Identification and Expression Analysis of MADS-Box Gene Family in Pinus koraiensis and Overexpression of PkMADS9 Promoting Early Flowering in Transgenic Arabidopsis

Abstract

Korean pine (Pinus koraiensis) is a vital woody oil tree species native to Northeast Asia, with its pine nuts serving as the primary global source of edible pine nuts globally due to their rich nutritional content. Currently, seed yield from Korean pine is low and unstable, failing to meet the market demand. The limited number of female cones is the primary factor restricting its yield. MADS-box family members are crucial in regulating the initiation, differentiation, and morphogenesis of floral organs. However, systematic identification and characterization of MADS-box proteins in Korean pine have not been reported. This study utilized transcriptome data from reproductive and vegetative buds during the flower bud differentiation stage of Korean pine to comprehensively identify MADS-box family members through bioinformatics analysis and molecular biology approaches. A total of 37 PkMADS-box genes were identified, including 6 type I and 31 type II (MIKC) genes, which were classified into 8 subfamilies. The physicochemical properties, conserved domains, conserved motifs, protein structures, gene expression profiles, and protein–protein interaction networks of these genes were analyzed. Key genes associated with physiological differentiation (flower induction) and sexual organogenesis were identified based on expression patterns during flower bud differentiation and flower organ development. Among these, PkMADS4 and PkMADS26 are likely involved in positively regulating flower induction, while PkMADS9 plays a role in the morphological differentiation of sexual organs in a dose-dependent manner and overexpression of PkMADS9 promoting early flowering in transgenic Arabidopsis. These genes were also identified as key candidates for regulating reproductive phase changes and strobilus development. This study provides a theoretical foundation for further investigation of MADS-box genes in reproduction and offers insights into genetic improvements aimed at enhancing the seed yield of Korean pine.

1. Introduction

MADS-box family transcription factors are widely distributed in eukaryotes and play a critical role throughout the plant life cycle, including responses to abiotic stress, floral organ identity, regulation of flowering time, and fruit and seed development [1,2,3]. The name “MADS” is derived from the initials of four genes: Minichromosome Maintenance 1 (MCM1) from yeast, AGAMOUS (AG) from Arabidopsis, DEFICIENS (DEF) from snapdragon, and the SerumResponse Factor (SRF) from Homo sapiens [4,5]. All MADS-box family members share a highly conserved MADS domain, typically comprising 50–60 amino acids, which is responsible for specific DNA binding (targeting CArG-box motifs) and dimerization, essential for regulating plant growth and development [6,7]. Plant MADS-box genes are classified into two major categories based on phylogeny and structure: type I (including Mα, Mβ, and Mγ groups) and type II (also known as MIKC-type) [8]. These types are distinguished by the presence or absence of the Keratin-like (K) domain [9,10]. Type I genes encode an MADS (M) domain and a variable C-terminal (C) domain [11], while Type II genes also encode M domain and C domain, additionally contain an intervening (I) domain and a keratin-like (K) domain, and are divided into MIKC^C and MIKC* groups [12].

Most MIKC members of the MADS-box family are key floral organ determining genes, playing essential roles in floral primordium differentiation, flower development, and flowering time control [13,14]. MIKC* members of plant MADS-box gene family are typically involved in the development of plant male and female gametophytes [15,16], while MIKC^C members are central to various stages of flower development and represent the most extensively studied functional group [17,18], they are further classified into 12 subgroups, including SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1), APETALA1/FRUITFULL (AP1/FUL), SHORT VEGETATIVE PHASE (SVP), APETALA3/PISTILLATA (AP3/PI), FLOWERING LOCUS C (FLC), AGAMOUS-LIKE6 (AGL6), AGAMOUS/SEEDSTICK (AG/STK), SEPALLATA (SEP), BSISTER (BS), AGL12, AGL15, and AGL17 [3,19,20]. Extensive research in angiosperm models has led to the development of the classic ABCDE model, providing a comprehensive framework for understanding floral organogenesis and development [21,22]. In Arabidopsis thaliana, the corresponding functional genes include the Class A gene AP1, which primarily controls the development of sepals and petals; Class B genes AP3 and PI, which regulate the development of petals and stamens; Class C gene AG, which governs the development of stamens and carpels; Class D gene STK, which influences ovule development; and Class E gene SEP1-4, which are involved in the formation and development of floral organs. SEP1-4 interact with A, B, C, and D class proteins to form higher-order complexes (floral quartets) essential for the specification of all floral organs [23,24]. Different combinations of the A, B, C, D, and E MADS-box gene function determine the identities of distinct floral organs. According to this model, A + E control sepals, A + B + E control petals, C + E control carpels, and D + E control ovules [25].

Korean pine (Pinus koraiensis Sieb. et Zucc.) is a vital ecological protection species and a woody oil tree species, playing significant roles in timber production, nut utilization, and ecological restoration [26]. It is widely distributed across Northeast China, the Russian Far East, the Korean Peninsula, and Japan [27,28]. Korean pine nuts are the primary global source of edible pine nuts, valued for their rich nutritional and pharmacological bioactive compounds [29,30]. However, the current seed yield of Korean pine in forest production is generally low, failing to meet the growing market demand. This is attributed to several factors, including apical fruiting, long reproductive cycles, shifting breeding objectives, and the lack of techniques to promote flowering [31,32,33,34]. Among these factors, the number of female cones (ovulate strobili) is a key limiting factor in seed yield. Furthermore, research on the regulatory mechanisms of reproductive development in Korean pine has progressed slowly due to the absence of essential research tools such as a reference genome, tissue culture systems, and genetic transformation platforms [35,36,37,38,39]. Although molecular breeding approaches to increase the number of female cones remain a complex challenge, identifying key genes associated with sex determination, floral induction, and floral organ development will provide valuable gene resources for molecular breeding in Korean pine. This strategy is critical for accelerating genetic improvement of Korean pine and achieving high and stable seed yields

To investigate the functional roles of PkMADS-box genes and identify potential regulators of floral organogenesis and development in Korean pine, this study comprehensively identified PkMADS-box family members using transcriptomic data from differentiating reproductive and vegetative buds. Bioinformatic analyses were conducted to determine their phylogenetic relationships, physicochemical properties, conserved domains, motifs, and protein structures. The expression profiling was performed to uncover dynamic gene expression patterns during flower bud differentiation, floral development, and across trees of varying ages. Additionally, the biological function of key gene PkMADS9 was characterized through heterologous transformation in Arabidopsis. This study lays the theoretical foundation and provides essential gene resources for molecular breeding aimed at enhancing cone yield in Korean pine.

2. Results

2.1. Identification and Characterization of MADS-Box Gene Family in Korean Pine

To identify all members PkMADS-box family genes of P. koraiensis, two strategies of the local BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 5 January 2025) search using TBtools software with protein sequences from P. tabuliformis, P. abies, G. gnemon, and A. thaliana, and the HMMER search using HMMER 3.0 software with MADS-box HMMs (PF00319 and PF01486) were conducted. A total of 37 MADS-box family genes were identified and designated as PkMADS1 to PkMADS37 by integrating local BLAST and HMMER search results, removing redundant sequences, confirming the MADS-box conserved domain (Table S1). The physicochemical properties of these 37 PkMADS-box proteins were analyzed. The coding sequence (CDS) lengths ranged from 312 bp (PkMADS37) to 1143 bp (PkMADS34), encoding polypeptides of 103 to 380 amino acids. Except for PkMADS33, PkMADS35, PkMADS36, and PkMADS37, which encoded about 110 amino acids, the remaining 33 members encoded more than 150 amino acids, with most having approximately 200 amino acids (Table S1).

The molecular weight of the PkMADS-box family proteins varied significantly, ranging from 11.39 kDa (PkMADS37) to 43.79 kDa (PkMADS34), with the majority of proteins having a molecular weight of approximately 25 kDa. Similarly, the theoretical isoelectric points of the 37 proteins also varied, ranging from 5.58 (PkMADS26) to 10.09 (PkMADS33), though most members had isoelectric points greater than 7, suggesting a high abundance of basic amino acids in these proteins. Analysis of the protein instability index revealed that except for PkMADS20, PkMADS27, PkMADS35, and PkMADS37, which had values below 40 and were classified as stable proteins, the remaining 33 proteins had instability index values exceeding 40, indicating that most PkMADS-box proteins are unstable. Furthermore, the grand average of hydropathicity (GRAVY) for all members was negative, indicating that PkMADS-box proteins are highly hydrophilic. Subcellular localization prediction showed that 37 PkMADS-box proteins were localized in the nucleus (Table S1).

2.2. Conserved Domain and Conserved Motif Analysis of PkMADS-Box Members

The conserved domain analysis of the 37 PkMADS-box members revealed distinct structural patterns (Figure 1A). Based on the results of 37 PkMADS-box members conserved domain analysis, there were 6 members (PkMADS32 to PkMADS37) contained only one MADS-box domain at the N-terminus, belonging to the Type I lineage of the MADS-box family. In contrast, the remaining 31 members (PkMADS1 to PkMADS31) contained both a conserved N-terminal MADS domain and a C-terminal K-box domain, categorizing them as the Type II (MIKC-type) lineage of the MADS-box family.

Conserved motif analysis indicated that all PkMADS-box family members contained Motif 5 and Motif 1. In addition to PkMADS35 and PkMADS37, the remaining 35 members also featured Motif 4, suggesting that Motifs 5, 1, and 4 are highly conserved across the family (Figure 1B). The distribution of these motifs at the N-terminus implies their potential association with the MADS domain. Among the 31 Type II members, all contained Motif 2, indicating its relevance to the K-box domain. Motifs 3 and 7 were found in most members, which may be linked to the I domain and C domain, respectively. Specific motifs were also identified in subsets of Type II members: Motif 9 was exclusive to PkMADS3, PkMADS4, and PkMADS25; Motif 10 was specific to PkMADS9–PkMADS11; Motif 8 was found only in PkMADS19, PkMADS23, and PkMADS24; and Motif 6 was restricted to PkMADS28–PkMADS31. The specific distribution of these motifs suggests they may contribute to the functional diversity of these proteins. Notably, PkMADS4 contains both Motif 9 and Motif 10. In contrast, PkMADS6, PkMADS12, and PkMADS14 lack Motif 3 and the motifs mentioned above, suggesting that these members may perform unique biological functions.

2.3. Phylogenetic, Classification and Conserved Motifs of PkMADS-Box Genes

To classify the MADS-box proteins of Korean pine, the ML phylogenetic trees were constructed based on sequence alignments of 37 P. koraiensis proteins, 69 A. thaliana proteins, 9 P. tabuliformis proteins, 5 P. abies proteins, and 1 G. gnemon protein. Within the Type I lineage (Figure 2), five members—PkMADS32, PkMADS33, PkMADS35, PkMADS36, and PkMADS37—were clustered with Arabidopsis Mα subfamily proteins, suggesting their affiliation with the Mα subfamily. PkMADS34 was grouped with Arabidopsis Mγ subfamily proteins, indicating its classification into the Mγ subfamily. Within the Type II lineage (Figure 3), the analysis further showed that the 31 MIKC members were primarily divided into six subfamilies: AG, AGL6, AP3/PI, SVP, SOC1, and GGM7, all of which are MIKC^C-type genes. Among these, four members—PkMADS6, PkMADS9, PkMADS10, and PkMADS11—were classified into the AP3/PI-like (Class B) subfamily. PkMADS1 was classified into the AG subfamily (Class C). Four members—PkMADS2, PkMADS3, PkMADS4, and PkMADS5—belonged to the AGL6 subfamily. Another four members—PkMADS12, PkMADS13, PkMADS14, and PkMADS15—were classified into the SVP subfamily. Two members, PkMADS7 and PkMADS8, were categorized into the GGM7 subfamily. The SOC1-like group was the largest, comprising 16 members (PkMADS16-PkMADS31). Notably, no members were classified into the AP1/FUL (Class A), SEP (Class E), FLC, or MIKC* subfamilies in the available Korean pine data.

2.4. Analysis of the Protein Secondary and Tertiary Structures of the PkMADS-Box Members

To further explore the structural diversity of the encoded proteins, secondary and tertiary structure analyses were conducted for the 37 PkMADS-box proteins (Table S2). Despite significant variation in amino acid sequence lengths, the secondary structure composition was relatively conserved, consisting of α-helices, extended strands, β-turns, and random coils. Among these, α-helices were the most abundant (32.63% to 75.00%), followed by random coils (11.11% to 58.16%) and extended strands (6.84% to 34.95%), while β-turns represented the smallest proportion (1.68% to 9.09%). In 36 of the 37 members, the proportion of α-helices exceeded that of random coils, with PkMADS34 being the only exception. Similarly, random coils were more prevalent than extended strands in all members except PkMADS37. The proportion of β-turns in most members ranged from 2.19% to 3.72%. The positional distribution of these structural elements is visualized in Figure 4A. Generally, α-helices are predominantly located at the C-terminus, while extended strands and β-turns are concentrated at the N-terminus. Random coils are mainly distributed in the central region and at the C-terminus.

The predicted tertiary structures of the PkMADS-box family proteins are presented in Figure 4B. Structural elements such as α-helices, β-turns, and random coils are clearly observable in the 3D models, consistent with the secondary structure predictions. Despite significant variation in overall spatial configurations due to differences in the length and arrangement of helices, sheets, and coils, high structural similarity was observed between specific pairs or groups. For example, PkMADS3 and PkMADS4, PkMADS13 and PkMADS15, and the PkMADS23–27 subgroup exhibited highly conserved 3D architectures. Notably, PkMADS10 and PkMADS11 shared a similar structure characterized by a high abundance of random coils.

2.5. Analysis of the Expression Pattern of the PkMADS-Box Genes During the Flower Bud Differentiation Stage

To investigate the functional roles of PkMADS-box genes during flower bud differentiation, gene expression profiling of the 37 identified PkMADS-box genes was performed based on transcriptomics data from the flower bud differentiation stage. The results revealed distinct spatiotemporal expression patterns (Figure 5).

In TB, high expression levels were observed for one AGL6-like gene (PkMADS2), one GGM7-like gene (PkMADS8), ten SOC1-like genes (PkMADS18, 20, 22–29), one Mγ gene (PkMADS34), and one Mα gene (PkMADS35). Specifically, PkMADS2 and PkMADS18 maintained high expression from the vegetative stage (S1) through the late differentiation stage (S3). PkMADS20, PkMADS22, PkMADS29, and PkMADS35 peaked during the mid-physiological differentiation stage (S2), while PkMADS23 to PkMADS27 showed high expression in the end-physiological differentiation stage (S3), suggesting they act as positive regulators of floral induction. In contrast, PkMADS8 exhibited peak expression at S1, significant downregulation at S2, and upregulation at S3, indicating a potential negative regulatory role. Notably, PkMADS2, PkMADS18, PkMADS25, and PkMADS26 were preferentially expressed in TB, identifying them as potential marker genes for TB differentiation. Moreover, PkMADS28 and PkMADS21 showed similar expression patterns in both TB and VB during S1 and S2, but their expression diverged significantly at S3. Combined with anatomical observations, this suggests that PkMADS28 and PkMADS21 are likely key characteristic genes for Os primordium differentiation.

In LTB, genes exhibiting elevated expression included AG-like (PkMADS1), AGL6-like (PkMADS5), GGM7-like (PkMADS7), AP3-like (PkMADS6, PkMADS9), SOC1-like (PkMADS17, PkMADS31), SVP-like (PkMADS12–14), and Mα (PkMADS33, PkMADS36) members. Among these, PkMADS12, PkMADS17, and PkMADS36 showed high expression at both S1 and S3, but significant downregulation at S2, suggesting a potential negative regulatory role during the critical physiological differentiation phase. In contrast, PkMADS4 and PkMADS13 were markedly upregulated only at S2 and downregulated at S3, with minimal expression in VB, indicating their function as positive regulators in floral induction with limited involvement in later stages of sexual organ development. Additionally, PkMADS7 and PkMADS9 exhibited an increasing expression trend from S1 to S3, peaking significantly at S3. While similar expression patterns were observed in the apical meristem and VB, especially lower expression in TB, these genes may influence the morphological differentiation of sexual organs through a dosage-dependent mechanism, based on the differentiation characteristics of this period. Notably, PkMADS1 (Class C) and PkMADS6 (Class B) were highly expressed exclusively in LTB at S3, suggesting a close association with lateral bud differentiation and LTB development, likely marking key genes for sex determination in the Ms.

In VB, fewer genes exhibited distinct expression patterns. PkMADS15 and PkMADS28–30 were significantly upregulated at S2, while PkMADS5, 7, 9, 10, 14, and 31 showed increased expression at S3, indicating their involvement in VB differentiation and needle formation. Among them, PkMADS10 exhibited a steady increase in expression across all three bud types from S1 to S3, with significantly higher levels in VB at S3 compared to reproductive buds, suggesting its role as a characteristic gene for VB differentiation and development. Finally, to validate the accuracy of the transcriptomic data and the expression patterns of PkMADS-box genes across the three bud types, qRT-PCR was performed using randomly selected PkMADS-box genes. As shown in Figure 6, the expression profiles obtained from qRT-PCR were highly consistent with the RNA-seq abundance data, confirming the reliability of the transcriptomic dataset and the specific expression patterns identified for the PkMADS-box family members.

2.6. Expression Pattern Analysis of PkMADS-Box Genes in Different Tree Ages and Floral Organ Development Stages

To explore the potential roles of PkMADS-box genes in the reproductive phase transition and floral organogenesis of Korean pine, samples of TB, VB, LTB, and TN from six different tree ages, as well as Os and Ms samples during the floral organ development stage from 30-year-old trees, were collected. The tissue-specific expression profiles of three candidate genes, PkMADS4, PkMADS9, and PkMADS26, were analyzed using qRT-PCR. In all three bud types, the expression levels of PkMADS4, PkMADS9, and PkMADS26 increased with tree age, peaking in 30-year-old (Y6) trees, which had transitioned into the reproductive stage (Figure 7). In needles, PkMADS4 exhibited a similar age-related expression pattern to that seen in the buds, while PkMADS9 and PkMADS26 also showed an upward trend with age, although their patterns were less pronounced. This age-dependent expression suggests that these three genes may play a critical role in the transition from vegetative to reproductive growth in Korean pine.

During floral organ development, PkMADS4, PkMADS9, and PkMADS26 displayed distinct temporal and spatial expression patterns, reflecting their functional specialization. PkMADS4 maintained a constitutive and stable expression profile across all developmental stages of both Os and Ms, suggesting its fundamental role in floral organ development. In contrast, the relative expression of PkMADS9 in both Os and Ms peaked at the stage when the bud morphology became distinguishable, then rapidly decreased, indicating its major role in the floral organ initiation stage. Notably, PkMADS26 showed extreme tissue specificity in Os, with expression levels rising sharply during floral organ development and pollination, while its expression in Ms was nearly undetectable, suggesting that PkMADS26 is a key regulator of Os development and the pollination process.

2.7. Overexpression of PkMADS9 Promotes Flowering in Transgenic Arabidopsis

To elucidate the biological function of PkMADS9 in P.s koraiensis, the CDS of PkMADS9 was transformed into Arabidopsis thaliana via the floral dip method, generating 10 independent transgenic lines (Figure S1). Three overexpressing lines (OE#1, OE#7, and OE#10) exhibiting relatively high expression levels were selected for subsequent phenotypic characterization. The results demonstrated that both bolting and flowering times of the PkMADS9 transgenic lines were significantly earlier than those of the WT (Figure 8A). The number of rosette leaves at flowering was significantly reduced in the PkMADS9 transgenic lines compared to the WT (Figure 8B). Furthermore, qRT-PCR revealed that the expression levels of positive flowering regulators (AtLFY, AtFT, and AtGI) and key floral organ development genes (AtAP1, AtAP3, AtPI, AtAG, and AtSEP3) were significantly upregulated in the transgenic lines (Figure 8C). These findings suggest that PkMADS9 overexpression promotes an early-flowering phenotype in Arabidopsis.

3. Discussion

The MADS-box gene family, particularly the MIKC-type, plays pivotal roles in floral primordium initiation, floral organ development, flowering time regulation, and ovule development in plants. With the rapid progress of high-throughput sequencing technologies, genome-wide characterization of the MADS-box family has been extensively carrie out across a wide range of angiosperms [40], including monocots like Oryza sativa [41], Triticum aestivum [19], Zea mays [42], Musa spp. [43], and Phyllostachys edulis [44] as well as dicots such as Arabidopsis [45,46], Glycine max [47], Malus domestica [48], Ipomoea batatas [49], Juglans mandshurica [50], Lonicera japonica [51], Rosa chinensis [52], and Cerasus pseudocerasus [53]. However, despite these advancements, a systematic identification and characterization of the MADS-box gene family in Korean pine remains unreported.

3.1. Identification Constraint and Structure Features of the PkMADS-Box Family Member

In comparison to angiosperms, gymnosperm genomes are typically vast (often exceeding 20 Gb), complex (containing high levels of repetitive sequences), and challenging to assemble, which has delayed their sequencing and molecular study relative to that of angiosperms. To date, only a few species have been successfully whole-genome sequenced and published, such as P. abies [54], Pinus taeda [55], Ginkgo biloba [56], G. montanum [57], Cycas panzhihuaensis [58], and P. tabuliformis [59], but a high-quality reference genome for Korean pine is currently unavailable. Based on previous research results on the reproductive biology of Korean pine, the key physiological differentiation period of strobili has been identified as spanning from early July to mid-August [30]. During this period, three types of buds were specifically selected for transcriptome sequencing, leading to the successful identification of 37 PkMADS-box gene family members in Korean pine from the obtained high-quality transcriptome data. In the present study, conserved domain identification and multiple sequence alignment revealed over ten additional candidate sequences with high homology to MADS-box genes but lacking complete MADS domains. Although these candidate sequences were excluded from further detailed analysis to ensure the accuracy and reliability of the identification result, their presence suggests that the actual size of the PkMADS-box family likely exceeds the reported 37 members. We clearly acknowledge that relying solely on transcriptomic data presents inherent limitations for gene family identification. Unlike whole-genome sequencing, transcriptomic data is limited to expressed genes and may fail to capture members that exhibit extremely low expression levels, highly tissue-specific patterns not covered by our sampling (such as roots or mature seeds), or expression is restricted to specific developmental stages or environmental stimuli. Consequently, the 37 PkMADS-box genes identified in this study likely represent the core active members involved in the flower differentiation of P. koraiensis, but not the complete members. The future availability of a high-quality P. koraiensis reference genome will be essential to provide a comprehensive information of this gene family and to facilitate the identification of these additional members, thereby accelerating research into the molecular mechanisms of reproductive development of P. koraiensis.

The structural characteristics of proteins are closely linked to their biological functions and regulatory mechanisms. For transcription factors, their specific spatial conformation not only dictates their binding affinity to target DNA but also influences the specificity of protein–protein interactions [60]. In the present study, subcellular localization predictions indicated that all 37 PkMADS-box family members are localized to the nucleus, aligning with their typical role as transcription factors that regulate gene expression within the nucleus [15]. Notably, secondary structure analysis revealed that PkMADS-box proteins are enriched with a high proportion of random coil regions (Table S3). These structural features are consistent with the characteristics of intrinsically disordered regions (IDRs), which are frequently observed in eukaryotic transcription factors and signaling proteins [61]. These disordered structures are generally hypothesized to confer a high degree of conformational plasticity, which may theoretically enable PkMADS-box proteins to interact with diverse partners, such as other MADS-box proteins or DNA motifs, thereby facilitating the formation of multimeric complexes required for their regulatory functions. Furthermore, although specific IDRs prediction remains to be conducted, the abundance of random coils suggests a potential for structural flexibility. In evolutionary contexts, such flexibility is often associated with the capacity to accommodate novel post-translational modification sites or facilitate rapid functional divergence [62], which may contribute to the adaptability of the PkMADS-box gene family in P. koraiensis.

3.2. Evolutionary Divergence of the PkMADS-Box Family Member

The number of MADS-box gene family members varies markedly among species, particularly in gymnosperms, which generally harbor fewer MADS-box genes than angiosperms due to the absence of frequent whole-genome duplication (WGD) events characteristic of angiosperms [63]. In contrast, angiosperms have undergone extensive expansion of the MADS-box family following WGD events, as seen in woody fruit trees such as pear (Pyrus bretschneideri, 95 PbMADS-box genes) [64], litchi (Litchi chinensis, 101 LcMADS-box genes) [65], Mango (Mangifera indica, 119 MiMADS-box gene) [66], and apple (Malus domestica, 146 MdMADS-box gene) [48]; in ornamental plants such as rose (R. chinensis 58 RcMADS-box gene) [52]; and in crops including rice (O. sativa, 75 OsMADS-box genes) [41], wheat (T. aestivum, 201 TaMADS-box genes) [19], and maize (Z. mays, 211 ZmMADS-box genes) [42]. In contrast, genome-wide studies in gymnosperms have identified substantially fewer members, including 22 GmMADS-box genes in G. montanum [57], 26 GbMADS-box genes in G. biloba [67], 47 ClMADS-box genes in Cunninghamia lanceolata [68], and 68 PtMADS-box genes in P. tabuliformis [69]. Although 278 MADS-box gene models were predicted in P. abies, many are likely pseudogenes derived from transposon activity [63].

The phylogenetic analysis in this study revealed that P. koraiensis lacks orthologs of the AP1/FUL (Class A) and SEP (Class E) subfamilies (Figure 2). This absence is consistent with genomic studies in P. abies [54] and P. tabuliformis [59], which support the hypothesis that the strict ABCE model is an angiosperm innovation. Instead, gymnosperms likely utilize a more ancestral B/C model, where organ identity might be specified by B and C class genes without the requirement for distinct A and E class genes [63,70,71]. Notably, although there are almost no WGD events in conifers, the SOC1-like subfamily in P. koraiensis has significantly expanded to 16 members. This finding is consistent with observations in P. tabuliformis, where the diversification of SOC1 homologs was driven by tandem duplication and transposable element activity [69,72]. The transition from vegetative to reproductive growth requires the integration of complex environmental cues (such as photoperiod and temperature) and internal age-related signals, with SOC1 serving as a central floral pathway integrator [73,74]. We speculate that this lineage-specific expansion of the SOC1-like subfamily in pine may serve as a compensation mechanism to compensate for the loss of WGD and promote the gene subfunctionalization. The differential expression patterns observed in this study strongly support this hypothesis: for instance, SOC1 homologous gene PkMADS26 showed strict tissue specificity expression in TB (Figure 5), while other members showed constitutive expression or age-dependent upregulation. This functional diversification may enable P. koraiensis to establish a fine regulatory network and coordinate its prolonged reproductive cycle in the absence of large-scale gene redundancy common in angiosperms [75,76].

3.3. The Expression Patterns and Potential Functions of Key PkMADS-Box Genes in Reproductive Development

The majority of genes involved in plant reproductive development belong to the MADS-box family. Within the classical ABCDE model of plant flowering, both structural genes—such as class A (AP1/FUL), class B (AP3/PI), class C/D (AG/STK), and class E (SEP)—and upstream regulators of flowering time, such as SVP and SOC1, are core members of this family, with their functions well-established in angiosperms [22,77]. Gymnosperms possess numerous orthologs of the angiosperm class B genes. Given the absence of petal structures in gymnosperms, current research indicates that the function of these class B homologs is primarily concentrated on the development of male reproductive organs, particularly in the specification of male cone identity determination [78]. For example, in P. abies, G. montanum, and Cryptomeria japonica, class B genes show significantly specific expression in male cones, with spatial and temporal expression patterns aligning closely with those in angiosperms, suggesting a conserved mechanism of reproductive organ development between the two lineages. Thus, similar to the class B genes in the angiosperm ABCDE model, gymnosperm B-class homologs play a critical role in male organ determination [79]. Among the four identified AP3/PI-like (class B) genes in this study, PkMADS6 and PkMADS9 were highly expressed in the LTB, which predominantly differentiates into male cones. Notably, PkMADS9 exhibited a highly specific expression pattern in the LTB, which is consistent with the role of class B genes in determining male identity in gymnosperms. Meanwhile, PkMADS9 was high homology with PaDAL11, PaDAL12, and PaDAL13 of P. abies. Although previous studies have characterized PaDAL11 to 13 as B-class genes specifically expressed in developing pollen-bearing organs [80], their potential roles in flower bud differentiation and the regulation of flowering time remain largely unexplored. In contrast, PkMADS11 was highly expressed in TB, which primarily differentiates into female cones, suggesting potential sub-functionalization of class B genes in Korean pine to finely regulate the development of distinct sexes.

In gymnosperms, unlike the carpel-specific role in angiosperms, Class C genes are typically expressed in both male and female cones to define reproductive identity against vegetative tissues [81,82]. Class C genes PkMADS1 (AG-like) were highly expressed only in LTB at the S3 stage, suggesting that the transcriptional activation of PkMADS1 occurred earlier in male buds, rapidly initiating the reproductive program during early morphological differentiation. In contrast, the peak expression of PkMADS1 in female buds (TB) may occur at a later stage of morphological development, beyond the S3 time point sampled in this study. This potential temporal divergence in PkMADS1 activation between sexes warrants further investigation in future studies. In angiosperms, AGL6 acts redundantly with SEP genes in floral organ specification, while in conifers, AGL6 homologs are known to act as floral integrators, involved in floral organ identity and evolved a unique function in phase transition [75,83]. Crucially, the AGL6 homolog PtDAL1 from P. tabuliformis has been characterized as a central age-dependent regulator and plays a major regulatory role in the transition from vegetative to reproductive growth [75,76]. Overexpression of PtDAL1 promotes early flowering in Arabidopsis and regulates the reproductive process in P. tabuliformis by modulating the expression of PtLFY and PtNLY via a regulatory cascade [76]. In this study, AGL6-like homolog PkMADS4 exhibited high homology to PtDAL1 of P. tabuliformis (Figure 3), and the expression of PkMADS4 was increased significantly with tree age (Figure 7) and upregulated during the S2 stage (Figure 5). This suggests that PkMADS4 might serve as an age clock and a floral activator in P. koraiensis, functionally conserved with PtDAL1. Furthermore, multiple SOC1-like homologs (PkMADS18, 21, 25, 26) were preferentially expressed in TB. Although SOC1 homologs in Arabidopsis typically function as floral integrators and organ identity regulators in angiosperms, SOC1 homologs in gymnosperm show functional divergence due to gene family expansion. The P. tabuliformis genome study revealed that SOC1-like genes have expanded via tandem duplications, facilitating the evolution of distinct spatiotemporal expression patterns [59]. In P. abies, certain SOC1 homologs are recruited to regulate bud dormancy and vegetative growth rhythms [84], whereas in C. japonica, specific SOC1-like genes CjMADS15 are involved in reproductive development [85]. The specific expression of PkMADS26 in this study suggests that it may have evolved a specialized role in the initiation and maintenance of ovulate strobilus, which needs to be further verified in future research.

Notably, in addition to spatial organ specificity, the dynamic temporal changes in MADS-box genes play a critical role in floral induction. During the key transition from vegetative to reproductive growth (S2 or S3), the AGL6-like gene PkMADS5 and SOC1-like genes (PkMADS16, 28, 30, 31) were significantly up-regulated across all bud types. This general up-regulation, independent of organ specificity, strongly suggests that these genes serve as key switches in the floral induction process of Korean pine, responding to environmental or developmental signals to initiate reproductive development. Similar patterns have been observed in related species such as P. tabuliformis and C. lanceolata [77]. Furthermore, gene sequence conservation does not always equate to functional identity. The complex evolutionary history of gymnosperms may lead to the retention of ancestral gene functions or the emergence of unique differentiation. For example, while the SOC1-like gene PtMADS11 in P. tabuliformis is highly similar to the SOC1 gene of A. thaliana, it cannot complement the late-flowering phenotype of the A. thaliana soc1-1-2 mutant [75].

3.4. Potential Interactions of Key PkMADS-Box Genes and Functional Characterization of PkMADS9

In the floral quarte model of angiosperms, MADS-box transcription factors form multimeric protein complexes to specify floral organ identity. Previous studies in G. gnemon have demonstrated the capability of gymnosperm MADS-box proteins to form multimeric complexes [78]. In P. tabuliformis, PtDAL1 (AGL6-like) acts as a hub in the interaction network, the PtDAL1 physically interacts with PtMADS11 (SOC1-like) to form a heterodimer that mediates the vegetative-to-reproductive transition [75]. Similarly, PtDAL10 (AGL15-like) has been identified as a direct target of PtDAL1 and is tightly engaged in the age-dependent reproductive pathway [77], and PtDAL1 can form ternary complexes with PtDAL10 and PtMADS11. Furthermore, interactome analyses using Y2H-seq identified a wide range of PtDAL1-interacting proteins, such as SOC1-like homologs PtDAL4, PtDAL9, PtDAL19, PtDAL35, PtMADS11, and PtMADS13, establishing a complex regulatory network [72]. In Larix kaempferi, the LaDAL1DAL1 (AGL6-like) was shown to interact with multiple SOC1-like members, LaSOC1-3, LaSOC1-5, and LaSOC1-6 [86]. In this study, PkMADS4 (orthologous to PtDAL1) exhibited broad and high-level expression in reproductive buds (Figure 7). Notably, its expression also followed a distinct age-dependent pattern like PtDAL1, it may act as a hub within the interaction network to facilitate complex formation. Therefore, we hypothesize that PkMADS4 may interact with PkMADS26 (orthologous to PtMADS11) to regulate floral induction, supported by their overlapping expression in developing buds. Moreover, in specific floral organs, PkMADS4 may recruit PkMADS9 (Class B) and PkMADS1 (Class C) to form tissue-specific complexes in LTB to specify male identity. Although conserved domain data and consistency with interaction models in P. tabuliformis and L. kaempferi strongly support the existence of MADS-box combinatorial protein networks in P. koraiensis, direct biophysical experimental verification is still required in future research.

Generally, MADS-box transcription factors function by forming multimeric complexes to regulate downstream targets [87]. In Arabidopsis, the floral transition is strictly regulated by repressors such as SVP and FLC, which inhibit FT and AP1 expression to prevent premature flowering [88,89]. Previous studies on gymnosperm MADS-box genes have shown that overexpression of non-flowering-time genes can induce early flowering in Arabidopsis, such as GbMADS9 (class B) of G. biloba, ref. [90] PtDAL1, PtMADS11, and PtDAL10 of P. tabuliformis [75,76,77]. In the present study, the overexpression of PkMADS9 (Class B) promoted flowering in Arabidopsis, which suggested that this gene may function as a floral activator in P. koraie nsis. Although PkMADS9 acts as a functional transcription factor, given the extensive evolutionary divergence (~300 million years) between gymnosperms and angiosperms, its specific expression in LTB suggested that its native role in P. koraiensis might be likely male organ determination rather than general floral induction. We hypothesized that the overexpression of PkMADS9 in Arabidopsis altered the endogenous flowering regulatory networks, thereby altering the expression of key downstream genes and resulting in the observed early flowering phenotype. This interference might be attributable to the structural conservation of protein–protein interaction domains (specifically the K-domain) in the exogenous PkMADS9, which may enable it to bind to and potentially promote endogenous repressors [91]. Therefore, consistent with the early flowering phenotype, our qRT-PCR analyses revealed the significant up-regulation of positive flowering regulators (AtLFY, AtFT, and AtGI) and key floral organ development genes (AtAP1, AtAP3, AtPI, AtAG, and AtSEP3). These results suggested that while PkMADS9 may serve as a male-determinant in pine, it acts as a potent activator in the heterologous Arabidopsis system by altered the conserved floral network.

In conclusion, the research on the regulatory mechanisms of reproductive development in gymnosperms lags behind that of angiosperms. Although the functions of the PkMADS-box genes identified in this study have been preliminarily investigated via heterologous expression, their unique function and regulatory mechanisms within gymnosperms require further verification through molecular biology experiments in the future study.

4. Materials and Methods

4.1. Plant Materials

The research site is located at the Korean pine seed orchard (42°49′ N, 127°09′ E) of Hongshi Forestry Co., Ltd. in Huadian City, Jilin Province, China. In 2022, plant samples for transcriptome sequencing were collected from 28-year-old trees in this orchard. Based on our team’s research on the characteristics of flower bud differentiation in Korean pine [30], bud samples from terminal buds (TB), vegetative buds (VB), and lateral terminal buds (LTB) were collected during different stages of vegetative growth (S1), mid-physiological differentiation (S2), and early morphological differentiation (S3).

Additionally, to analyze the expression pattern of PkMADS-box genes, samples from different tree ages and floral organ developmental stages of Korean pine were collected in 2025. The age gradient included trees of 2, 5, 10, 15, 20, and 30 years old (renamed as Y1 to Y6, respectively). TB, VB, LTB, and needle samples from the TB (TN) were collected at the early stage of morphological differentiation. Floral organ samples were also gathered from four developmental stages of the ovulate strobilus (Os) and three developmental stages of the microstrobilus (Ms), including the distinguishable period of strobilus bud morphology, the development period I of strobilus, the development period II of strobilus, and the pollination period of Os. Detailed information and numbering of all samples can be found in Table S3. For each sampling, samples from four trees were pooled as a replicate, with a total of three biological replicates. The samples were then frozen in liquid nitrogen and stored at −80 °C.

4.2. Identification of PkMADS-Box Family Genes in Korean Pine

cDNA libraries were constructed, and transcriptome sequencing was conducted on the bud samples (TB, VB, and LTB) of Korean pine. The sequencing was carried out using the Illumina platform by Wuhan Metware Biotechnology Co., Ltd. (Wuhan, China). The transcriptome data were subsequently used to identify MADS-box transcription factors in Korean pine (PRJNA1215794). Protein sequences for MADS-box genes from Pinus tabuliformis, Picea abies, Gnetum gnemon, and A. thaliana were obtained from the NCBI database (https://www.ncbi.nlm.nih.gov/, accessed on 5 January 2025) and TAIR website (https://www.arabidopsis.org/, accessed on 5 January 2025) (Table S4).

To identify all MADS-box family genes in Korean pine, two identification strategies were employed: a local BLAST search using TBtools software (version II) [92] and a Hidden Markov Model (HMM) search using HMMER 3.0 software (http://hmmer.org/, accessed on 5 January 2025). First, the MADS-box reference protein sequences from P. tabuliformis, P. abies, G. gnemon, and A. thaliana were used for a local BLAST search (default parameters: E-value < 1 × 10⁻⁵) to obtain candidate MADS-box family members within the Korean pine transcriptomic database. Second, HMM files for the MADS domain (PF00319) and K domains (PF01486) were downloaded from the Pfam database (http://pfam.xfam.org, accessed on 5 January 2025) and used to identify candidate PkMADS-box family members using HMMER 3.0 with default parameters (E-value < 1 × 10⁻⁵). The results from both identification strategies were integrated, and redundant sequences were removed to obtain candidate PkMADS-box members. The PkMADS-box family members were analyzed using the SMART (http://smart.embl-heidelberg.de/, accessed on 6 January 2025) and the NCBI-CDD database (https://www.ncbi.nlm.nih.gov/cdd, accessed on 6 January 2025) databases to verify the presence and integrality of conserved domains. To ensure PkMADS-box family member sequences quality, candidate PkMADS-box members were filtered based on the following criteria: the sequences with an incomplete MADS domain were excluded; the sequences shorter than 100 amino acids were removed to avoid potential pseudogenes or incomplete fragments. The classification of Type I and Type II lineage MADS-box genes was determined by the presence or absence of a conserved K-box domain at the C-terminus [8,11,12]. The conserved motifs of the PkMADS-box protein sequences were analyzed using the MEME tool (https://meme-suite.org/meme/tools/meme, accessed on 6 January 2025).

4.3. Analysis of Physicochemical Property and Structure of Protein

The physicochemical properties of PkMADS-box proteins were analyzed by ultilizing ExPASy website (https://www.expasy.org/, accessed on 8 January 2025), the tool Compute pI/Mw (https://web.expasy.org/compute_pi/, accessed on 8 January 2025) was utilized to analyze the molecular weight, instability index, isoelectric point, and instability index, and GRAVY value, the tool ProtScale was utilized to analyze the GRAVY value. The online website tool Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/, accessed on 8 January 2025) to analyze the subcellular location. The secondary and tertiary structure prediction of PkMADS-box proteins were analyzed using website tool SPOMA (https://npsa.lyon.inserm.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_sopma.html, accessed on 8 January 2025) online tool and Swiss-model (https://swissmodel.expasy.org/, accessed on 8 January 2025), respectively.

4.4. Phylogenetic and Classification Analysis

To investigate the phylogenetic relationships and the classification of the PkMADS-box genes, the MADS-box protein sequence of P. koraiensis, P. tabuliformis, P. abies, G. gnemon, and A. thaliana were aligned using the Muscle Wrapper V5 tool in TBtools software. The phylogenetic tree was then created using maximum likelihood (ML) method in MEGA 11 software, with parameters of 1000 bootstraps. The resulting ML tree was visualized in website iTOL (http://itol.embl.de/, accessed on 8 January 2025).

4.5. Expression of PkMADS Genes in Buds and Floral Organs by qRT-PCR Analysis

Total RNA extraction and cDNA synthesis were performed using the RNAprep Pure Plant Kit (Tiangen, Beijing, China) and the cDNA Synthesis Kit (Takara, Kyoto, Japan), respectively. qRT-PCR was performed using TB Green Premix Ex Taq II (Takara, Kyoto, Japan) on the CFX Opus 96 system (Bio-Rad, Hercules, CA, USA) in a 25 μL reaction volume. The qRT-PCR programs were as follows: 95 °C for 30 s, 30 cycles of 95 °C for 5 s and 60 °C for 30 s, 95 °C for 10 s, 65 °C for 5 s, 95 °C for 5 s. All reactions were performed with three biological replicates, with 18S-RNA serving as the reference gene. 18S-RNA was served as the internal reference gene, and all qRT-PCR analysis were performed with three biological replicates and three technical replicates. Relative gene expression levels of PkMADS-box genes were calculated using the 2^−ΔΔCt method. Primer sequences were designed by Primer Premier 5.0 and listed in Table S5.

4.6. Plasmid Construction, Transformation, and Analysis of Related Genes Expression of Arabidopsis

The CDS of PkMADS9 was amplified from the cDNA of P. koraiensis terminal buds by PCR using specific primers that introduced SacI and XbalI restriction sites (Table S5). The CDS of PkMADS9 was recombined in the PCAMBIA2300-GFP vector and subsequently transformed into Agrobacterium GV3101 strain. Wild type Arabidopsis was transformed via flower dipping method to obtain overexpressed transgenic plants [93], and transgenic lines were selected on MS medium supplemented with 50 mg/L kanamycin. All plants were grown in a growth chamber under LD conditions (16 h light/8 h dark) at 22 °C. Phenotypic analysis, including the number of rosette leaves and flowering time, was performed using the T3 generation transgenic plants. Statistical significance was determined by t-test. The expression levels of genes related to flower development and flowering were analyzed in both transgenic lines and wild-type Arabidopsis using qRT-PCR. AtActin was employed as the internal reference gene (Table S5), and all qRT-PCR analysis were performed with three biological replicates and three technical replicates. The relative expression levels of the genes were calculated using the 2^−ΔΔCt method.

4.7. Statistical Analyses

All statistical analyses were performed using SPSS 26.0. Data are presented as mean ± standard deviation (SD). The Statistical significance was performed using one-way ANOVA and Tukey’s test using depending on the experimental design and data. Figures were generated in Origin 2024, and final illustrations were created in Adobe Illustrator 2022. Error bars in the figures represent the SD of the mean. Means with different lowercase letters in the figures denote significant differences (p < 0.05).

5. Conclusions

In this study, the MADS-box gene family in Korean pine was systematically identified and characterized based on transcriptomic data. A total of 37 PkMADS-box genes were identified, including 6 Type I and 31 Type II (MIKC) genes. These genes were classified into 8 subfamilies, with the SOC1 subfamily being the most prevalent (16 members). All PkMADS-box proteins are localized in the nucleus. Their secondary structures are enriched in random coils, providing the conformational flexibility necessary for forming multimeric complexes that regulate reproductive development. Through expression profiling during floral induction, floral organ development, and the reproductive phase transition, the spatiotemporal expression patterns of PkMADS-box genes were elucidated. Notably, PkMADS4, PkMADS9, and PkMADS26 were identified as key candidate factors regulating the reproductive phase transition and strobilus development. Furthermore, overexpression of PkMADS9 promoting early flowering in transgenic Arabidopsis. Although further functional verification and interaction analyses are needed to unravel the underlying molecular mechanisms, this study lays a theoretical foundation for understanding the reproductive regulation mechanisms and facilitating molecular breeding to enhance seed yield in Korean pine.