Characterization of BrSTM16 from Brassica rapa and functional analysis of its role in regulating cold tolerance

Abstract

The SHOOTMERISTEMLESS (STM) gene family controls shoot apical meristem homeostasis and participates in plant development and abiotic stress adaptation in higher plants. While the conserved roles of STM in maintaining meristem stability are well documented, the functions and regulatory mechanisms of BrSTM genes in Brassica rapa cold tolerance remain elusive, hindering our understanding of stress mediated meristem plasticity. Here, we systematically characterized the Brassica rapa STM gene family and identified 31 BrSTM members, all harboring the conserved KNOXⅠ domain and belonging to the class I KNOX family. Further functional analysis focused on the key candidate BrSTM16. Its promoter contains cold responsive and meristem specific regulatory elements, and the gene displays predominant expression in shoot apical meristems, with distinct tissue-specific and temperature dependent cold responses. Higher cold induction in cold-sensitive varieties indicates that BrSTM16 acts as a vital regulator linking meristem morphology and development to cold acclimation. Gene edited BrSTM16 mutants exhibited reduced survival rates and antioxidant enzyme activities under cold stress, confirming its positive role in regulating Brassica rapa cold tolerance. Transcriptome analysis suggested that BrSTM16 may modulate cold-induced metabolic homeostasis via the phenylpropanoid biosynthesis pathway. The disrupted expression of key pathway genes (PAL, 4CL, F5H) in knockout mutants was closely correlated with impaired cold resistance. Subcellular localization verified that BrSTM16 localizes to the nucleus. Six BrSTM16 interacting proteins with cold-responsive expression patterns were identified by yeast two-hybrid screening, and they participate in the BrSTM16 mediated cold-regulatory network. This study clarifies the characteristics of the BrSTM family and preliminarily reveals that BrSTM16 integrates phenylpropanoid metabolism and protein interaction networks in meristems to govern cold tolerance, providing reliable candidate genes and theoretical foundations for Brassica rapa cold-resistance molecular breeding.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09488-5.

Introduction

In natural environments, agricultural crops commonly encounter various abiotic stresses including drought, high salinity, and cold stress all of which greatly limit their final yield and biomass production. To mitigate these environmental threats, plants have evolved multi-layered adaptive defense mechanisms. Under drought stress, adaptive responses such as reduced leaf area and enhanced root growth is common [1]. Similarly, salt stress, triggers morphological adjustments, including tighter epidermal cells arrangement and significant thickening of the cuticle and wax layers [2]. These modifications simultaneously impede salt infiltration into mesophyll tissues and decrease transpirational water dissipation [3, 4]. From a cellular perspective, external stress stimuli accelerate the rapid generation of reactive oxygen species. These reactive oxygen species act as signal transmitters to activate downstream regulatory pathways, supporting plant adaptive responses to harsh environments [5]. Of these abiotic stresses, cold stress severely threatens crop survival and productivity in temperate and high-latitude regions, and the morphological characteristics of plant meristems are closely correlated with cold tolerance, which is particularly prominent in Brassica rapa.

Brassica rapa, a species with considerable cold tolerance, is widely cultivated in northern China. Prior to winter arrival, obvious morphological adaptive changes occur in Brassica rapa. Aboveground vegetative growth is decelerated, underground roots proliferate well, and the shoot apical meristem, namely growth cone, is buried underground, presenting a creeping growth pattern [6]. Featuring a dome-like morphology, the growth cone is formed by versatile stem cells that serve as the origin of all aerial structures in plant growth. Notably, the morphological variation of the growth cone is a key indicator of cold hardiness in Brassica rapa. Empirical observations indicate a strong correlation between growth cone morphology and cold hardiness: varieties with a concave (depressed) growth cone exhibit superior freezing tolerance compared to those with a convex (protruding) growth cone. This specific morphological difference of pre-winter shoot apical meristem is tightly associated with plant cold resistance, representing a core phenotypic basis for cold adaptation in Brassica rapa. However, the molecular regulatory mechanisms underlying this meristem morphology mediated cold tolerance remain largely unelucidated, which urgently needs in depth exploration.

As a typical member of Class I KNOX homeobox transcription factors, the STM gene exhibits considerable homology to maize KNOTTED1 (KN1), and functions as a primary regulator responsible for growth cone stability maintenance [7]. STM is crucial for preserving the stem cells undifferentiated condition by preventing their transition into specialized cell lineages, and its loss of function leads to growth cone arrest [8]. Its expression is ubiquitous in the growth cone but is precisely downregulated at sites of organ initiation, often correlating with local auxin maxima [9]. The function and subcellular trafficking of STM are tightly regulated. In Arabidopsis, STM requires heterodimerization with BELL-like homeodomain (BLH) proteins for nuclear localization and function. Furthermore, its intercellular movement via plasmodesmata is selective and modulated by associated proteins, such as the microtubule-associated protein MPB2C in maize [10]. Recently, FTIP3 and FTIP4 were identified in Arabidopsis as critical regulators that retain STM in the nucleus of growth cone cells, preventing its mis-localization to the plasma membrane. Mutants lacking FTIP3/4 show aberrant STM membrane localization, loss of nuclear function, premature growth cone termination, and a bushy dwarf phenotype [11].

Apart from governing plant growth and morphogenesis, STM exerts considerable effects on abiotic stress resilience. It contributes to improved drought resistance via direct induction of Isopentenyl Transferase 7 (IPT7) expression in growth cone tissues, which further facilitates cytokinin biosynthesis [12, 13]. Such interplay presents a bidirectional regulatory loop. Rising endogenous cytokinin concentrations enhance STM expression, and externally supplied cytokinin can ameliorate the impaired meristematic phenotype of STM mutants [14, 15]. ABA-mediated MYB96 can directly target the STM promoter. The expression of STM is upregulated in myb96-ox, while ABA fails to relieve the inhibition of STM expression in the myb96-1 mutant. The drought tolerance of the 35 S: STM-MYC line is enhanced, confirming that MYB96-STM is a core module in the abscisic acid-mediated drought stress signaling pathway [16]. Despite its established roles in model plants, the function of STM in the winter hardy growth cone of Brassica rapa is unknown.

Accumulating studies have proven that transcription factors, phytohormone signaling and flavonoid-mediated secondary metabolism synergistically regulate plant abiotic stress resistance and meristem stress adaptation in diverse crops. Existing abiotic stress studies have uncovered diverse stress regulatory mechanisms in crop plants, core transcriptional networks govern stress-related testa pigmentation and defense metabolism in peanut under adverse conditions. Flavonoid biosynthetic genes enhance drought tolerance by modulating secondary metabolism in safflower, and AP2/ERF transcription factors regulate flavonoid biosynthesis to strengthen plant stress resistance [17–19], and exogenous hormone application enhances peanut salt tolerance by remodeling internal stress signaling pathways [20]. For stress-resistant medicinal crop safflower, the key flavonoid synthetic enzyme DFR participates in abiotic stress adaptation by modulating leucoanthocyanidin accumulation under adverse conditions [21], whereas CtCYP71A1 confers safflower drought resistance via regulating lignin deposition and cell wall stress stability [22]. Exogenous gibberellin can alleviate salt-induced oxidative damage and strengthen safflower salt tolerance by activating flavonoid antioxidant defense pathways [23]. Besides, MYB transcription factor-mediated anthocyanin metabolism serves as a core defensive strategy against multiple abiotic stresses in plants [24]. Notably, most of these studies merely explore independent stress-responsive pathways, limited research focuses on the synergistic regulatory mechanism of KNOX-type STM genes linking hormone signaling, apical meristem morphological adaptation and flavonoid defensive metabolism to facilitate plant cold resistance, which remains poorly characterized in cold-resistant Brassica rapa.

Given the critical involvement of STM in both meristem maintenance and stress responses, this research aimed to explore its potential role in linking morphology to cold tolerance in Brassica rapa. A comprehensive genome-wide analysis coupled with experimental validation of the STM gene family was performed in Brassica rapa. Integrating transcriptome profiling with RT-qPCR analysis, we pinpointed BrSTM16 as a cold-responsive gene and obtained initial experimental evidence supporting its involvement in low-temperature tolerance. Moreover, we constructed a yeast two-hybrid library for screening upstream and downstream proteins that interact with BrSTM16, and further analyzed the transcriptional profiles of these screened genes. This research expands our understanding of the evolutionary development of BrSTM family members, supplies valuable candidate genes for breeding resilient crop varieties, and facilitates further exploration of the molecular mechanism by which BrSTM16 modulates plant environmental adaptability.

Materials and methods

Plant materials, growth conditions, and stress treatments

Two distinct Brassica rapa accessions with contrasting cold adaptability were selected as experimental materials in this research. The full names of all plant seeds and materials used in this study are as follows: Brassica rapa Longyou 7 (abbreviated as L7), Brassica rapa Longyou 99 (abbreviated as L99), Arabidopsis thaliana, Nicotiana benthamiana, Brassica napus Weaster (abbreviated as WT), and BrSTM16 gene-edited Brassica napus lines (abbreviated as BrSTM16 edited lines). All the above-mentioned plant materials were provided by the State Key Laboratory of Crop Science in Arid Habitats, Gansu Agricultural University. Among them, L7 possesses superior freezing resistance, whereas L99 shows relatively weak tolerance to low temperature. Seeds were treated with 10% (v/v) hydrogen peroxide for surface sterilization over a 30-minute period, subsequently washed extensively with distilled water, and then allowed to germinate. To achieve consistent vernalization, rapeseed seedlings were initially exposed to cyclic conditions comprising 14-hour illumination at 28 °C and 10-hour dark periods at 0 °C. Following this procedure, all materials were grown in a phytotron under day/night temperatures of 25 °C/20°C with identical photoperiods, reaching the seven-leaf stage [25]. Uniformly developed seedlings were subjected to cold stress treatments at 4 °C, 0 °C, and − 4 °C for 72 h; plants grown continuously at 25 °C served as the control. Post-treatment sampling was performed independently on roots, stems, leaves, and growth cones. All samples were rapidly frozen in liquid nitrogen and stored at − 80 °C for subsequent RNA isolation and RT-qPCR analysis, with three independent biological replicates per treatment. For subcellular localization analysis, Nicotiana benthamiana seeds underwent 3-day low-temperature stratification at 4 °C. Subsequently, the seeds were sown in sterilized growth medium and grown in greenhouse conditions for 4 weeks prior to transient infiltration.

Genome-wide identification and physicochemical analysis of STM genes

The HMM profile corresponding to the KNOX1 (STM) domain (Pfam accession PF037090) was retrieved from the Pfam database (http://pfam.xfam.org/) [26]. The profile was used to search Brassica rapa protein sequences (NCBI, SRR18959686) with HMMER v3.3.2 and default e-value cutoffs [27]; candidate protein sequences were subsequently analyzed for the presence of evolutionarily conserved domains using the NCBI Conserved Domain Database (CDD). (https://www.ncbi.nlm.nih.gov/cdd/) [28]. The physicochemical properties—including molecular weight, theoretical isoelectric point (pI), and instability index—of the identified STM proteins were predicted using ExPASy ProtParam. (http://web.expasy.org/protparam/). Subcellular targeting of the proteins was computationally predicted using Cell-PLoc 2.0. (http://www.csbio.sjtu.edu.cn/bioinf/Cell-PLoc-2/) [29].

Phylogenetic, gene structure, and cis element analysis

Genome assemblies and annotation documents of six species were acquired from Ensembl Plants, involving Arabidopsis thaliana, Brassica rapa subsp. pekinensis, Brassica napus, Brassica rapa, rice and maize. ClustalW was employed to align STM protein sequences among these species. A neighbor-joining phylogenetic tree was then constructed in MEGA11, with 1000 bootstrap replicates applied for statistical validation [30]. Visual display and functional annotation of the phylogenetic tree were accomplished using the web-based iTOL platform accessible at https://itol.embl.de [31]. A maximum of ten conserved motifs among BrSTM genes were characterized via motif scanning analysis using MEME v5.5.5 [32]. Exon-intron organization of genes was plotted with TBtools-II [33]. Upstream 2000 bp sequences flanking the translation start site were extracted as promoter sequences by TBtools. Cis-element prediction was subsequently carried out using the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

Chromosomal localization, gene duplication, and synteny analysis

TBtools was used to map BrSTM gene chromosomal positions, while MCScanX (default parameters) analyzed gene duplication events. Tandem duplicates referred to homologous genes with ≤ 10 intervening genes on the same chromosome. Ka/Ks analysis for duplicated gene pairs was performed with KaKs_Calculator 3.0 to assess non-synonymous and synonymous substitution rates. Corresponding Ks data were utilized to infer the occurrence time of gene duplication. Besides, synteny analysis results among Arabidopsis thaliana, Brassica rapa and Brassica napus were plotted with the Advanced Circos function in TBtools [34].

Expression pattern analysis of BrSTM genes

Public RNA-seq expression datasets of multiple Brassica rapa tissues were extracted from the BRAD repository (http://brassicadb.cn), involving root, stem, leaf, flower, silique and callus samples [35]. These data were integrated with in-house transcriptome datasets generated from cold-stressed seedlings (as per Section "Plant materials, growth conditions, and stress treatments"). Transcript levels (FPKM) were log₂-transformed using the formula log₂ (FPKM + 1). Heatmaps were generated using TBtools to visualize tissue-specific and stress-responsive expression profiles. RT-qPCR detection was carried out to verify expression patterns of the screened genes.

RT-qPCR, vector construction, and generation of transgenic Arabidopsis

Total RNA was isolated from collected samples using commercial plant RNA kits, and genomic DNA contamination was removed via on-column DNase I treatment. First-strand cDNA synthesis was achieved with the PrimeScript RT reagent kit. RT-qPCR analysis was performed on the QuantStudio system with SYBR Green master mix (Table S1). Using Brassica rapa Actin as internal control, relative expression values were computed by the 2^−ΔΔCT method [36].

The full-length cDNA of BrSTM16 was amplified by PCR using PrimeSTAR High Fidelity DNA Polymerase (Takara), recombined into the pDONR vector with BP Clonase II enzyme mix (Invitrogen), and then LR-recombined with the destination vector pEarlyGate101 driven by the 35 S promoter using LR Clonase II enzyme mix (Invitrogen) to transfer the gene into the target vector. The recombinant construct was introduced into Agrobacterium GV3101 competent cells via electroporation, followed by floral dip transformation in Arabidopsis thaliana Col-0 plants [37]. T1 transgenic seeds were germinated on 1/2MS medium with 10 mg L⁻¹ phosphinothricin. After successive generation screening, homozygous T3 lines were chosen for further functional investigation.Cold stress tolerance was evaluated by subjecting four-week-old transgenic and wild-type plants to freezing temperatures (-4 °C) for durations of 3, 6, 12, or 24 h, followed by a 7-day recovery period under normal growth conditions (22 °C), after which survival rates were determined. Physiological indices (e.g., relative electrolyte leakage, malondialdehyde content) were measured, with all experiments performed in three biological replicates.

Subcellular localization of BrSTM16

The full-length BrSTM16 coding region without termination codon was ligated in-frame upstream of GFP in the pEarlyGate101-GFP plasmid. Both recombinant fusion vector and 35 S-driven GFP empty vector were transferred into Agrobacterium tumefaciens GV3101 strain. After cultivation to a defined OD₆₀₀ value, bacterial pellets were re-suspended in freshly prepared infiltration medium (10 mM MES, 10 mM MgCl₂, 150 µM acetosyringone, pH 5.6) and infiltrated into the abaxial leaf surface of 5-week-old tobacco seedlings. After dark incubation for 48–72 h, confocal laser scanning microscopy was applied to detect GFP fluorescence at 488 nm excitation and 500–530 nm emission. DAPI staining was additionally conducted to mark cellular nuclei [38].

Prediction and analysis of miRNA–BrSTMs and BrSTM16–protein interactions

The psRNATarget web server was adopted to identify potential microRNAs that target the CDS regions of BrSTM genes. In addition, homologous protein interaction analysis of BrSTM16 was conducted on STRING v12.0 database. Using Arabidopsis thaliana as reference organism, only interactions with a combined score above 0.6 were retained for subsequent analysis [39]. Cytoscape v3.10.0 was employed for the retrieval and visualization of the constructed interaction network. Nodes with no predicted protein-protein connections (isolated nodes) were removed from the network.

Transcriptome sequencing

Plant tissues were sampled from the growth cone of 7 leaves stage seedlings from BrSTM16 knockout lines and corresponding wild type plants, with three independent biological replicates per genotype.Total RNA was extracted using TRIZOL (Invitrogen, Carlsbad, CA, USA). After the samples were qualified, the transcriptome library was constructed. The process included rRNA removal, random fragmentation of RNA, synthesis and purification of double-stranded cDNA, end repair, addition of A-tails and adapter ligation, AMPure XP beads for fragment selection, USER enzyme treatment and PCR enrichment to obtain the cDNA library. After the library construction was completed, the initial quantification was performed using Qubit 3.0 (concentration ≥ 1 ng/µL), followed by the detection of insert fragment size with Qsep400. Finally, the effective concentration was accurately quantified by qPCR (> 2nM) to complete the library quality control to meet the sequencing requirements [40]. The raw reads generated by high-throughput sequencing were in fastq format and uploaded to the NCBI database (accession number: PRJNA1467764).

Yeast two-hybrid (Y2H) assay

BrSTM16 coding region was ligated into pGBKT7 to generate the recombinant bait plasmid. To examine bait autoactivation and potential toxicity, this construct was co-transformed with empty pGADT7 into Y2HGold yeast strain, which was then grown on DDO and QDO selective media with X-α-Gal addition. Meanwhile, a Brassica rapa cDNA library was established within the pGADT7 vector. For subsequent screening assays, yeast cells were co-transformed with pGBKT7-BrSTM16 and the constructed cDNA library, and spread onto QDO/X/A medium with Aureobasidin A supplementation. Parallel positive and negative control combinations were arranged in the experiment. Following 3–5 days of incubation at 30 °C, distinct blue colonies were isolated. Corresponding prey plasmids were extracted and sequenced, and direct interaction validation was completed via back-transformation of candidate prey vectors with the original bait construct [41].

Results

Identification and physicochemical properties of the BrSTM gene family in Brassica rapa

In order to elucidate the regulatory functions of STM family members in response to external abiotic stressors, comprehensive genome-wide characterization was carried out on this gene family in Brassica rapa.This analysis identified 31 STM family members in the Brassica rapa genome, designated as BrSTM1 to BrSTM31 according to chromosomal order (Table S2). Bioinformatic characterization of the predicted proteins revealed considerable variation in their physicochemical attributes. The amino acid lengths varied from 327 residues (BrSTM6) to 672 residues (BrSTM26), corresponding to predicted molecular weights between 36.7 kDa (BrSTM17) and 73.4 kDa (BrSTM1). The instability indices were all above 34.8, indicating that these proteins are generally unstable. Considerable differences were observed in theoretical pI values across family members. Among them, BrSTM6 possessed the lowest value of 4.91 (acidic), while BrSTM27 exhibited the highest pI of 8.87 (basic). All 31 BrSTM proteins were consistently predicted to localize exclusively in the nucleus a finding that aligns with their anticipated function as transcriptional regulators. This variability in sequence length and biochemical properties suggests a corresponding functional diversity within the BrSTM protein family.

Phylogenetic analysis and evolutionary relationships of the STM gene family

To further elucidate the phylogenetic divergence and evolutionary characteristics of STM genes across diverse plant lineages, six representative species were deliberately selected, with a focus on comparative evolutionary analysis between Brassicaceae (dicot) and monocot species. A total of 153 STM genes were identified from these six species: Arabidopsis thaliana (12, AtSTM), Brassica rapa subsp. pekinensis (35, BraSTM), Brassica rapa (31, BrSTM), Brassica napus (38, BnSTM), Oryza sativa (14, OsSTM), and Zea mays (23, ZmSTM). Among them, Brassicaceae species were chosen as they are closely related to Brassica rapa and share common whole-genome duplication events, while rice and maize were included as typical monocot models for inter-species evolutionary comparison. A phylogenetic tree was then constructed using these STM protein sequences, which distributed the 153 genes into seven distinct clades designated as Clade I to VII (Fig. 1 and Table S3). The distribution of members among clades was uneven. Clade VII was the largest, containing 53 members, followed by Clade IV with 41 members. Clades II and III contained 22 and 20 members, respectively, while Clades V, VI, and I were smaller, with 9, 6, and 2 members each.

Fig. 1.

Phylogenetic analysis of STM gene family across various species. (different colors represent distinct subfamilies)

Evolutionary analysis revealed distinct structural features associated with each clade. Clade I, containing only two BnSTM members, is characterized by a conserved MEINOX domain. Members of Clade II possess a unique N-terminal extension and an acidic region. Clade III genes feature a proline and glutamine rich N-terminal region. Clade IV shows the greatest degree of evolutionary divergence. Clade V members have a distinctive C-terminal structure. Clade VI is defined by the presence of conserved KNOX1 and KNOX2 domains. Notably, Clade VII contains most of the Brassica rapa STM members, which cluster closely with homologs from Brassica napus and Brassica rapa subsp. pekinensis, suggesting this clade may represent an ancestral group with conserved functions.

Species distribution analysis indicated that Arabidopsis STM genes are primarily located in Clades I, II, III, and IV. In contrast, members from the monocots O. sativa and Z. mays are predominantly enriched in the highly divergent Clade IV, which may reflect clade specific adaptations related to monocot development. Members of the Brassica rapa STM family were scattered throughout every major evolutionary clade, implying remarkable lineage-specific expansion and functional divergence of this gene set. Based on functional annotations of orthologs, these genes are predicted to be involved in fundamental processes such as growth cone maintenance, lateral organ development, and potentially, tissue-specific responses to environmental stimuli. The uneven phylogenetic distribution and distinct structural characteristics of each clade likely result from natural selection during speciation and establish a genetic basis for functional diversity of the STM gene family.

Gene structure, conserved motifs, and domain analysis of BrSTMs

To uncover intrinsic structural traits of Brassica rapa BrSTM genes, comprehensive analyses of evolutionary relationships, conserved motifs and protein functional domains were carried out. Phylogenetic analysis classified all BrSTM proteins into three independent evolutionary subgroups (Fig. S1A). This grouping pattern is consistent with that of Arabidopsis STM proteins but differs from the classification observed in Zea mays, suggesting lineage-specific evolutionary paths. Analysis of conserved motifs revealed that all 31 BrSTM proteins contain both Motif 1 and Motif 3 (Fig. S1B), indicating these motifs are essential core elements for the family’s function. Groups I and II lack Motif 4 but uniquely possess Motif 2 and Motif 5, suggesting these groups have undergone distinct functional specialization. The presence or absence of other motifs, such as the widely distributed Motif 6, further implies that variations in motif composition contribute to functional diversification within the family.

Analysis of conserved protein regions revealed the universal presence of Homeobox_KN motif across all BrSTM members. This domain is recognized as a typical DNA-binding element belonging to the KNOX subfamily of homeobox transcription factors (Fig. S1C). This domain enables BrSTM transcription factors to act as molecular switches, directly regulating downstream target genes involved in meristem maintenance and organogenesis. Groups I and II additionally possess POX and POX superfamily domains, which are associated with peroxidase activity. Such results indicate potential dual biological activities of these proteins. On one hand, they regulate intracellular ROS metabolism to counteract oxidative injury; on the other hand, they facilitate lignin biosynthesis and cell wall structural reinforcement, ultimately improving comprehensive stress resilience against both biotic and abiotic stimuli [42].

In contrast, Group III proteins are characterized by a more complex domain architecture, including the canonical KNOX1 and KNOX2 domains which are critical for maintaining meristem indeterminacy and autoregulation, respectively and an ELK domain [43]. The ELK domain mediates protein–protein interactions, thereby expanding the regulatory network of these transcription factors [44]. Notably, all proteins with a predicted isoelectric point (pI) below 6.0 clustered in Group III, while those with a pI above 6.0 were exclusive to Groups I and II. This distinct biochemical property supports the phylogenetic division and may indicate a more recent evolutionary origin for Group III.

Analysis of gene architecture revealed that closely related members within the phylogenetic tree tend to share highly conserved exon–intron organization patterns (Fig. S1D). A clear group specific pattern was observed in untranslated region (UTR) composition, most Group I members possess three or more UTR exons, whereas Groups II and III members uniformly contain only two. The presence of multiple UTR exons in Group I may facilitate more complex post-transcriptional regulation [45]. All BrSTM genes contain coding sequence (CDS) exons, with Group III members exhibiting a highly uniform and distinctive structure of five or more CDS exons. This allows for the encoding of proteins with more complex amino acid sequences and sophisticated spatial architectures. Such structural complexity can provide a greater diversity of functional domains and regulatory interfaces, potentially underpinning the specialized and precise roles of Group III proteins in critical processes like meristem maintenance, organ differentiation, and stress adaptation [46].

Analysis of cis-acting elements of BrSTMs gene promoter

In order to probe into transcriptional modulation mechanisms of BrSTM family members, we performed cis-element prediction on their 2000-bp upstream promoter regions through the PlantCARE platform. In total, 28 different categories of functional regulatory motifs were detected (Fig. S2A and Table S4). These elements were functionally categorized into groups involved in light responsiveness, hormone response, plant metabolism and growth regulation, and adaptive responses to biotic and abiotic stress conditions. In-depth motif classification indicated that functional elements involved in light signaling, ABA mediation, MeJA regulation, as well as anaerobic adaptation were abundantly present in the promoter regions of the majority of BrSTM family genes. Notably, the promoters of three genes BrSTM8, BrSTM16, and BrSTM23 contained both low temperature response elements (LTRs) and meristem specific regulatory elements (Fig. S2B). Every BrSTM gene promoter possessed at least one element related to environmental stress or developmental regulation, and hormone responsive elements were ubiquitous among all family members (Fig. S2C).

These findings suggest that BrSTM gene expression is finely tuned through the combinatorial action of a broad spectrum of cis-regulatory motifs, implicating them in fundamental processes such as meristem regulation, stress adaptation, and photosynthetic organ development. The widespread occurrence of elements linked to cold stress responses in most promoters may reflect an evolutionary adaptation of this gene family to cold acclimation. The colocalization of low temperature and meristem specific elements in BrSTM8, BrSTM16, and BrSTM23 further indicates that these specific genes are likely key regulatory nodes integrating meristem development with low temperature signaling, providing a molecular basis for their potential role in conferring cold tolerance in Brassica rapa.

Chromosomal localization, collinearity and evolutionary analysis of STM genes

Based on the phylogenetic analysis (Fig. 1), we performed chromosomal mapping to investigate the distribution of STM genes in Arabidopsis thaliana, Brassica rapa subsp. pekinensis, and Brassica rapa. The results confirmed that STM genes are distributed across multiple chromosomes in all three species (Fig. S3). A notable proportion of these genes were located in sub-telomeric chromosomal regions. The number of identified STM genes increased from Arabidopsis thaliana (12) to Brassica rapa subsp. pekinensis (35) and Brassica rapa (31), consistent with the whole genome triplication events characteristic of the Brassica lineage. This suggests a notable lineage specific expansion of the STM gene family within the Brassica species, likely driven by gene duplication events followed by functional diversification. Despite this expansion, the STM genes in these related Brassicaceae species showed a degree of conservation in their genomic organization, underscoring the fundamental role of this family in core developmental regulation.

Furthermore, several STM genes in Brassica rapa were found in localized gene clusters, which may be the result of recent segmental duplications or local tandem duplication events. Collectively, evolutionary diversification of STM genes is largely attributed to family expansion and internal structural remodeling across the genome. Such evolutionary events are likely responsible for improved environmental adaptability and abundant phenotypic variations in Brassica rapa.

Species evolution is driven by genomic structural variation and natural selection. Gene collinearity, which reflects the conservation of ancestral gene order among related species, provides critical evidence for reconstructing evolutionary relationships and tracing gene family expansion. Intraspecific collinearity analysis identified 4, 18, and 35 collinear gene pairs within the STM families of Arabidopsis thaliana, Brassica rapa, and Brassica rapa subsp. pekinensis, respectively, showing a progressive increase (Fig. 2A). These collinear chromosomal segments were located on diverse chromosomes, revealing that segmental duplication serves as the major evolutionary force driving family size expansion of STM genes Moreover, intergenomic collinearity investigation uncovered 63 evolutionarily conserved gene orthologs across Brassica rapa subsp. pekinensis and Brassica rapa. In contrast, only 31 corresponding homologous pairs were found in the syntenic comparison of Arabidopsis thaliana against Brassica rapa (Fig. 2B). This pattern is consistent with their closer phylogenetic relationship and shared polyploidization history.

Fig. 2.

Collinearity analysis of the STM gene. A Intraspecific collinearity. B Interspecific collinearity

To quantify evolutionary selective pressure across homologous orthologs, substitution rate analysis of Ka/Ks was conducted (Table S5). Consistently, all calculated values remained under 1, demonstrating that intense purifying selection substantially governed the evolutionary differentiation of the STM gene family. This maintains the high functional conservation of these genes by eliminating deleterious mutations. Previous studies support the functional importance and conservation of this family. In Arabidopsis thaliana, AtSTM acts as a key regulator of growth cone maintenance and integrates environmental signals (e.g., low temperature, drought) into developmental programs [47]. Similarly, the STM orthologs identified in Brassica rapa subsp. pekinensis govern meristematic activity and display inducible expression under salt and low-temperature conditions, verifying their conserved dual roles in developmental regulation and stress responses across cruciferous species [48].

In conclusion, the STM gene family demonstrates strong sequence and functional conservation across Brassicaceae species. Its expansion is closely linked to genomic duplication events. While maintaining its core function in meristem regulation, this family appears to have undergone network expansion and optimization, likely enhancing plant adaptability under natural selection and playing a critical role in species evolution and environmental adaptation.

Prediction and analysis of miRNA-BrSTM interactions

In this study, we computationally predicted miRNAs capable of targeting BrSTM genes and systematically assessed their potential involvement in regulating plant growth, developmental processes, and stress adaptation. A total of 88 distinct miRNAs were found to target 30 BrSTM genes, with several of these genes being subject to regulation by up to four different miRNAs. and a few genes could even be co-targeted by 9 miRNAs, suggesting the existence of a complex regulatory network between miRNAs and BrSTM (Fig. 3). Collectively, our data reveal that miRNAs implement elaborate and accurate regulation on BrSTM expression, exerting key regulatory effects throughout plant development, morphogenesis, and stress tolerance processes. Among all predicted interactions, bra-miR9561-5p is capable of targeting 8 BrSTM family members (BrSTM3, BrSTM5, BrSTM10, BrSTM11, BrSTM17, BrSTM18, BrSTM20, BrSTM21), implying that this miRNA participates in various biological processes including plant development and stress responses through the coordinated regulation of multiple BrSTM genes. The results of this study reveal the complex interaction between miRNAs and BrSTM, laying a theoretical foundation for further elucidating their regulatory functions in plant physiological processes and for subsequent studies on gene function and molecular mechanisms.

Fig. 3.

Analysis of the interactions between candidate miRNAs and their target BrSTM in the potential regulatory network. The black lines represent the interaction relationships between nodes, and the darker the node color, the more target nodes it has

Analysis of the expression patterns of BrSTMs family members

To characterize the expression profiles of BrSTM genes, we performed hierarchical cluster analysis on RNA-seq data from various Brassica rapa tissues obtained from the BRAD database. The findings demonstrated clear tissue-specific expression patterns (Fig. S4A, and Table S6). For instance, BrSTM6, BrSTM7, BrSTM13, BrSTM16, and BrSTM18 showed high expression in callus tissue. Elevated expression in flowers was observed for BrSTM16, BrSTM18, BrSTM23, and BrSTM26. BrSTM16 and BrSTM25 were expressed in leaves, roots, and stems, while expression in siliques was particularly prominent for BrSTM16 and BrSTM18. Notably, BrSTM16 displayed consistently high expression across all tested tissues, suggesting a broad, constitutive regulatory role. I In contrast, BrSTM23, BrSTM24, BrSTM28, and BrSTM29 exhibited extremely weak expression signals or were completely repressed across all tested tissue types.

To gain deeper insights into the expression changes of BrSTM genes induced by cold stress, transcriptome profiling was performed using RNA-seq datasets from two Brassica rapa varieties Longyou 7 (L7) and Longyou 99 (L99), subjected to low-temperature conditions. The expression levels of individual members varied dynamically during cold exposure. Notably, BrSTM16, BrSTM25, and BrSTM26 were significantly upregulated following 4 °C treatment (Fig. S4B, and Table S7). Intriguingly, most BrSTM family members exhibited stronger basal or inducible expression in the low cold-tolerance cultivar L99 than in the resistant cultivar L7, indicating that the expression of certain BrSTM genes may be negatively correlated with cold tolerance.

To validate these cold-responsive expression patterns, we conducted RT-qPCR analysis on leaves, roots, and growth cones (apical meristems) of both cultivars after 0 °C treatment for 4 and 24 h. Relative to the normal temperature control, the expression of most BrSTM genes rose gradually as cold stress duration increased, peaking at 24 h (Figs. 4 and 5 and Table S8). The induction was tissue-dependent. Several genes, including BrSTM7, BrSTM11, BrSTM16, BrSTM22, BrSTM23, BrSTM28, BrSTM30, and BrSTM31, showed pronounced cold induction. For example, in L7 after 24 h, BrSTM16 expression increased 24.6-fold in the growth cone, 11.5-fold in roots, and 7.9-fold in leaves. Notably, for most of these responsive genes, the highest relative expression levels at 24 h were observed in the apical meristem (growth cone) of L7, suggesting a particularly important role for BrSTM genes in mediating the meristem specific response to cold stress in winter Brassica rapa.

Fig. 4.

Expression characteristics of BrSTMs genes in different tissues of L7 under 0°C cold stress. Statistical analysis was performed using two-way ANOVA followed by Tukey’s multiple comparison test. Data in the figure represent the average of three independent replicates. Error bars indicate the standard deviation (SD) of the three replicates. Statistical significance between treatments was indicated by different symbols: ns indicates p > 0.05, no significant difference; *, **, and *** indicate significant differences at the levels of p < 0.05, p < 0.01, and p < 0.001, respectively

Fig. 5.

Expression characteristics of BrSTMs genes in different tissues of L99 under 0°Ccold stress. Statistical analysis was performed using two-way ANOVA followed by Tukey’s multiple comparison test. Data in the figure represent the average of three independent replicates. Error bars indicate the standard deviation (SD) of the three replicates. Statistical significance between treatments was indicated by different symbols: ns indicates p > 0.05, no significant difference; *, **, and *** indicate significant differences at the levels of p < 0.05, p < 0.01, and p < 0.001, respectively

BrSTM16 gene expression profiling and morphological changes of Brassica rapa before and after overwintering

Given that the BrSTM16 promoter contains low temperature response and meristem-specific regulatory elements (Fig. S2), and transcriptome data showed its constitutive tissue expression (Fig. S4 Table S9), we performed detailed expression analysis via RT-qPCR. This confirmed significant BrSTM16 expression variation across tissues in two Brassica rapa cultivars, with the highest expression in the growth cone (33-fold in L7 and 61-fold in L99 vs. leaves), indicating the shoot apex as its primary functional site. To reliably capture the response characteristics of target tissues under continuous low-temperature stress, three temperature gradients 0 °C, 4 °C, and − 4 °C were applied for a 72-hour continuous cold treatment. Analysis of cold responses revealed that gene expression exhibited clear tissue-specificity and temperature-dependent induction patterns, with the effects being most pronounced in stems and growth cone (Fig. 6B-F). Notably, BrSTM16 expression was consistently higher in the cold-sensitive cultivar (L99, protruding growth cone) than in the cold-resistant one (L7, sunken growth cone). Specific responses included: leaf expression suppressed except modest induction at -4 °C (1.6-fold in L7, 2.1-fold in L99); root upregulation (6.1-fold) only in L7 at -4 °C; and progressive expression increase in stems and meristems with decreasing temperature, peaking at -4 °C (20.1-fold in L7, 33.3-fold in L99 in meristems). This higher basal and induced expression in L99 reflects compensatory activation to counter cold stress, identifying BrSTM16 as a key cold-responsive regulator linking meristem development/morphology to cold acclimation. Field observations corroborated these results. Line L7 exhibited a prostrate seedling habit and maintained a compact, low growth cone architecture post-overwintering, while the semi erect line L99 retained an erect habit with exposed, protruding growth cones and suffered severe cold injury (Fig. 6A). Given that growth cone morphology was evaluated via preliminary field observations, the observed morphological differences suggest an association with cold tolerance. Combined with the high meristematic expression of BrSTM16, this preliminary evidence supports that BrSTM16 may contribute to overwintering cold adaptability by regulating plant architecture and growth cone morphology.

Fig. 6.

Phenotypic Morphology and Tissue-specific Expression of BrSTM16 in Brassica rapa During Overwintering. A a-c: L7 before overwintering; d-f: L99 before overwintering; g-i: L7 after overwintering; j-l: L99 after overwintering. B Expression levels of BrSTM16 in different tissues. C Expression of BrSTM16 gene in leaves under low-temperature stress. D Expression of BrSTM16 gene in stems under low-temperature stress. E Expression of BrSTM16 gene in roots under low-temperature stress. F Expression of BrSTM16 gene in growth cones under low-temperature stress. Statistical analysis was performed using two-way ANOVA followed by Tukey’s multiple comparison test. Data in the figure represent the average of three independent replicates. Error bars indicate the standard deviation (SD) of the three replicates. Statistical significance between treatments was indicated by different symbols: ns indicates p > 0.05, no significant difference; *, **, and *** indicate significant differences at the levels of p < 0.05, p < 0.01, and p < 0.001, respectively

Phenotypic and physiological analysis of BrSTM16 overexpression lines in Arabidopsis

Basta-based screening was performed to identify positive transgenic Arabidopsis plants, and homozygous lines were isolated via genetic segregation analysis. Molecular identification confirmed the stable integration and overexpression of BrSTM16, and the S2 line was chosen for subsequent functional investigation (Fig. S5). Phenotypic comparison between the wild-type (WT) and the high-expressing BrSTM16 overexpression line (S2) revealed that transgenic seedlings developed broader and exhibited larger leaves relative to wild-type plants under normal conditions (Fig. S6A). However, plant structure showed no significant variation at later growth phases including flowering and maturity (Fig. S6B and C).

To assess the role of BrSTM16 in plant cold resistance, 4-week-old WT and transgenic overexpression lines were exposed to -4 °C for varying durations, followed by 7 days of room-temperature recovery. Key physiological indicators of the plants were quantitatively analyzed. Phenotypic and survival rate statistics showed that at -4 °C for 3 h and 6 h, both WT and overexpression plants grew normally with no significant phenotypic differences. However, when the stress duration was extended to 12 h and 24 h, the survival rates of the two types of plants showed a highly significant difference (P < 0.01). All WT plants died, while the survival rates of overexpression plants reached 89% and 67%, respectively, and most of the surviving plants could recover normal growth (Fig. S6D and E). Concurrently, the transcript abundance of BrSTM16 exhibited a pronounced increasing trend as the duration of low-temperature stress prolonged (Fig. S6F). As exposure time prolonged, enhanced enzymatic activities of superoxide dismutase and peroxidase were observed across the two genotypes. On the contrary, catalase activity and free proline content first climbed to a peak temporarily and then displayed a downward trend (Fig. S6G-J). Of particular interest, the temporal variation in physiological responses varied markedly across different genotypes. When subjected to short-duration low-temperature treatment (0–3 h), WT plants exhibited higher initial levels of SOD, POD, and PRO. As the treatment progressed, the overexpression lines surpassed WT in these metrics. Particularly, at the 24 h time point, remarkable improvements were detected in transgenic overexpression lines compared with wild-type controls. Specifically, their SOD activity, POD activity, and proline accumulation were elevated by 9.6%, 28.5%, and 34.7%, respectively. CAT activity remained significantly elevated in the overexpression lines throughout the treatment period. This coordinated enhancement of the antioxidant system increased SOD activity for superoxide anion scavenging, elevated POD and sustained CAT activity for hydrogen peroxide detoxification, and greater accumulation of the osmoprotectant proline collectively indicates a more robust and sustained physiological response to low-temperature stress in BrSTM16-overexpressing line. These changes likely contribute to reduced oxidative damage, maintenance of membrane integrity, and improved osmotic adjustment, elucidating a key physiological mechanism by which BrSTM16 overexpression enhances cold tolerance.

Knockout of the BrSTM16 ortholog in Brassica napus impairs cold stress tolerance

To verify the conserved function of BrSTM16 during cold responses, the orthologous gene of BrSTM16 in Brassica napus was edited using CRISPR/Cas9 (designated as bnstm16-KO) (Fig. S9). Weaster (WT) Brassica napus and bnstm16-KO lines were exposed to 0 °C for 0, 4, 8, 12, and 24 h. Phenotypic observations showed that leaves of bnstm16-KO plants began to wilt and droop after 8 h of treatment, and the difference became increasingly prominent with prolonged stress (Fig. 7A). Physiological measurements revealed that MDA content and relative electrolyte leakage were significantly higher in bnstm16-KO plants than in WT at all time points. In contrast, activities of CAT, POD, and SOD were markedly lower in the knockout lines. MDA peaked at 8 h, while CAT and SOD activities peaked at 12 h, and POD activity peaked at 8 h (Fig. 7B–F). Collectively, knockout of the BrSTM16 ortholog in Brassica napus aggravated cold-induced membrane damage, dampened antioxidant enzyme activation, and significantly reduced cold tolerance, confirming that BrSTM16 orthologs positively regulate cold responses in Brassica napus. Given the high sequence similarity and conserved function of STM genes among Brassica species, the impaired cold tolerance in Brassica napus knockout lines further supports that BrSTM16 acts as a positive regulator of cold resistance in Brassica. rapa.

Fig. 7.

Comparison of cold stress phenotypes and physiological indices between its knockout mutants and the wild type. A Comparison of plant phenotypes after 0 °C low temperature treatment for 4, 8, 12, and 24 h. B Relative conductivity; C superoxide dismutase (SOD); (D) Physiological indicators of catalase (CAT); (E) Malondialdehyde (MDA); (F) peroxidase (POD). Statistical analysis was performed using two-way ANOVA followed by Tukey’s multiple comparison test. Data in the figure represent the average of three independent replicates. Error bars indicate the standard deviation (SD) of the three replicates. Statistical significance between treatments was indicated by different symbols: ns indicates p > 0.05, no significant difference; *, **, and *** indicate significant differences at the levels of p < 0.05, p < 0.01, and p < 0.001, respectively

Transcriptome sequencing was performed using growth cone tissues from WT and bnstm16-KO Brassica napus to globally characterize transcriptional changes caused by loss of function of the BrSTM16 ortholog. Biological replicates showed high consistency (Pearson correlation coefficient > 0.86). PCA revealed clear separation between genotypes, with PC1 accounting for 84.8% of total variation. In total, 8,926 DEGs were identified (|log2(fold change)| ≥ 1, FDR < 0.05), including 4,323 upregulated and 4,603 downregulated genes (Fig. S8A–D). GO enrichment showed that DEGs were enriched in photosynthesis, ROS scavenging, and carbohydrate metabolism. KEGG analysis highlighted significant enrichment in environmental adaptation, primary metabolism, MAPK signaling, and hormone transduction pathways (Fig. S8E–F).

Expression profiling of transcription factors revealed extensive transcriptional reprogramming upon knockout of the BrSTM16 ortholog in B. napus. Key genes involved in starch and sucrose metabolism were significantly altered, including those encoding trehalose-6-phosphate synthase (TPS) and sucrose phosphate synthase (SPS). Among them, BnC05g0711070, BnC04g0635270, BnA05g0186770, and BnA03g0147290 exhibited prominent upregulation in the bnstm16-KO mutant. In addition, transcription factors associated with TMEM222, a membrane-related protein involved in ethylene biosynthesis regulation, were also significantly induced (Fig. 8). These results indicate that the BrSTM16 ortholog modulates multiple downstream pathways involved in sugar metabolism, hormone signaling, and transcriptional regulation, thereby shaping the cold-responsive gene network and contributing to chilling and freezing tolerance in Brassica napus.

Fig. 8.

Hierarchical clustering heatmap of differentially expressed transcription factor genes in brstm16-KO plants. (A) MYB family; (B) bHLH family; (C) homeobox family; (D) GATA family; (E) SRF family; (F) HSF family; (G) ZBTB family; (H) zf-C2H2 family; (I) CCCH family. The heatmap of differentially expressed genes is constructed based on gene expression levels (FPKM), and data rows are standardized using the Z-score method for clustering analysis. Red and yellow indicate significant upregulation and downregulation of gene expression in brstm16-KO, respectively

Whole-transcriptome identification of miRNA differential expression and its regulatory function analysis

Whole-transcriptome sequencing identified 325 miRNAs shared between bnstm16-KO and WT plants, among which 123 were differentially expressed (79 upregulated and 44 downregulated). Hierarchical clustering classified these differentially expressed miRNAs into distinct co-expression modules, revealing condition-specific expression patterns and conserved regulatory trends under cold stress (Fig. S9).

GO enrichment analysis of miRNA target genes showed significant enrichment in lignin catabolism, phenylpropanoid metabolic flux, sucrose catabolism, and superoxide metabolism. KEGG enrichment further indicated that target genes were involved in zeatin biosynthesis, α-linolenic acid metabolism, sulfur metabolism, starch and sucrose metabolism, and phenylpropanoid biosynthesis, which are closely associated with hormone balance, energy supply, secondary metabolism, and stress tolerance (Fig. S10). These results indicate that BrSTM16 participates in plant growth, development, and stress adaptation by orchestrating multiple metabolic and regulatory pathways.

Transcriptome profiling revealed that BrSTM16-dependent differentially expressed genes were highly enriched in the phenylpropanoid biosynthesis pathway. We further characterized 15 core enzyme genes in this pathway, including PAL, CAD, CCR, and 4CL. Expression analysis showed that genes encoding PAL and F5H were significantly downregulated in bnstm16-KO plants, whereas 4CL was significantly upregulated, while other enzyme genes showed varied expression changes (Fig. 9). These observations indicate that BrSTM16 does not simply activate or repress the entire pathway, but modulates the phenylpropanoid pathway in a gene-specific and fine-tuned manner to maintain metabolic homeostasis.

Fig. 9.

Heatmap of phenylpropanoid biosynthesis pathway and related gene expression. Red and green respectively indicate the upregulation and downregulation of gene expression in brstm16-KO plants. The heatmap of differentially expressed genes is constructed based on gene expression levels (FPKM), with data rows standardized using the Z-score method

Further regulatory prediction showed that the 15 key enzyme genes in the phenylpropanoid pathway were targeted by 50 upstream regulators, among which 18 genes were potentially targeted by 23 differentially expressed miRNAs (Fig. 10). This indicates that the phenylpropanoid pathway is under robust post-transcriptional control by miRNAs, with one miRNA targeting multiple genes and individual genes regulated by multiple miRNAs. Notably, novel_miR_174 and novel_miR_229 were both downregulated under cold stress and jointly targeted multiple PAL genes, including BnA01g0036810, BnA02g0046410, BnA04g0179970, and BnC06g0749310. These two miRNAs may therefore contribute to cold tolerance by repressing PAL expression at the post-transcriptional level.

Fig. 10.

Phenylpropanoid metabolic pathway - miRNA interaction network and miRNA expression analysis. A Phenylpropanoid metabolism pathway gene and target miRNA regulatory network; (B) Heatmap of miRNA expression levels. The heatmap of differentially expressed genes is constructed based on gene expression levels (FPKM), and data rows are standardized using the Z-score method for clustering analysis. Red and yellow indicate significant upregulation and downregulation of gene expression in brstm16-KO, respectively. The darker the node color, the greater the number of target genes

In summary, BrSTM16 fine-tunes the phenylpropanoid biosynthesis pathway to maintain metabolic homeostasis under low-temperature stress. In the bnstm16-KO mutant, this balanced regulation is disrupted, leading to uncoordinated expression of PAL, C4H, 4CL, and other rate-limiting enzyme genes. This dysregulation reduces phenylpropanoid metabolic flux, impairs cell wall stability, weakens ROS-scavenging capacity, and disturbs carbon partitioning, ultimately compromising plant cold tolerance.

Subcellular localization of BrSTM16

To investigate intracellular localization of BrSTM16, the gene coding sequence was fused to C-terminal of GFP in pSuper1300 vector. Recombinant and control vectors were transiently expressed in tobacco leaves by Agrobacterium infiltration method. After 24 h dark incubation and 48 h normal culture, confocal microscopy analysis was performed. The fusion protein was specifically located in the nucleus, while the solitary GFP exhibited broad distribution inside the entire cell (Fig. 11). On the basis of above localization results, it can be concluded that BrSTM16 is a nucleus-targeted protein, which is in accordance with the prior functional prediction as a transcription factor.

Fig. 11.

Subcellular localization of BrSTM16 protein. GFP: blank control; BrSTM16-GFP: BrSTM16-GFP fusion protein; mCherry, nuclear localization protein; Scale bar: 50 μm

Predication of the BrSTM16 protein interaction network

Many important traits are governed by complex regulatory networks. To explore the potential functional context of BrSTM16, we inferred its protein–protein interaction network via the STRING database, using Brassicaceae species as the reference. The resulting network positions BrSTM16 as a central hub, indicating that BrSTM16 potentially functions as an essential regulatory hub mediating the interplay between meristem growth modulation and plant tolerance to cold stress. (Fig. 12A).

Fig. 12.

Identification, verification of BrSTM16 protein interactions and analysis of its expression at low temperature. A Protein interaction network of BrSTM16 in Brassica rapa. Nodes represent proteins, with the central node indicated in orange. Black lines represent interactions between nodes. In the interaction network, darker colors indicate greater importance of the corresponding proteins; (B) One-to-one verification of BrSTM16 positive clone yeast two-hybrid; (C- E) Expression of BrSTM16-interacting protein genes under low-temperature conditions. Statistical analysis was performed using two-way ANOVA followed by Tukey’s multiple comparison test. Data in the figure represent the average of three independent replicates. Error bars indicate the standard deviation (SD) of the three replicates. Statistical significance between treatments was indicated by different symbols: ns indicates p > 0.05, no significant difference; *, **, and *** indicate significant differences at the levels of p < 0.05, p < 0.01, and p < 0.001, respectively

The predicted interactions indicate several functional modules. First, BrSTM16 is connected to multiple homeobox transcription factors, which could reinforce its core regulatory role in meristem maintenance and stability under stress. Second, an interaction with the floral meristem identity regulator LFY1 suggests a potential link to flowering time control, possibly coordinating reproductive development with environmental stress signals. Notably, a predicted interaction with a PPM-type protein phosphatase (M4CJW1) points to a potential mechanism for stress signal perception. As PPM phosphatases can act as negative modulators of abscisic acid (ABA) signaling, this interaction might mediate ABA-dependent phosphorylation cascades that modulate BrSTM16 activity, stability, or the expression of its downstream targets.

Collectively, these predicted interactions outline a sophisticated regulatory module. This network would enable BrSTM16 to simultaneously influence meristematic differentiation and cellular cold stress responses, allowing Brassica rapa to balance necessary growth with the activation of protective mechanisms under low temperature conditions, thereby enhancing environmental adaptability.

Yeast two-hybrid screening for BrSTM16 interacting proteins

With the aim of identifying potential interacting proteins of BrSTM16, we conducted comprehensive screening experiments via the yeast two-hybrid system. A total of 49 candidate positive clones were obtained. Gene ontology analysis of the corresponding proteins revealed diverse functions, including enrichment in cellular transport, catabolism, signal transduction, genetic information processing, and metabolic pathways. Database annotation identified several promising interacting factors, including transcription factors (bHLH121, ZHD2, BEL1-like, PHL7), cold response proteins (COR78, CIPK7, LOS1), proteins related to meristem development (LEA, COV1), ubiquitously expressed proteins (NBR1, HAKAI), metabolic enzymes (AGM1, SAM-2), DNA/RNA-binding proteins (SMUBP-2, PTB1, PABP2), and proteins of unknown function (Table S10).

Six representative candidates were selected from the 49 preliminary proteins for pairwise Y2H validation. We prioritized candidates participating in cold stress response, apical meristem development, stress-mediated transcriptional regulation, and metabolic homeostasis. Specifically, the screened proteins cover core functional modules of plant cold resistance, including calcium signal transduction (CIPK7), cold stress protection and protein translation regulation (COR78, LOS1), meristem morphogenesis and stress-responsive transcription regulation (BLH1, PHL7), and stress-resistant metabolic regulation (SAM-2). These six proteins act as key regulators linking meristem development and cold acclimation, and are functionally matched with the potential biological roles of BrSTM16. Verification of these interactions provides a basis for elucidating the molecular mechanism of BrSTM16-mediated apical meristem cold tolerance and refining the regulatory network of cold adaptation in Brassica rapa.

Full-length coding sequences of these genes were cloned into the pGADT7 prey vector. Yeast strains co-transformed with pGBKT7-BrSTM16 and each recombinant prey plasmid grew well and displayed blue coloration on QDO/X/A selective plates, confirming specific protein–protein interactions (Fig. 12B). No colonies were observed in the negative control co-transformed with pGBKT7-BrSTM16 and empty pGADT7. These results demonstrate that BrSTM16 physically interacts with multiple regulatory factors involved in transcription, cold stress signaling, and cellular metabolism.

Expression analysis of genes encoding BrSTM16 interacting protein under cold stress

Given that interacting proteins are often co-expressed, we examined the expression of the six validated BrSTM16-interacting genes (COR78, CIPK7, LOS1, SAM2, PHL7, and BLH1) in response to cold stress. Briefly, CIPK7 is a core serine/threonine kinase in calcium signaling that decodes Ca²⁺ signals via interaction with CBL proteins and participates in multiple abiotic stress responses including cold, salt, and drought. BLH1, a BEL1-like homeodomain transcription factor, regulates meristem development, organ morphogenesis, and stress responses, often by interacting with KNOX family proteins. PHL7, an MYB family transcription factor, contributes to phosphate homeostasis, salt tolerance, and nutrient balance. COR78 is a well-known cold-induced protein that stabilizes cell membranes and proteins to alleviate cold injury. LOS1 encodes a eukaryotic translation elongation factor EF-2 essential for protein synthesis under low temperature. SAM-2 catalyzes the synthesis of S-adenosylmethionine (SAM), a major methyl donor involved in methylation, hormone metabolism, and stress adaptation, and contributes to membrane stability and antioxidant defense under cold stress. Tissue-specific expression analysis was conducted on these genes in leaves, roots and apical meristematic tissues of L7 and L99 transgenic lines at 3 h and 24 h intervals under 4 °C low-temperature conditions.(Fig. 12C-H).

Remarkable transcriptional activation of all six genes was triggered under cold stress conditions, accompanied by obvious tissue-specific differences and time-dependent expression characteristics throughout the experiment. Overall, greater expression accumulation was observed in apical growing regions relative to leaf and root tissues, and continuous low-temperature treatment further promoted their expression upregulation. Notably, in the growth cone of the cold-tolerant L7 cultivar after 24 h, the expression of COR78, CIPK7, SAM2, PHL7, and BLH1 was strongly upregulated, reaching levels 58-fold, 22-fold, 14-fold, 10-fold, and 7-fold higher than the control, respectively. In the growth cone of L99, LOS1 expression was prominently induced, showing a 13-fold increase after 24 h.

The strong and preferential co-upregulation of these interacting genes with BrSTM16 in the apical meristem strongly reinforces the functional link between BrSTM16 and its partners in this tissue. By interacting with calcium signaling kinases, cold-protective proteins, translation factors, metabolic enzymes, and meristem-related transcription factors, BrSTM16 forms a multi-component regulatory network in the growth cone that coordinately modulates cold signaling, protein homeostasis, stress metabolism, and meristem stability, thereby enhancing cold acclimation and freezing tolerance in Brassica rapa.

Discussion

Molecular characteristics and evolution of the BrSTM gene family

The plant growth cone consists of pluripotent stem cells that give rise to all aerial plant structures. In Arabidopsis, the homeodomain transcription factor WUSCHEL (WUS) maintains stem cell homeostasis via the CLAVATA (CLV) signaling pathway [49]. As a canonical KNOX-family transcription factor, STM is ubiquitously expressed in growth cone tissues but significantly suppressed in nascent organ primordia [50]. The subcellular localization and intercellular movement of STM proteins are indispensable for its biological function. In Arabidopsis, STM interacts with BLH proteins to form heterodimers, which facilitates STM nuclear import and sustains growth cone developmental activity [51]. In maize, the STM homolog KN1 traffics via plasmodesmata in an MPB2C chaperone-dependent manner. Nevertheless, the trafficking mechanism and interacting regulators of STM in Brassica rapa are still poorly characterized.

Genome wide identification characterized 31 BrSTM members in Brassica rapa. All BrSTM proteins contain a nuclear-localized Homeobox_KN domain, and Group I/II proteins possess additional POX domains related to ROS homeostasis. Group III proteins harbor intact KNOX1, KNOX2 and ELK domains responsible for meristem maintenance, protein autoregulation and protein interaction, respectively [52, 53]. Group III genes also have more conserved motifs and multi-exon structures for functional diversification [54]. Variable intron numbers among BrSTM genes reveal evolutionary intron turnover that contributes to family functional divergence. Notably, intron-deficient BrSTM7 and BrSTM16 are rapidly induced by abiotic stress, supporting the common perspective that genes with fewer introns exhibit quicker stress-responsive activation.

Phylogenetic relationships and gene family expansion

Based on phylogenetic tree reconstruction of STM proteins from six representative species, STM orthologs from Brassicaceae plants form a compact clade, indicating shared evolutionary ancestry (Fig. 1). Moreover, a greater number of collinear gene pairs was identified between Brassica rapa and Brassica rapa subsp. pekinensis compared with other species, corroborating their relatively recent evolutionary divergence (Fig. 2). The evolutionary expansion of the STM family in the genus Brassica is mainly caused by the combined effects of segmental and tandem duplications, common drivers of gene family growth and genome evolution [55]. Such duplications generate paralogs with similar functions, as seen with BrSTM15/16/18/25, which showed correlated expression patterns across tissues and stresses (Fig. S5). The conservation of gene function is tightly linked to the preservation of their transcriptional regulatory mechanisms [56].

Functional roles of BrSTM genes in development and stress response

Through a comprehensive analysis of promoter sequences, abundant regulatory cis-acting motifs responsive to photoperiod, hormonal signals, and adverse environmental stresses were identified in the promoters of BrSTM genes. Significantly, the promoters of BrSTM8 / 16 / 23 contain both low-temperature response (LTR) and meristem-specific elements (Fig. S3). This co-localization suggests these genes are poised to integrate cold signals with developmental programs in the growth cone, a key site for cold acclimation [57]. Previous studies have demonstrated that LTR elements do not function merely as isolated cold-responsive elements; rather, they act as central signal integrators, responding extensively to multiple signalling pathways including ethylene, ABA, salicylic acid, MeJA and oxidative stress. ABA signaling modules are widely involved in modulating abiotic stress tolerance by adjusting hormone sensitivity in plants [58]. They thus serve as key regulatory modules for the integration of various stress signals and hormonal signals in plants. This signal-integrating property of LTR elements also provides a plausible and compelling explanation for the molecular mechanism by which the BrSTM gene promoter, containing an LTR element, is able to respond to multiple stresses.

Consistent with a role in undifferentiated growth, several BrSTM genes, particularly BrSTM7, were highly expressed in callus tissue (Fig. S5A). Under cold stress, specific members like BrSTM7 / 11/ 16/ 22/ 23/ 28/ 30/ 31 were strongly induced, especially in the apical meristem of the less cold-tolerant cultivar L-99 (Figs. 4 and 5). Phylogenetically related genes (BrSTM7/16/22/23/28/30) often showed similar expression patterns, indicative of functional conservation within clades. BrSTM16 emerged as a key candidate due to its constitutive yet stress-inducible expression, with highest levels in the apical meristem. Its lower basal expression in the cold-tolerant, recessed-meristem cultivar L7 hints at a link between BrSTM16 expression levels and meristem morphology.

Functional validation of BrSTM16 in cold tolerance

Brassica rapa, Arabidopsis thaliana and Brassica napus all belong to the Brassicaceae family. Brassica rapa, Arabidopsis thaliana and Brassica napus have close phylogenetic relationships, and STM homologous genes possess relatively high sequence conservation across the three species. Such evolutionary conservation provides a credible prerequisite for cross-species functional analysis of BrSTM16. Dual genetic assays including Arabidopsis heterologous overexpression and Brassica napus CRISPR/Cas9 knockout validated the conserved positive role of BrSTM16 orthologs in regulating cold tolerance in Brassica napus. Interspecific differences in ploidy, apical meristem structure and downstream signals may cause slight functional divergence of BrSTM16, whereas its core function in cold acclimation remains intact. Similar to small RNA mediated modules modulating specialized metabolism, transcription factors regulate secondary metabolism to coordinate plant growth and stress adaptation [59].

Transcriptomic analysis further indicated that BrSTM16 enhances cold tolerance by modulating phenylpropanoid biosynthesis. BrSTM16 maintains low-temperature metabolic homeostasis via regulating phenylpropanoid flux, which is disrupted in bnstm16-KO mutants. Loss-of-function of BrSTM16 inhibits the expression of PAL, C4H and 4CL, hindering phenylpropanoid metabolism, impairing cell wall stability and ROS scavenging, disrupting carbon partitioning, and finally reducing cold tolerance [60, 61]. Previous safflower research has proven that PAL positively controls flavonoid accumulation and antioxidant capacity, which is consistent with the stress-induced flavonoid accumulation and dynamically regulated biosynthetic genes widely observed in plant stress responses, explaining the defective cold performance of bnstm16-KO rapeseed with disturbed PAL/4CL expression [62, 63]. BrSTM16 positively activates plant antioxidant and osmotic defense pathways under cold conditions. The synergistic defense modulation mitigates cold-induced oxidative damage and cellular homeostasis disorder, thereby improving plant cold resistance, which is consistent with conserved plant stress regulatory patterns [64].

Transgenic plants with constitutive overexpression of BrSTM16 exhibited markedly improved cold stress tolerance. According to earlier literature, numerous transcription factors regulate plant stress adaptation by modulating intracellular antioxidant metabolism. For example, the apple-derived transcription factor MdGH3 and the Arabidopsis transcription factor MYB96 confer enhanced resistance to chilling and drought stresses—partially through strengthening ROS detoxification cascades [65–67]. Unlike the above-mentioned single regulation mode, BrSTM16 can not only effectively activate antioxidant pathways but also simultaneously regulate the synthesis and accumulation of osmoprotectants, such findings indicate that BrSTM16 functions as a core regulatory element that coordinates multiple physiological defense pathways to collectively enhance plant tolerance to cold stress. Overall, BrSTM16 retains conserved roles in stress response and possesses distinctive molecular features associated with stress resistance. Accordingly, it provides valuable candidate genes and foundational theoretical insights for molecular breeding of cold-tolerant rapeseed and other crops.

Cold triggered BrSTM16 expression varies between two Brassica rapa cultivars. Cold sensitive L99 accumulates more apical BrSTM16 than cold tolerant L7, yet genetic evidence confirms BrSTM16 positively enhances cold resistance. Such expression discrepancy originates from genotype-determined apical morphology and cold-induced compensatory transcription. The exposed apical meristem of L99 undergoes severe freezing damage and triggers passive BrSTM16 upregulation for defense, whereas the compact apical structure of L7 confers intrinsic cold tolerance and reduces BrSTM16 expression. L99’s cold vulnerability is governed by unfavorable meristem structure and impaired cold signaling, which cannot be offset by BrSTM16 upregulation.

Beyond cultivar compensatory expression, meristem preferential BrSTM16 expression sustains basal cold adaptation of overwintering Brassica rapa. Controlled by meristem specific promoter motifs, BrSTM16 functions in cold-vulnerable apical tissues to maintain ROS homeostasis, osmotic balance, sugar metabolism and hormone signaling, thus protecting apical cells from freezing damage. Overall, differential BrSTM16 expression is a damage-induced compensatory response, and BrSTM16 acts as a meristem-specific positive regulator of cold tolerance in Brassica rapa.

Protein interaction network of BrSTM16

Subcellular localization confirmed BrSTM16 is a nuclear protein (Fig. 11). Yeast two-hybrid screening identified 49 potential interactors, and six were validated: the cold-responsive proteins COR78 and CIPK7, the metabolic enzyme SAM2, and the transcription factors PHL7, BLH1, and LOS1 (Fig. 12). These interactions place BrSTM16 at the nexus of stress signaling and developmental regulation. COR78 is a key marker gene for plants’ response to low temperatures. In Arabidopsis thaliana, it can rapidly sense cold stress through the cis-regulatory sequences at its 5’ end and achieve precise expression regulation at the transcriptional level. Its expression induction does not rely on the ABA signaling pathway, indicating that this gene has an independent and crucial regulatory pathway when plants cope with abiotic stress [68]. The interaction with BLH1 is particularly significant, as BLH-KNOX heterodimerization is essential for nuclear localization and function of KNOX proteins [69, 70]. BLH1 also interacts with ABA signaling components, potentially linking BrSTM16 to hormone-mediated stress responses [71, 72]. Similarly, PHL7 is involved in phosphate and ion homeostasis under stress [73]. The co-upregulation of these interacting genes with BrSTM16 in the cold-stressed growth cone supports their functional cooperation in a meristem-specific cold acclimation network.

Conclusion

This study conducted genome-wide identification and functional analysis of the STM gene family in Brassica rapa. Thirty-one BrSTM genes were identified with diversified structure, evolutionary expansion and stress- or development-related cis-elements. BrSTM16 contains cold-responsive and meristem-specific promoter elements, which links cold signaling to apical meristem regulation. Functional assays demonstrate the conserved positive role of BrSTM16 in cold tolerance across Brassica rapa. BrSTM16 modulates multiple downstream pathways including phenylpropanoid biosynthesis, sugar metabolism, ROS homeostasis and hormone signaling to mediate plant cold adaptation. BrSTM16 physically interacts with diverse cold-related proteins, metabolic enzymes and transcription factors. The coordinated expression of these interactors in apical meristems indicates that BrSTM16 participates in protein complexes to balance meristem development and cold stress adaptation(Fig. 13).

Fig. 13.

Stress response pattern of the Brassica rapa BrSTM16

Collectively, our findings demonstrate that BrSTM16 acts as an important positive regulator of cold tolerance in Brassica rapa by modulating antioxidant systems, osmotic adjustment, phenylpropanoid metabolism, and downstream gene networks. This work expands our understanding of the STM gene family’s roles in cold stress responses and provides a valuable candidate gene for the genetic improvement of cold resilience in Brassica rapa.

Abbreviations

Actin

Reference internal control gene

4CL

4-coumarate-CoA ligase

BLH

BELL-like homeodomain

BRAD

Brassica database

CAD

Cinnamyl alcohol dehydrogenase

CAT

Catalase

CCR

Cinnamoyl-CoA reductase

CDD

Conserved Domain Database

CRISPR/Cas9

Clustered Regularly Interspaced Short Palindromic Repeats/CRISPR-associated protein 9

DEGs

Differentially Expressed Genes

F5H

Ferulate 5-hydroxylase

FDR

False Discovery Rate

FPKM

Fragments Per Kilobase of transcript per Million mapped reads

GFP

Green Fluorescent Protein

Ka

Non-synonymous substitution rate

KEGG

Kyoto Encyclopedia of Genes and Genomes

Ks

Synonymous substitution rate

miRNA

MicroRNA

PAL

Phenylalanine ammonia-lyase

PCA

Principal Component Analysis

PEG-6000

Polyethylene Glycol 6000

RT-qPCR

Quantitative reverse transcription PCR

SPS

Sucrose phosphate synthase

TPS

Trehalose-6-phosphate synthase

Y2H

Yeast two-hybrid assay

Authors’ contributions

Yuanyuan Zhang: Conceptualization, Data curation, Formal analysis, Validation, Writing – original draft. Xiaolei Tao: Data curation, Formal analysis, Validation. Yuhong Zhao: Data curation, Formal analysis. Abbas Muhammad Fahim: Writing – review and editing. Shiyi Li: Investigation. Zhengnan Xu: Validation. Hao Sun: Software. Yifan Wang: Software. Lijun Liu: Supervision, Resources. Junyan Wu: Methodology, Project administration. Yuanyuan Pu: Methodology. Gang Yang: Methodology.Wangtian Wang: Methodology. Wancang Sun: Methodology. Li Ma: Conceptualization, Supervision, Funding acquisition, writing – review and editing.

Funding

This research was funded by the National Natural Science Foundation of China (32260519), Gansu Province Youth Talent (Team Project) (2026QNTD018), 2026 China Agricultural University Partner Assistance Joint Research Fund (GSAU-DKZY-2026-001), Gansu Provincial Joint Research Fund (24JRRA837), Star of Innovation for Postgraduate Students in Gansu Province in 2026 (2026CXZX-855), Gansu Agricultural University Young Faculty Mentor Fund Grant Program (GAU-QDFC-2025-01), the Fuxi talent project of Gansu Agricultural University (Gaufx-05Y01), the Lanzhou Youth Science and Technology Talent Innovation Program (2024-QN-45), the Science and Technology Program of Gansu Province (24ZDNA007), and the China Agriculture Research System of MOF and MARA (CARS-12-09).

Data availability

Data will be made available on request. The datasets analysed during the current study are available in the NCBI repository (SRR18959686, PRJNA1467764).

Declarations

Ethics approval and consent to participate

This article does not contain any research conducted by the author involving human participants or animals. All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by Gansu Agricultural University.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.