacZ. Overnight liquid cultures of rhizobia were centrifuged and resuspended in ½-strength basic nodulation medium to an OD₆₀₀ of 0.05. Each M. truncatula plant, grown in green zeolite, was inoculated with 5 mL of the rhizobial suspension [33].

Molecular cloning

KOD Plus DNA polymerase (TOYOBO, Japan) was used for amplifying target DNA and cDNA fragments. Ligation was performed using the ClonExpress Ultra One Step Kit (Vazyme, China) through In-fusion cloning. For constructing pDNF1-CDKB1;1/CDKB2;2/CDKL13-pKGW-RR, we amplified the full-length genomic/coding sequences of these three genes (3267 bp, 1743 bp, and 2913 bp, respectively) and directly ligated them into the target vector (Table S3). For promoter-pMDC163 constructs, we amplified both full-length and truncated promoter regions along with 5' UTR sequences of CDKB1;1 (2691 bp), CDKB2;2 (2653 bp), and CDKL13 (2118 bp), which were then directly cloned into pMDC163. All cloning primers are listed in Table S4.

X-Gluc staining

X-Gluc staining was performed to evaluate: (1) rhizobial nifH::GUS reporter activity in M. truncatula nodules, and (2) promoter activities of selected CDKB1;1/CDKB2;2/CDKL13 genes in transformed roots and nodules. The staining protocol followed established methods with the following modifications: Freshly harvested nodule and root tissues were fixed in GUS staining buffer containing 10 mM X-Gluc and incubated at 37 °C for approximately 12 h. Post-staining, samples were destained through sequential ethanol washes (75%) until complete tissue clearing was achieved. Stained tissues were examined and photographed using a TL500 microsystem (Leica, Germany) coupled with a DMC6200 digital camera.

RNA extraction, cdna synthesis, and quantitative real-time PCR (qRT-PCR)

Total RNA was extracted from nodules using TRIzol reagent (Invitrogen, 15596026). To remove genomic DNA contamination, the RNA samples were treated with the Turbo DNA-free kit (Invitrogen, AM1907). First-strand cDNA was synthesized from 1 µg of total RNA using the iScript cDNA Synthesis Kit (Bio-Rad, 1708890) according to the manufacturer’s instructions.

qRT-PCR was performed in 20 µL reactions using PerfectStart® Green qPCR SuperMix (Transgen Biotech, AQ601-01) on an Eppendorf Mastercycler ep Realplex system. Each reaction included three biological replicates, and melting curve analysis was conducted to confirm amplification specificity. The threshold cycle (Ct) values were averaged across replicates, and gene expression levels were normalized to the reference gene MtPDF2 (MtrunA17_Chr6g0484701), which has been validated as a stable internal control in M. truncatula [46]. Primer sequences are provided in Supporting Information Table S4.

Flow cytometry technique

To determine the endoreduplication levels of Control, pDNF1::CDKB1;1 pDNF1::CDKB2;2 and pDNF1::CDKL13 nodule cells, at the specified timepoints, nodules inoculated with correspondent S. meliloti Rm1021 nifH::GUS strains were rinsed 3 times with sterile water, the nodules were homogenized in cold Galbraith buffer (45 mM MgCl₂, 30 mM C₆H₅Na₃O₇·2H₂O (trisodium citrate dihydrate), 20 mM MOPS, 0.1% (v/v) Triton X-100, pH 7.0, with 10 mM 2-mercaptoethanol added before use, to release the nuclei. The cell lysate was then passed through a 75 μm nylon mesh (Millipore, USA) into a new 1.5 mL Eppendorf tube. The extracted nuclei were stained with DAPI for 5 min were analyzed using a Beckman-Coulter ELITE ESP flow cytometer (Beckman, the United States) following the manufacturer’s instructions. The raw data were processed using the FlowJo software for detailed analysis.

Identification, phylogenetic, gene structure, conserved motifs, and protein domain analysis of CDK

To identify and classify cyclin-dependent kinase (CDK) families in M. truncatula, we performed comparative analyses using A. thaliana CDKs (AtCDKs) as references. Multiple sequence alignments were conducted, and a phylogenetic tree was constructed using MEGA7 with the neighbor-joining (NJ) method and 1000 bootstrap replicates for robustness assessment. The final phylogenetic tree was visualized and annotated using the Interactive Tree of Life (iTOL) platform (https://itol.embl.de/). To further validate the CDK classification, the conserved motifs were detected in MEME (https://meme-suite.org/meme/tools/Meme) and visualized in TBtools. Protein domain architecture was analyzed using SMART (https://smart.embl.de/), an online tool for protein domain identification. Gene structures of the M. truncatula CDKs were mapped using TBtools.

Chromosomal distribution

The chromosomal locations of predicted CDK genes in M. truncatula were retrieved from the genome sequence and the GFF3 annotation files, and the "Gene Location Visualize (Advanced)" function in TBtools was used to map and display the genes on the corresponding chromosomes and visualized using TBtools.

Analysis of gene duplication and collinearity

Identification of gene duplication events in M. truncatula was performed using the Multiple Collinearity Scan toolkit (MCScanX) with stringent BLASTp parameters (e-value < 1 × 10^–5). For comparative evolutionary analysis of CDK gene family evolution, four model species were selected: M. truncatula (reference genome), G. max, L. japonicus, and A. thaliana. Collinear gene pairs were systematically identified across these species using TBtools, followed by comprehensive visualization of both intra-genomic duplication events and inter-species syntenic relationships to elucidate CDK gene family expansion patterns.

Gene expression analysis

Tissue-specific expression profiles of CDK genes across various developmental stages were obtained from two comprehensive transcriptomic databases: MtExpress (https://medicago.toulouse.inrae.fr/MtExpress) and Symbimics (https://iant.toulouse.inra.fr/symbimics/). The expression data were analyzed and visualized as a clustered heatmap using TBtools software, revealing distinct spatiotemporal expression patterns of CDK gene family members.

Protein tertiary structure prediction, protein–protein interaction (PPI) network construction, and gene ontology (GO) analysis

To elucidate the structural conservation of key CDK proteins in M. truncatula, we performed homology-based tertiary structure modeling for CDKB1;1, CDKB2;2, and CDKL13 using SWISS-MODEL(https://swissmodel.expasy.org/). The similarity between protein pairs was quantitatively evaluated using three metrics: sequence similarity, TM-score, and RMSD. Sequence similarity was calculated through pairwise alignment using the Needleman-Wunsch available on the NovoPro website (https://www.novopro.cn/tools/needle.html). Tertiary structure alignment was performed using TM-align (https://zhanggroup.org/TM-align/), from which the TM-score and RMSD values were obtained. To explore functional interactions, we constructed a PPI network and GO Enrichment Analysis using STRING (https://cn.string-db.org/).

Results

Genome-wide identification and evolutionary analysis of CDK genes in M. truncatula

To identify CDK proteins in M. truncatula, 29 A. thaliana AtCDK protein sequences served as queries to search against the M. truncatula proteins using the BLASTP program. Following the removal of incomplete and low-similarity sequences, a total of 29 putative CDK genes were identified, comprising 15 CDKs and 14 CDKLs, which were unevenly distributed across the eight chromosomes (Fig. 1A). Chromosomes 3 and 4 harbored the highest number of CDK genes, with six each, while chromosome 5 contained the fewest, with only one CDK gene (CDKL14) (Fig. 1A). These proteins ranged from 294 (CDKA1) to 810 (CDKG1;2) amino acids, with molecular weights between 33.91 kDa (CDKA2) and 91.72 kDa (CDKG1;2) (Table S1). The isoelectric points (pI) ranged from 4.54 (CDKF1) to 9.73 (CDKE2), and instability index from 31.3 (CDKB1;1) to 60.09 (CDKL13), indicating varying structural stabilities (Table S1). The aliphatic indices of the CDK proteins ranged from 59.46 to 96.53, and all exhibited negative grand average of hydropathicity values, indicating that they are hydrophilic in nature (Table S1).

Fig. 1.

Identification and evolutionary analysis of CDK and CDKL Genes in M. truncatula. A Chromosomal localization of the 29 identified CDK genes. Gene density along the chromosomes is represented by a color gradient, with white indicating low density and blue indicating high density. B Phylogenetic analysis of CDK genes in M. truncatula and A. thaliana. The phylogenetic tree was constructed using MEGA7 software, with distinct colors indicating different CDK subgroups. C Collinearity analysis of CDK genes between M. truncatula (Mt) and three other species -A. thaliana (AT), G. max (Gm), and L. japonicus (Lj). Gray lines denote syntenic blocks between species, while red lines emphasize collinear CDK gene pairs

To elucidate phylogenetic relationships within the CDK gene family, we cgenerated an unrooted tree using 58 CDKs from M. truncatula and A. thaliana (Table S2), revealing eight distinct subgroups, CDKA-G and CDKL (Fig. 1B). Analysis of the amino acid sequences within CDK subgroups further revealed that each possesses a characteristic cyclin-binding domain specific to its subgroup (Fig. S1). While the CDKC and CDKG subgroups displayed similar gene counts in A. thaliana and M. truncatula, the CDKA, CDKE, and CDKF subgroups showed increased gene numbers in M. truncatula. However, CDKB and CDKD subgroups exhibited reduced gene numbers (Fig. 1B). Comparative collinear analysis revealed that M. truncatula shares 71 collinear CDK gene pairs with G. max, involving 22 CDKs from M. truncatula and 57 from G. max. Additionally, 36 collinear pairs were identified between M. truncatula and L. japonicus (25 CDKs from each species), and 22 collinear pairs between M. truncatula and A. thaliana, involving 16 and 14 CDKs from M. truncatula and A. thaliana, respectively. These results indicate that CDK genes exhibit significantly higher conservation in M. truncatula, G. max, and L. japonicus compared to A. thaliana (Fig. 1C). Furthermore, intraspecific synteny analysis identified eight segmentally duplicated CDK gene pairs within the M. truncatula genome (Fig. S2), indicating that local gene duplication events may have driven the expansion of the CDK gene family in M. truncatula.

Subfamily-specific conservation and divergence in CDK gene structures and protein motifs

The evolution of multigene families is likely driven by structural diversity and divergence in conserved protein domains [47, 48]. To further explore the structural diversity of CDK genes in M. truncatula, exon/intron organization was examined by aligning full-length cDNA sequences with their corresponding genomic DNA. Our results showed that the coding regions of CDK genes are interrupted by 1 to 18 introns (Fig. 2A). Except for CDKL14, 28 CDK genes contain untranslated regions (UTRs) at both the 5' and 3' ends. Among all CDK subgroups in M. truncatula, the CDKA subgroup exhibits the most conserved gene architecture, typically containing eight coding sequences (CDS), consistent with AtCDKA1 (Fig. 2A and Fig. S3A). Similarly, CDKB1;1 shares a highly comparable exon/intron structure with its AtCDKB1;1. In contrast, CDKB2;2 and members of the CDKC subgroup each possess one fewer exon than their respective A. thaliana counterparts. The exon numbers in the CDKD and CDKL subgroups are identical between the two species, indicating strong structural conservation. Interestingly, the CDKE subgroup in M. truncatula displays the most complex gene structure, with up to 16–19 exons per gene, which contrasts with the simpler configuration in A. thaliana. CDKF genes exhibit the simplest architecture, comprising only two CDSs (Fig. 2A and Fig. S3A). These results reveal substantial structural diversity among M. truncatula CDK genes, with conserved architectures in certain subfamilies and lineage-specific variations in others, reflecting both functional specialization within the gene family.

Fig. 2.

Comparative analysis of phylogenetic relationships, gene structures, and conserved motifs of the CDK family in M. truncatula.A Phylogenetic tree and Exon–intron organization of 29 CDK proteins, constructed using MEGA7 and TBtools. Distinct clades are color-coded. Exons (orange), untranslated regions (UTRs, blue), and introns (black lines) are indicated. A table showed the number of exons (E), introns (I), coding sequence (C), and UTRs (U) of CDK genes. B Protein motif structure analysis of 29 CDKs in M. truncatula.The identified protein motifs are represented by colored boxes (Motif 1-Motif 10). Black lines indicate relative protein lengths. A table shows whether motifs 1–10 are present in various CDK protein sequences. Gray dots indicate absence, while orange dots indicate presence. C and D Motif 1 and Motif 8 are specific to CDKL proteins. Protein domains were predicted using the MEME database

CDKs are a class of kinases with highly conserved functional domains. We analyzed the conserved domains of 29 CDK proteins from M. truncatula using the SMART database and found that they all contain a serine/threonine kinase catalytic (S_TKc) domain (Fig. S4A), which is highly similar to the conserved domains of the 29 CDK proteins from A. thaliana (Fig. S4B). To further analyze the differences in M. truncatula CDK protein sequences, we used the MEME database to identify 10 conserved motifs, the core amino acid sequences of which show high similarity to those found in A. thaliana (Fig. 2C-D, Fig. S3C, Fig. S5G, and Fig. S6), indicating strong evolutionary conservation. Remarkably, each CDKL gene in both M. truncatula and A. thaliana contains all ten conserved motifs (Fig. 2B and Fig. S3B), highlighting a high degree of structural conservation within the CDKL subfamily across species. Compared to CDKL, 13 CDK proteins in M. truncatula lacked motif 1, motif 8, and motif 10, whereas CDKs in A. thaliana were only missing motif 1 (Fig. 2C-D and Fig. S3C). Additionally, variation in motif composition was observed among different CDK subgroups. Both M. truncatula and A. thaliana CDKA, CDKD, CDKE, CDKF, and CDKG subgroups lacked motif 1 and motif 10, and in M. truncatula, these subgroups also showed the absence of motif 8 (Fig. 2B). Notably, some CDKs even lacked motif 3 and motif 4. Among all subgroups, CDKB proteins in both species were consistently deficient in motif 1, motif 3, motif 8, and motif 10, suggesting potential functional similarity across species (Fig. 2B).

To gain deeper insights into the structural divergence of CDK proteins, we conducted a comprehensive analysis of the S_TKc domain architecture and its associated conserved motifs in M. truncatula (Fig. S5). The S_TKc domain sequences from all 29 CDK proteins were extracted and aligned (Fig. S5A-F). Our analysis revealed that a complete S_TKc domain encompasses motifs 2–9 (Fig. S5), with motif 8 showing particularly strong conservation within the CDKL subfamily, consistent with the motif distribution pattern observed (Fig. 2B). Interestingly, motif 1 and motif 10 were not found within the S_TKc domain, which is consistent with their exclusive presence in CDKL subfamily members (Fig. S5A-F and Fig. 2B), further emphasizing a key structural distinction between CDKs and CDKLs.

Tissue-specific and symbiosis-responsive expression patterns of CDK genes in M. truncatula.

Tissue-specific expression analysis is essential for elucidating the functional specificity of CDK genes across various tissues in M. truncatula. To analyze their expression patterns, we utilized the MtExpress database [49], which provides comprehensive RNA-seq data, to assess CDK gene expression across a range of tissues, including leaves, inflorescence apices, vegetative apices, embryos, seed coats (26 d), mature seeds, shoots, roots, and nodules. The results demonstrate that the majority of CDKs are predominantly expressed in tissues with active mitosis, such as leaves and roots, while exhibiting the lowest expression levels in embryos (Fig. 3A). The specific expression of CDKL1 in inflorescence apices suggests a potential role in inflorescence development, floral organ formation, or stem cell maintenance. CDKL14 is expressed in tissues characterized by extensive cell differentiation, such as embryos, seed coats (26 d), mature seeds, and shoots. This could be a pointer to the significant role played in its involvement in embryogenesis, seed development, and seedling growth (Fig. 3A). Significantly, M. truncatula, as a model legume for studying nitrogen-fixing symbiosis [31], possesses a specialized root organ—the nodule—which is absent in A. thaliana [38]. In M. truncatula nodules, several CDK genes, including CDKB1;1, CDKB2;2, CDKL1, CDKL2, CDKL8, and CDKL13, exhibit markedly low expression levels (Fig. 3A). Spatial expression analysis of mature nodules revealed that CDKB1;1, CDKB2;2, and CDKL13 are highly expressed in the meristem zone, where mitosis is active, yet they are downregulated in the infection to fixation zones, where endoreduplication predominates (Fig. 3B). Thisnodule-specific suppression of CDKB1;1, CDKB2;2, and CDKL13 suggests that downregulation of these CDKs may be functionally relevant to endoreduplication.

Fig. 3.

Spatiotemporal expression profiles of CDK genes in M. truncatula. A Tissue-specific expression profiles of 29 CDK genes. The analyzed tissues include leaves (L), inflorescence apices (IA), vegetative apices (VA), embryo (E), seed coat (26 d, SC), mature seed (MS), shoot (S), root (R), and nodule (N). The color scale represents normalized expression levels (log2-transformed values). B Heatmap of CDKs gene expression levels in different zones of mature nodules. ZI, Meristem zone; ZIId, Distal infection zone; ZIIp, Proximal infection zone; ZI, Infection Zone; ZIII, Fixation zone. The color scale represents normalized expression levels (log2-transformed values). C Expression heatmap of CDK genes at different time points after rhizobial inoculation. Hour post-inoculation(hpi), MO, mock inoculation, Sm, the plants inoculation with Sinorhizobium meliloti 2011. The color scale represents normalized expression levels (log2-transformed values)

Compared to mature nodules, during the early stages of nodule formation of M. truncatula, the root endodermis cells undergo extensive cell division, exhibiting an active cell cycle and high CDK expression levels [50, 51]. We systematically analyzed their dynamic expression patterns in M. truncatula roots at 48 and 96 h post-inoculation (hpi) with Sinorhizobium meliloti 2011. The results showed that CDKB1;1, CDKB2;2, CDKC2, CDKD1;1, CDKL2, CDKL3, and CDKL9 were transcriptionally induced following rhizobial infection (Fig. 3C). On the other hand, expression of CDKC1, CDKE2, CDKG1;1, CDKG1;2, CDKG2, CDKL11, and CDKL13 was specifically repressed in response to inoculation (Fig. 3C). In summary, CDKB1;1 and CDKB2;2 are highly expressed in mitotically active nodule tissues but are downregulated in endoreduplication-dominant regions, whereas CDKL13 exhibits consistently low expression across all nodule developmental stages.

Validation of CDKB1;1, CDKB2;2, and CDKL13 showed dramatically lower expression in nodules

To validate the expression profiles of three key CDK genes obtained from the MtExpress database in M. truncatula, we performed RT-qPCR analysis on CDKB1;1, CDKB2;2, and CDKL13 in wild-type roots and nodules. The results confirmed their significantly reduced expression in nodules compared to roots (Fig. 4A-C). To more clearly assess the expression patterns of CDKB1;1, CDKB2;2, and CDKL13 in root and nodules of M. truncatula, approximately 2500 bp of their promoter sequences were cloned to generate promoter–GUS fusion constructs (pCDKB1;1::GUS, pCDKB2;2::GUS, and pCDKL13::GUS) (Table S3). These constructs were introduced into wild-type plants via Agrobacterium rhizogenes-mediated hairy root transformation to obtain composite transgenic plants with positively transformed roots. Following inoculation with Sinorhizobium medicae ABS7 hemA::lacZ at different time points, CDKB1;1 and CDKB2;2 exhibited strong β-glucuronidase (GUS) activity in roots, nodule primordia, and the meristem zones of mature nodules (Fig. 4D), those regions characterized by active mitotic activity [37], but were markedly repressed in the infection and fixation zones, where endoreduplication predominates [52]. Conversely, CDKL13 exhibited root-specific expression, showing no detectable activity in nodule primordia, meristem, or mature nodule zones (Fig. 4D). The findings demonstrate a progressive downregulation of CDKB1;1, CDKB2;2, and CDKL13, with this spatially restricted repression. Strikingly, CDKB1;1 and CDKB2;2 show specific expression in the meristem zone of nodules, while CDKL13 is not expressed in any nodule zones. This distinct expression pattern suggests that CDKB1;1, CDKB2;2, and CDKL13 may differentially regulate endoreduplication in nodule cells.

Fig. 4.

Downregulation of CDKB1;1, CDKB2;2, and CDKL13 in mature nodules. A to C. qRT-PCR validation of CDKB1;1, CDKB2;2, and CDKL13 expression in uninoculated roots versus 21 dpi nodules. Expression was normalized to MtPDF2. Data represent means ± SD of three biological replicates. Statistical differences were determined using Student's T-test. *: significant differences between treatments at p ≤ 0.05. **: significant differences between treatments at p ≤ 0.01. ***: significant differences between treatments at p ≤ 0.001. D Promoter-GUS reporter assays showing spatiotemporal expression patterns. CDKB1;1 and CDKB2;2 promoters were active in nodule primordia but suppressed in infection-to-fixation zones, while CDKL13 showed no detectable expression. dpi, days post-inoculation; wpi, week post-inoculation. Scale bars = 750 μm

CDKB1;1, CDKB2;2, and CDKL13 are structurally and functionally conserved kinases

AtCDKB1;1 and AtCDKB2;2 play essential roles in the cell cycle [53], with their expression tightly associated with cell division activity in meristems and other proliferative tissues [20]. These kinases are involved in regulating mitotic progression, and their functional inhibition has been linked to the activation of endoreduplication [28]. Members of the CDKL subfamily in animals are implicated in transcriptional regulation, RNA processing, intercellular communication, and cell cycle control [54]. Through comprehensive bioinformatic analyses, we identified three CDK genes, CDKB1;1, CDKB2;2, and CDKL13, that are specifically downregulated in nodules (Fig. 4), suggesting their potential involvement in initiating endoreduplication during nodule development. To further assess their function, we performed protein structure modeling and homology comparisons. The predicted tertiary structure of CDKB1;1 shared 90.4% sequence similarity with its A. thaliana homolog AtCDKB1;1, with a TM-score of 0.98 and an RMSD of 0.86 Å. Similarly, CDKB2;2 exhibited 90.9% sequence similarity with AtCDKB2;2, accompanied by a TM-score of 0.98 and an RMSD of 1.08 Å. CDKL13 showed 77.7% sequence similarity with its Arabidopsis counterpart AtCDKL9, with a TM-score of 0.96 and an RMSD of 1.33 Å. These results suggest that CDKB1;1, CDKB2;2, and CDKL13 may perform similar functions to their respective homologs in A. thaliana, indicating potential functional conservation.

Protein–protein interaction (PPI) network analyses provided mechanistic insights into the regulatory roles of CDKB1;1 and CDKB2;2 in the cell cycle of M. truncatula. CDKB1;1 was found to interact with multiple cell cycle regulators across different phases (Fig. 5D). It associates with components of the G1/S transition pathway, including the RBR1/RBL1–E2F–CDC6/MCM module, which orchestrates replication origin licensing and S-phase entry [54]. Interactions with Cyclin B, WEE1, and CDC25 suggest involvement in G2/M transition control via CDK phosphorylation status modulation [55]. CDKB1;1 also connects with ORC2, CDC6, and MCM2-like proteins (Fig. 5D), indicating a role in DNA replication initiation, and with CDC16, a subunit of the APC/C complex, pointing to involvement in cell cycle checkpoint regulation and mitotic exit. Similarly, CDKB2;2 interacts with replication initiation factors (ORC2, CDC6, MCM-like, CDC7), reinforcing its role in S-phase progression (Fig. 5F). It also connects with WEE1-like kinases and ATR, implicating it in checkpoint activation and genomic stability maintenance under replication stress. Associations with CDC16 and APC1 further suggest participation in Cyclin degradation and mitotic exit. CDKL13 likely interacts with CycT1-1/2/3/4 and the Pol II complex (RPB3), potentially regulating transcriptional elongation through phosphorylation of the Pol II CTD domain [56]. The interaction between CDKL13 and DDB1 suggests its potential involvement in DNA damage response or oxidative stress adaptation [57]. Furthermore, CDKL13's association with SPT4 may influence RNA splicing processes [56]. CDKL13 appears to participate in cell cycle regulation and transcription-coupled repair by modulating the binding between CycH1-1 and CDKD1;2 (Fig. 5H). These findings collectively indicate that CDKL13 may function as a multifunctional kinase in plants, serving as an integrative hub that connects cell cycle signaling, transcriptional elongation, and RNA metabolism. We also performed Gene Ontology (GO) enrichment analysis for CDKB1;1, CDKB2;2, and CDKL13, along with their interacting proteins. The results showed that CDKB1;1 is broadly involved in multiple cell cycle transition processes (Fig. 5E), while CDKB2;2 is primarily associated with the regulation of mitotic processes (Fig. 5G). In contrast, CDKL13 participates in the positive regulation of DNA-templated transcription elongation, mainly functioning during the G1 and S phases (Fig. 5I), which is consistent with the results of the PPI analysis. In light of these results, we suggest that CDKB1;1 and CDKB2;2 are integral components of the core cell cycle regulatory network in M. truncatula. Their selective downregulation in nodules may serve as a key regulatory mechanism for promoting endoreduplication, which is essential for nodule cell differentiation.

Fig. 5.

Structural and functional conservation of CDKB1;1, CDKB2;2, and CDKL13 protein. A to C Comparative analysis of tertiary structures revealed CDKB1;1, CDKB2;2, and CDKL13 as structural orthologs of AtCDKB1;1, AtCDKB2;2, and AtCDKL9, respectively. A sequence similarity greater than 50% indicates high sequence conservation. The TM-score ranges from 0 to 1, where higher values denote greater structural similarity; a TM-score above 0.5 suggests that the proteins share the same fold type. An RMSD value below 1.5 Å indicates that the two protein tertiary structures are nearly identical. D, F, and H Protein–protein interaction networks of CDKB1;1, CDKB2;2, and CDKL13 were predicted using the STRING database. E, G, and I Gene Ontology (GO) enrichment analysis of CDKB1;1, CDKB2;2, and CDKL13 was performed. False discovery rate (FDR) < 0.05

Ectopic expression of CDKB1;1, CDKB2;2 or CDKL13 in nodule impairs symbiotic nitrogen fixation

To investigate the involvement of CDKB1;1, CDKB2;2, and CDKL13 in the nitrogen-fixing symbiosis within nodules, we first generated composite transgenic plants individually ectopic-overexpressing CDKB1;1, CDKB2;2, or CDKL13 under the control of the nodule-specific DNF1 promoter, using dsRed as a selection marker. qRT-PCR analysis showed that CDKB1;1 expression in transgenic nodules was approximately fivefold higher than in controls, confirming successful overexpression (Fig. 6A). We found that most control nodules were pink indicating strong leghemoglobin expression, while the majority of CDKB1;1-overexpressing nodules appeared white (Fig. 6B). Using the pnifH::GUS reporter in Sinorhizobium meliloti Rm1021 nifH::GUS to assess nitrogen fixation activity of positively transformed nodules, we observed that only 18.35% of overexpression CDKB1;1 nodules showed complete GUS staining, while 29.36% exhibited staining only in infection zones. The GUS reporter driven by the nifH promoter was not expressed in overexpression CDKB1;1 nodules, accounting for 52.29%, representing a 2.3-fold increase compared to controls, demonstrating failed activation of the nitrogenase complex (Fig. 6C). These findings suggest that ectopic overexpression of CDKB1;1 disrupts symbiotic nitrogen fixation.

Fig. 6.

Ectopic expression of CDKB1;1 or CDKB2;2 in nodule impairs nitrogen fixation. A and D Transcript levels of CDKB1;1 or CDKB2;2 were significantly upregulated in their respective transgenic lines (pDNF1::CDKB1;1, and pDNF1::CDKB2;2) relative to controls (untransformed hairy roots), as quantified by qRT-PCR at 21 dpi. Plants were inoculated with Sinorhizobium meliloti Rm1021 nifH::GUS. Expression was normalized to MtPDF2. Results were presented as means ± SD. Statistical differences were determined using Student's T-test. **: significant differences between treatments at p ≤ 0.01. ***: significant differences between treatments at p ≤ 0.001. B and E. Phenotypic comparison of transgenic and control nodules. (B) Phenotype in ectopic expression of CDKB1;1. (E) Phenotype in ectopic expression of CDKB2;2. Top: Nodule color categories (fully pink, partially pink, white). Bottom: nifH::GUS expression patterns (fully blue, partially blue, white). dsRed served as a transformation marker. Control represents negative roots from the same seedlings as the positive roots. Scale bars = 750 μm. C and F. The pie chart results show the percentage of fully blue, partially blue, and fully white nodules between transgenic nodules (C, pDNF1::CDKB1;1; F, pDNF1::CDKB2;2) and the control in the nifH::GUS reporter assay

Similarly, expression of CDKB2;2 increases twofold in the overexpression of CDKB2;2 transgenic nodules (Fig. 6D). Among nodules formed, 26.17% showed uniform GUS staining, 33.64% displayed staining only in the infection and differentiation zones, and 40.19% were completely no GUS staining. Compared to the control, the proportion of pink nodules was reduced by half, while the number of white nodules increased twofold (Fig. 6E-F). These findings suggest that CDKB2;2 overexpression also compromises nitrogen fixation, although the effect is milder than that observed with CDKB1;1 overexpression. However, Overexpression of CDKL13 under the DNF1 promoter led to significantly elevated expression levels in nodules, as confirmed by qRT-PCR (Fig. S7A). Phenotypic analysis revealed only a 12.52% reduction in pink nodules, alongside a 6.32% increase in partially pink and a 8.2% increase in white nodules, relative to controls (Fig. S7B-C), suggesting that CDKL13 likely has slight impact on nodule nitrogen fixation capacity. Taken together, these results demonstrate that CDKB1;1 and CDKB2;2, as classical cell cycle regulators, can directly affect the nodule symbiosis nitrogen fixation.

CDKB1;1 and CDKB2;2 regulate the transition from mitosis to endoreduplication in nodule cells

Endoreduplication is a key event during nodule development, leading to polyploidization of nodule cells. This process establishes the specialized microenvironment required for efficient nitrogen fixation [7]. The initiation of endoreduplication is generally associated with a decline in CDK activity [9, 15]. Therefore, we hypothesize that the nodule-specific downregulation of CDKB1;1 and CDKB2;2 may contribute to the activation of endoreduplication in nodule cells, promoting polyploidy and sustaining nitrogen-fixing capacity in nodules. Ploidy analysis by flow cytometry demonstrated that overexpression of CDKB1;1 nodules contained a significantly higher proportion of 2 C to 8 C cells, but a markedly reduced number of high-ploidy cells (16C–64C) (Fig. 7A and C-D). In CDKB2;2- overexpressing nodule cells, there was no significant difference in the proportion of 4 C cells, the proportions of 2 C and 8 C cells were significantly increased, whereas the frequencies of 16 C to 64 C were markedly reduced compared to controls (Fig. 7B and E-F), indicating that overexpressing CDKB1;1 or CDKB2;2 impaired progression to higher ploidy states, the transition from 8C-to-16C in nodule cells. In agreement with these results, the expression levels of CCS52A, Topo VIA, and APC2, which are known as endoreduplication marker genes [43, 58, 59], were significantly downregulated in transformed nodules (Fig. 7G-I), This finding was further confirmed that ectopic expression of CDKB1;1 or CDKB2;2 disrupts nodule cell endoreduplication.

Fig. 7.

Ectopic expression of CDKB1;1 or CDKB2;2 in nodules disrupts endoreduplication. A and B. Flow cytometry analysis showing reduced ploidy levels in pDNF1::CDKB1;1 or pDNF1::CDKB2;2 nodules compared to respective control. Nuclear DNA content was measured from isolated nuclei. C to F The percentage distribution of cells at different ploidy levels in pDNF1::CDKB1;1 or pDNF1::CDKB2;2 nodules compared to the respective control. (C and E) 2C-4C cells and (D and F) 8C-64C cells. Results were presented as means ± SD. Statistical differences were determined using Student's T-test. *: significant differences between treatments at p ≤ 0.05. **: significant differences between treatments at p ≤ 0.01. ****: significant differences between treatments at p ≤ 0.0001. G to I Downregulation of endoreduplication-related genes CCS52A, Topo VIA, and APC2 in transgenic nodules (pDNF1::CDKB1;1 or pDNF1::CDKB2;2). Expression levels of target genes were normalized to those of MtPDF2. Data represent means ± SD of three biological replicates. Statistical differences were determined using Student's T-test. *: significant differences between treatments at p ≤ 0.05. **: significant differences between treatments at p ≤ 0.01

Meanwhile, we also examined the polyploidization levels of overexpression CDKL13 nodule cells. Notably, no significant changes were observed in 4C–32C populations, while a slight increase in 2 C cells and a moderate reduction in 64 C cells were detected (Fig. S8A-C). The expression of CCS52A was specifically downregulated, whereas transcript levels of Topo VIA and APC2 remained unchanged in CDKL13 overexpression nodules (Fig. S8D-F). These results collectively indicate that tight regulation of CDKB1;1 and CDKB2;2 expression is critical for proper endoreduplication in nodule cells, which in turn supports the high metabolic demands of symbiotic nitrogen fixation.

Discussion

Mitosis and endoreduplication are two closely related yet functionally distinct biological processes essential for plant growth and development [60]. Together, they constitute a regulatory network that coordinates plant cell proliferation and differentiation [1, 9]. The CDK gene family, as a central regulator of the cell cycle, plays a pivotal role in both processes [14]. At present, CDK and CDKL genes have been identified in the model plant A. thaliana as well as in other species such as rice [61], Cotton (Gossypium spp.)[62], Phalaenopsis aphrodite [63], peanut [64], and quinoa [65, 66].

Our comprehensive genome-wide analysis reveals that a total of 29 CDK genes, including 15 canonical CDKs and 14 CDKLs that were classified into eight conserved subgroups (CDKA-CDKG and CDKL). While this classification mirrors that of A. thaliana [64, 66, 67], the absence of the CDKA subgroup is observed in both quinoa [65] and cotton [62]. The finding suggested a potential evolutionary pattern of conservation and possible functional conservation with A. thaliana. Comparative collinear analysis further highlights this evolutionary divergence, with M. truncatula sharing the most collinear CDK gene pairs with G. max and L. japonicus, but significantly fewer with A. thaliana (Fig. 1C). This strongly supports closer orthologous relationships of CDKs among leguminous species [67]. Given that root nodule formation is a distinctive feature of legumes, these conserved CDKs likely play crucial roles in regulating nodule development. Structural analysis of gene architecture and protein motifs supports this functional diversification [68]. Despite overall conservation of core domains, such as the S_TKc domain (Fig. S4), substantial variation in exon/intron organization and motif composition was observed among subgroups [62, 65]. The CDKA subgroup maintains a highly conserved exon structure across M. truncatula and A. thaliana (Fig. 2A and Fig. S3A), whereas the CDKE subgroup in M. truncatula exhibits significantly greater structural complexity, with up to 19 exons, compared to its Arabidopsis counterparts. In contrast, CDKF genes show a simplified gene structure, with only two coding exons (Fig. 2A), a feature consistent with those observed in other legume species like peanut [61, 69], suggesting CDKs exhibit higher conservation among legume species, which may be associated with legume-specific functional diversification or increased regulatory complexity. Motif analysis revealed that all CDKL proteins in both M. truncatula and A. thaliana share ten conserved motifs [65, 69], highlighting exceptional evolutionary conservation within this subgroup (Fig. 2B and Fig. S3B). Several motifs, particularly motif 1, motif 3, motif 8, and motif 10, were frequently absent in canonical CDK proteins(Fig. 2B, Fig. S3B, and Fig. S5-6), especially in the CDKB subgroup, across both species. The absence or variation of specific motifs may correlate with changes in cyclin-binding affinity or substrate specificity [70] and could underlie the subfunctionalization or neofunctionalization of certain CDK genes following duplication events. Together, these findings suggest that CDK family evolution in M. truncatula involves both retention of core regulatory functions and divergence of specific members to meet legume-specific developmental needs.

CDK genes exhibit a conserved spatial expression pattern in M. truncatula. Core cell cycle regulators (CDKA, CDKB, CDKC, and CDKG) display highly conserved tissue-specific expression profiles, predominantly enriched in mitotically active organs such as roots, stems, and leaves (Fig. 3A). This expression pattern has also been validated in other plant species, including peanut and A. thaliana [69, 71, 72], supporting their fundamental role in cell cycle regulation. The selective suppression of CDKB1;1, CDKB2;2, and CDKL13 in nodules, particularly in the infection to fixation zones (Figs. 3B and 4), suggests their specialized roles in mitotic progression and endoreduplication. These genes exhibit upregulated expression in mitotically active tissues, consistent with their putative function in promoting cell division. Conversely, their transcriptional suppression in nodules, where endoreduplication dominates, implies a regulatory mechanism that restricts mitotic CDK activity in favor of polyploidization, a critical adaptation in symbiotic nitrogen-fixing organs. Structural modeling revealed that M. truncatula CDKB1;1, CDKB2;2, and CDKL13 share high tertiary structural similarity and sequence identity with their Arabidopsis orthologs AtCDKB1;1, AtCDKB2;2, and AtCDKL9 (Fig. 5A-C), indicating strong evolutionary conservation. Despite this, functional divergence is evident: protein–protein interaction analyses and GO enrichment results show that CDKB1;1 and CDKB2;2 function as core regulators of cell cycle phase transitions (Fig. 5D-G), engaging in DNA replication initiation, G2/M progression, and mitotic exit, whereas CDKL13 is primarily involved in the positive regulation of transcriptional elongation from DNA templates, likely during the G1/S phases (Fig. 5H-I). This distinction suggests that CDKB1;1 and CDKB2;2 act as direct modulators of cell cycle progression, while CDKL13 may influence cell cycle dynamics indirectly through transcriptional regulation (Fig. 5I). These findings underscore the differential specialization of CDK and CDKL subfamilies in legumes and support a model in which the coordinated suppression of mitotic CDKs facilitates endoreduplication-driven differentiation, representing a crucial adaptation for functional nodule development and symbiotic nitrogen fixation.

As a mitotic driver, Ectopic overexpression of CDKB1;1 or CDKB2;2 may lead to persistent activation of mitosis, thereby disrupting the onset of endoreduplication in nodules (Fig. 7). Ectopic overexpression of CDKB1;1 or CDKB2;2 significantly increased the proportion of white nodules lacking nitrogen-fixing capacity (Fig. 6). This phenotype was accompanied by a sharp decline in high-ploidy nuclei (16C–64C) and an accumulation of lower ploidy levels (2C–8C) (Fig. 7C-F), indicating that the nodule-specific suppression of CDK activity is critical for regulating the transition from 8 C to higher ploidy states during nodule development. Collectively, these results highlight that CDKB1;1 and CDKB2;2 are key regulators of endoreduplication in nodule cells, which is essential for the establishment of effective nitrogen-fixing symbiosis. Acting as a pivotal regulator of endoreduplication entry, CCS52A is required for the proper initiation of endoreduplication; its deficiency disrupts the entire process[25–27, 43]. Intriguingly, the phenotype of CDKB1; 1 or CDKB2;2 overexpression closely resembles that of CCS52A silencing mutants, which also exhibit defective nodule development, loss of nitrogen fixation ability, and markedly reduced symbiotic cell ploidy, such as 16 C to 64 C [3, 43]. We conclude that the nitrogen-fixing capacity of nodule cells is tightly coupled with cellular ploidy levels with the critical endoreduplication transition point occurring at the 8C-to-16C switch during nodule cell development. Disruption of endoreduplication directly compromises symbiotic nitrogen fixation in nodules. In contrast, CDKL13 overexpression had only a minor impact on nodule nitrogen-fixing function and did not significantly alter nuclear ploidy distribution (Fig. S7-S8). PPI and GO enrichment analyses further indicate that CDKL13 may function at the nexus of transcription regulation, RNA processing, and cell cycle coordination (Fig. 5H-I). Current evidence suggests that animal CDKL5 may regulate RNA splicing indirectly through phosphorylation of splicing factors [73]. Furthermore, we found that CDKL13 is not involved in either early nodule development or mitotic processes in the apical meristematic zone of mature nodules (Fig. 3B). Thus, our results further demonstrate that most CDKL genes in M. truncatula nodules likely do not directly regulate cell cycle progression, and CDKB1;1 and CDKB2;2 emerge as core cell cycle regulators in nodule development, providing new mechanistic insights into legume nitrogen-fixing symbiosis.

Conclusions

This study comprehensively analyzed the CDK and CDKL gene families in M. truncatula, elucidating their phylogenetic relationships, chromosomal distribution, gene structure, conserved motifs, protein domains, expression patterns, and functional roles in nodule development and symbiotic nitrogen fixation. The findings revealed that CDK genes in M. truncatula are distributed across eight chromosomes, with distinct evolutionary patterns including gene number increase and contraction in different subfamilies. The expression analysis showed that specific CDK genes, such as CDKB1;1, CDKB2;2, and CDKL13, are highly expressed in roots but significantly downregulated in nodules, particularly in the infection-to-fixation zones. Ectopic overexpression of CDKB1;1 or CDKB2;2 in nodules disrupts nitrogen fixation and impairs endoreduplication, especially CDKB1;1, highlighting that suppression of mitotic CDK activity is a prerequisite for symbiotic polyploidization. These findings emphasize the critical role of CDKs in coordinating the transition from mitosis to endoreduplication during nodule development and offer mechanistic insights into the regulation of symbiotic nitrogen fixation in legumes.

Abbreviations

CDK

Cyclin-dependent kinase

CDKL

Cyclin-dependent kinase-like

PPI

Protein–protein interaction

GO

Gene ontology

CYC

Cyclin

RBR

Retinoblastoma-related protein

64C

64 C-value

UTR(U)

Untranslated region

CDS(C)

Coding sequence

E

Exons

I

Introns

GUS

β-Glucuronidase

Mt

M. truncatula

AT

A. thaliana

Gm

G. max

Lj

L. japonicas

L

Leaves

IA

Inflorescence apices

VA

Vegetative apices

E

Embryo

SC

Seed coat

MS

Mature seed

S

Shoot

R

Root

N

Nodule

ZI

Meristem zone

ZIId

Distal infection zone

ZIIp

Proximal infection zone

ZI

Infection Zone

ZIII

Fixation zone

Hpi

Hour post-inoculation

Dpi

Days post-inoculation

Wpi

Week post-inoculation

MO

Mock inoculation

Sm

The plants inoculation with Sinorhizobium meliloti 2011

FDR

False discovery rate

qRT-PCR

Quantitative Real-Time PCR

TM

Template modeling

RMSD

Root mean square deviation

Authors’ contributions

RW and HP designed the research. RW performed most of the experiments. RW analyzed the experimental results. RW wrote the manuscript.

Funding

Not applicable.

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.