Differential Expression of MYB29 Homologs and Their Subfunctionalization in Glucosinolate Biosynthesis in Allotetraploid Brassica juncea

Abstract

Brassica juncea (L.) Coss. var. foliosa Bailey contains high glucosinolate (GSL) levels that define its flavor and defense properties. However, the regulatory mechanisms controlling GSL biosynthesis in Brassica crops remain unclear. Here, four MYB29 homologs were identified in allotetraploid Brassica juncea. These BjuMYB29 proteins localize to the nucleus and possess transcriptional activation activity. Evolutionary analysis suggests polyploidization-driven expansion of MYB genes contributed to GLS diversification in Brassica species. Expression profiling showed distinct spatiotemporal and herbivory-responsive patterns among BjuMYB29 homologs. Heterologous expression of BjuA03.MYB29 and BjuA10.MYB29 in Arabidopsis enhanced insect resistance via GSL accumulation. Although both homologs upregulate aliphatic GSL biosynthetic genes, they differentially regulate indolic GSLs, with BjuA03.MYB29 suppressing and BjuA10.MYB29 enhancing their accumulation, potentially through differential control of CYP79B2. These results reveal subfunctionalization among MYB29 homologs in GSL regulation. This functional diversification of MYB29 homologs offers novel targets for precision breeding of Brassica crops with customized GSL profiles to optimize pest resistance and nutritional quality.

1. Introduction

The order Capparales includes the model plant Arabidopsis thaliana as well as economically important Brassica crops, such as Chinese cabbage, turnip, broccoli, cauliflower, cabbage, and mustard, many of which are rich in glucosinolates (GSLs) [1,2]. Glucosinolates, a diverse class of sulphur- and nitrogen-containing secondary metabolites, are a defining biochemical feature of the order Capparales. Structurally, GSLs exhibit remarkable diversity, with approximately 200 naturally occurring variants identified in plants [3,4,5]. This structural variation arises from their biosynthesis from eight different precursor amino acids, followed by extensive side-chain modifications. Based on their amino acid origin, GSLs are classified into three major groups: aliphatic, indolic, and aromatic GSLs [6,7].

GSLs are sulfur-rich secondary metabolites that play multifaceted roles in plant-environment interactions and human health. As crucial defense compounds, they provide protection against herbivores and pathogens while also mediating responses to abiotic stresses [8,9]. Beyond their ecological functions, GSLs contribute to the distinctive flavors of Brassica vegetables and possess documented biomedical value, particularly for their potential cancer-preventive properties [10,11]. The complete GSL biosynthetic pathway has been extensively characterized in A. thaliana, where researchers have identified not only the enzymatic genes responsible for side-chain elongation, core structure formation, and side-chain modification, along with transporters of biosynthetic intermediates [2,7,12,13], but also the key transcriptional regulators. Central to this regulatory network are six subgroup-12 R2R3-MYB transcription factors that positively regulate GSL biosynthesis [2,14]. Specifically, MYB28, MYB29, and MYB76 activate aliphatic GSL production [15,16,17,18], while MYB34, MYB51, and MYB122 control indolic GSL biosynthesis [19,20]. This well-defined regulatory framework in Arabidopsis provides fundamental insights into the molecular mechanisms controlling GSL accumulation, serving as an essential reference for investigating the more complex regulatory systems in polyploid Brassica crops.

Brassicaceae crops provide essential vegetables, oils, and condiments worldwide, with GSL content being a key quality trait due to its impact on flavor, defense, and nutritional value [21,22,23,24]. While GSL-related genes are generally conserved with their Arabidopsis orthologs, the diversity in GSL composition across Brassica species implies the existence of complex, lineage-specific regulatory mechanisms [21,22,25]. Furthermore, although genome-wide analyses have identified numerous GSL-related genes and revealed a significant expansion of the R2R3-MYB family to 55 conserved members due to polyploidization [14,26,27,28,29,30,31], the functions of most of these transcription factors remain uncharacterized. This knowledge gap is particularly prominent for MYB29 homologs in the allotetraploid Brassica juncea, where the functional divergence and regulatory specificity among multiple paralogs are largely unknown. Elucidating the roles of these expanded MYB29 copies is crucial to deciphering how polyploidy has reshaped GSL regulatory networks and offers new opportunities for improving stress resistance and quality traits in this important crop.

In this study, we identified and characterized four MYB29 homologs (designated as BjuMYB29s) from Brassica juncea (L.) Coss. var. foliosa Bailey, an economically significant vegetable crop. Integrating evolutionary, expression, transgenic, and functional evidence, we demonstrated that MYB29 homologs in B. juncea exhibit differential expression and have undergone subfunctionalization in the regulation of GSL biosynthesis. Our findings suggest that BjuMYB29s represent promising molecular targets for metabolic engineering approaches aimed at enhancing both pest resistance and nutritional value in B. juncea cultivars.

2. Materials and Methods

2.1. Plant Material and Growth Conditions

The plant material used in this study was the Brassica juncea cultivar ‘Shuidong’ (classified as B. juncea (L.) Coss. var. foliosa Bailey), a leaf-type mustard. For cultivation, plants were maintained in controlled growth chambers set to a 16-h light/8-h dark photoperiod and day/night temperatures of 24 °C and 22 °C, respectively. All A. thaliana plants used in this study were of the Col-0 ecotype and grown under controlled conditions (22 °C day/18 °C night) with identical light/dark cycles. Both species were grown in a standardized growth medium composed of grass charcoal, vermiculite, and perlite (3:2:1 ratio by volume).

2.2. Gene Identification and Phylogenetic Analysis

Homologs of AtMYB29 in the B. juncea genome were identified through BLAST (SequenceServer 2.0.0) searches against the Brassica Database (BARD; http://www.brassicadb.cn/, accessed on 9 March 2022). Gene-specific primers were designed to amplify both genomic DNA and coding sequences of putative BjuMYB29 genes. PCR amplification products were cloned into the pCE2-TA vector (Yeasen Biotechnology, Shanghai, China) and verified by Sanger sequencing. Sequence alignments were performed using DNAMAN 9.0 software (Lynnon Biosoft, San Ramon, CA, USA) with default parameters.

Protein sequences of R2R3-MYB subgroup-12 members from four Brassica species (B. juncea, B. rapa, B. oleracea, and B. nigra) were retrieved and aligned. Phylogenetic reconstruction was performed using MEGA 11.0 software [32] with the neighbor-joining method (bootstrap = 1000 replicates). The resulting phylogenetic tree was visualized and annotated using the TVBOT online platform (https://www.chiplot.online/tvbot.html, accessed on 5 June 2024) [33].

2.3. Subcellular Localization Analysis

The coding sequences of BjuMYB29s were PCR-amplified and cloned into the pFGC-eGFP vector to generate CaMV35S promoter-driven GFP fusion constructs. Both the empty pFGC-eGFP control vector and recombinant pFGC-CaMV35S:BjuMYB29s-GFP constructs were transformed into Agrobacterium tumefaciens strain GV3101. For transient expression assays, young leaves of 4-week-old transgenic tobacco (Nicotiana benthamiana) plants stably expressing the nuclear marker H2B-RFP were infiltrated with Agrobacterium suspensions using needleless syringes. After 48 h of incubation under dark conditions, subcellular localization was examined using a confocal laser scanning microscope (Leica SP8 MP; Leica Microsystems GmbH, Wetzlar, Germany). Primer sequences used for cloning are provided in Table S1.

2.4. Transcription Activation Assay

The coding sequences (CDS) of BjuMYB29s were integrated into pGBKT7 vector using EcoRI and BamHI restriction enzymes to generate the bait plasmids pGBKT7-BjuMYB29s. The empty pGBKT7 vector was used as a negative control, and the pGBKT7-AD vector was employed as a positive control. All constructed plasmids were then transformed into yeast strain AH109. The transformed yeast cells were spotted onto synthetic dropout (SD) selection media lacking tryptophan (SD/-Trp) and SD medium lacking tryptophan, adenine, and histidine (SD/-Trp-Ade-His) at 28 °C for 3–4 days to test their transcriptional activity. Primers used for transcription activation assay are listed in Table S1.

2.5. Vector Construction and Stable Plant Transformation

To generate the overexpression construct, the CDS of BjuA03.MYB29 and BjuA10.MYB29 were cloned into the SmaI/XbaI sites of the binary vector pBI121 under the control of the CaMV35S promoter. The promoter sequence upstream of the ATG codon of BjuA03.MYB29 and BjuA10.MYB29 were amplified with gene-specific primers (Table S1) and cloned into pBI101 vector flanking the GUS reporter gene to create the fusion construct. The resulting recombinant plasmids were verified by sequencing and subsequently transformed into Agrobacterium tumefaciens strain GV3101. For stable transformation, A. thaliana (Col-0) plants were transformed using the floral dip method, as this species provides a conserved GSL pathway background that is ideal for heterologous functional studies. Transgenic seedlings were selected on Murashige and Skoog (MS) medium containing 75 mg/L kanamycin. Positive transformants were confirmed by PCR amplification of the transgene. To assess heterologous expression levels, RT-qPCR analysis was performed using specific primers (Table S1). Homozygous T₃ or T₄ transgenic lines were used for all subsequent functional analyses.

2.6. Expression Profiling of BjuMYB29s in Brassica juncea

To characterize the spatiotemporal expression patterns of BjuMYB29s, we performed quantitative analysis across multiple tissues of the B. juncea inbred line ‘znxlh-3’. Samples were collected from: (1) 7-day-old seedlings (cotyledons and hypocotyls), (2) 4-week-old seedlings (leaves and roots), and (3) bolting-stage plants (stems [node and internode], inflorescences, and young siliques). All tissues were immediately frozen in liquid nitrogen and stored at −80 °C. Quantitative reverse transcription PCR (RT-qPCR) was performed using gene-specific primers (Table S1) on an ABI 7500 FAST Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with Taq Pro Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China). The B. juncea Actin gene served as the internal control for normalization. Relative expression levels were calculated using the 2^−ΔΔCt method. Three independent biological replicates, each with three technical replicates, were analyzed for statistical reliability. Transgenic Arabidopsis lines ProBjuA03.MYB29:GUS and ProBjuA10.MYB29:GUS were selected for GUS staining at the cotyledon stage, seedling stage, and in stems, leaves, internodes, inflorescences, and siliques of flowering plants.

To analyze BjuMYB29s response to insect herbivory, four-week-old uniform seedlings of mustard variety ‘znxlh-3’ were used. Following 5 h of pre-starvation, third-instar beet armyworm (Spodoptera exigua Hübner) larvae were placed on the plants (three per seedling). Leaf samples were collected at 0, 1, 6, 12, 24, 36, and 72 h after feeding, and BjuMYB29s expression was quantified via qRT-PCR (Table S1). GUS staining was conducted on leaves of ProBjuA10.MYB29:GUS transgenic Arabidopsis before and after 36 h of beet armyworm infestation to visually examine the herbivory-induced response of BjuA10.MYB29.

2.7. Insect Feeding Experiment

To evaluate the effect of BjuA03.MYB29 and BjuA10.MYB29 overexpression on insect resistance, we performed feeding bioassays using four-week-old transgenic Arabidopsis plants and second-instar beet armyworm larvae. For each transgenic line, a minimum of ten plants were randomly selected for analysis. Prior to the experiment, larvae were starved for 5 h to standardize their feeding behavior. Three pre-starved larvae were then carefully placed at the center of each rosette. After seven days of feeding under controlled conditions, larval weights were measured (n = 20 per genotype), and their morphological condition was documented using stereomicroscopy (Leica M125 C; Leica Microsystems GmbH, Wetzlar, Germany). To ensure experimental reproducibility, the entire bioassay was conducted twice as independent biological replicates for each transgenic line. Data were analyzed to compare larval growth performance across different genotypes.

2.8. HPLC Analysis of Glucosinolate Content

The extraction, purification, and analytical determination of GSLs were conducted based on the methods described by Cai et al. [34] and Tao et al. [35], with minor adjustment. Briefly, tissues underwent sequential hot-water extraction, followed by purification using DEAE-Sephadex A-25 columns (Pharmacia, now Cytiva; Uppsala, Sweden) and enzymatic desulfation with sulfatase. Desulfoglucosinolates were analyzed by reversed-phase high-performance liquid chromatography (HPLC). Detailed HPLC parameters, including column specifications, mobile phase composition, gradient program, and detection wavelength, are provided in Supplementary Methods S1. Glucosinolate quantification was performed using sinigrin as an internal standard, with concentrations expressed as μmol per gram fresh weight (μmol/g FW).

2.9. Dual-Luciferase Reporter Assay

The CDS of BjuA03.MYB29 and BjuA10.MYB29 were cloned into the pGreenII 62-SK vector (BioVector NTCC Inc., Beijing, China) to generate effector constructs, with the empty pGreenII 62-SK vector serving as negative control. Promoter regions (~2000 bp upstream of the translation start site) of BjuA03.MAM3, BjuB02.MAM2, BjuB03.SOT16, BjuB05.SOT16, BjuA03.AOP2 and BjuB05.AOP2 were amplified and inserted into the pGreenII 0800-LUC reporter vector. All constructs were transformed into Agrobacterium tumefaciens GV3101 (pSoup) strain. For transient expression assays, bacterial suspensions containing effector and reporter constructs (1:1 ratio) were co-infiltrated into mature Nicotiana benthamiana leaves. After 48-h incubation in darkness, luminescence was detected following application of 0.3 mg/mL luciferin potassium salt solution using a CCD imaging system. Quantitative analysis was performed by measuring the LUC/REN ratio via RT-qPCR using infiltrated leaf samples.

2.10. Statistical Analysis

All data were analyzed and visualized using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). Statistical significance was determined by Student’s t-test in SPSS 11.5 (IBM, Chicago, IL, USA), with significance levels set at * p < 0.05 and ** p < 0.01 for indicating significant and highly significant differences, respectively.

3. Results

3.1. Identification and Characterization of Four BjuMYB29 Homologs in Brassica juncea

Bioinformatic analysis of the Brassica Database (BRAD) revealed four MYB29 homologs in the B. juncea genome, designated as BjuA03.MYB29, BjuA10.MYB29, BjuB02.MYB29, and BjuB08.MYB29, each located on distinct chromosomes. Using B. juncea (L.) Coss. var. foliosa Bailey as experimental material, we successfully isolated the full-length coding sequences of these BjuMYB29 genes through PCR amplification with gene-specific primers designed based on BRAD reference sequences. Sequence characterization uncovered an interesting feature of BjuA10.MYB29, which produces two alternative transcript variants. The predominant variant, BjuA10.MYB29.1, maintains the canonical three-exon/two-intron structure conserved among all other BjuMYB29 homologs. In contrast, the alternative splice variant BjuA10.MYB29.2 exhibits a simplified two-exon/one-intron organization and encodes a truncated 237-amino acid protein (711 bp), while the other BjuMYB29 genes encode full-length proteins ranging in length from 330 to 359 amino acids (990–1077 bp) (Table S2).

Sequence alignment analysis demonstrated that BjuMYB29 proteins share 46–74% amino acid identity with A. thaliana MYB29, with BjuA10.MYB29.1 showing the highest conservation (74%) and the truncated BjuA10.MYB29.2 variant exhibiting the lowest similarity (46%) (Table S3). Multiple sequence alignment revealed that all full-length BjuMYB29 proteins (except the truncated BjuA10.MYB29.2) contain two evolutionarily conserved domains: (1) the canonical R2R3-MYB DNA-binding domain and (2) a predicted nuclear localization signal (NLS) at their N-termini (Figure 1A). Although the N-terminal regions of these proteins show high conservation with Arabidopsis subgroup-12 R2R3-MYB transcription factors (AtMYB28, AtMYB29, and AtMYB76), which are established master regulators of aliphatic GSL biosynthesis, their C-terminal domains display significant divergence, indicating potential functional differentiation.

Phylogenetic reconstruction of R2R3-MYB subgroup-12 members across Brassicaceae species revealed several important evolutionary trends (Figure 1B). The evolutionary analysis revealed a conserved yet dynamically expanded MYB regulatory network in Brassica species. Orthologs of MYB28, MYB29, MYB34, MYB51, and MYB122 were not only conserved but also expanded in copy number across Brassica genomes, forming distinct monophyletic clades. In contrast, MYB76 homologs were conspicuously absent from all examined Brassica species. The MYB29 family, for instance, expanded from a single copy in A. thaliana to two copies in diploid Brassica species, culminating in four homologs in the allotetraploid B. juncea, a pattern of copy number increase that was similarly observed for the MYB28, MYB34, MYB51, and MYB122 families. The BjuMYB29 proteins displayed clear phylogenetic divergence, separating into two major clusters: (1) BjuA03.MYB29 and BjuB08.MYB29, and (2) BjuA10.MYB29.1, BjuA10.MYB29.2, and BjuB02.MYB29. This phylogenetic distribution pattern is highly consistent with sequence conservation characteristics, and the polyploidization events during Brassica evolution suggest the potential for broader functional diversification. The expansion and divergence of MYB29 homologs in B. juncea may represent an evolutionary adaptation to enhance regulatory complexity in GSL biosynthesis pathways.

3.2. Subcellular Localization and Transcriptional Activity of BjuMYB29 Proteins

To characterize the subcellular localization properties of BjuMYB29 proteins, we generated eGFP fusion constructs and transiently expressed them in tobacco epidermal cells (Figure 2A). Confocal microscopy analysis revealed distinct localization patterns among the different BjuMYB29 members. Both BjuA03.MYB29 and BjuB08.MYB29 displayed exclusive nuclear localization, as evidenced by strong GFP fluorescence in the nucleus that completely overlapped with the nuclear marker H2B-RFP, with minimal signal detected in the cytoplasm. In contrast, BjuA10.MYB29.1, BjuA10.MYB29.2, and BjuB02.MYB29 exhibited dual localization, showing significant fluorescence signals in both nuclear and cytoplasmic compartments. These differential localization patterns suggest potential functional specialization among BjuMYB29 family members, with some acting as strict nuclear transcription factors while others may have additional cytoplasmic functions.

To investigate the transcriptional activation potential of BjuMYB29 proteins, we conducted transcription activation assays using pGBKT7/BjuMYB29s fusion constructs transformed into AH109 yeast strains (Figure 2B). The experimental design included two controls, empty pGBKT7 vector as negative control and pGBKT7/GAL4 as positive control. Growth assays on selective SD/-Trp/-His/-Ade medium demonstrated that yeast cells expressing either pGBKT7/BjuMYB29s constructs, or the positive control (pGBKT7/GAL4) exhibited robust growth (Figure 2C). These findings provide direct experimental evidence that all BjuMYB29 proteins function as transcriptional activators and possess intrinsic transactivation activity in nucleus.

3.3. Spatiotemporal Expression Patterns of BjuMYB29s During Brassica juncea Development

Comprehensive expression profiling of BjuMYB29s across various tissues revealed distinct spatiotemporal regulation patterns (Figure 3). RT-qPCR analysis demonstrated that BjuA03.MYB29 and BjuB08.MYB29 exhibited broad expression profiles with particularly high transcript levels in roots and stems, suggesting their potential roles in vegetative growth and development. In contrast, BjuA10.MYB29.1 and BjuB02.MYB29 showed more restricted expression patterns, predominantly accumulating in hypocotyls, stems and inflorescences, while BjuA10.MYB29.2 displayed exclusive expression in inflorescence tissues, indicating specialized functions in reproductive development. These expression patterns showed a strong correlation with phylogenetic relationships, where evolutionarily closely related homologs shared similar tissue-specific expression profiles, while those more distantly related exhibited progressively divergent expression patterns (Figure 1B). Notably, BjuA03.MYB29 consistently showed the highest expression levels among all members, a finding corroborated by stronger GUS activity in ProBjuA03.MYB29:GUS transgenic Arabidopsis compared to ProBjuA10.MYB29:GUS plants (Figure 4). In addition, in the leaves, promoter activity of BjuA03.MYB29 was observed in the mid-vein, primary and secondary veins, as well as toward the leaf margins. In contrast, the promoter activity of BjuA10.MYB29 was predominantly localized to the mid-vein, primary and secondary veins (Figure 4D,K). The observed differential expression patterns suggest that BjuMYB29s have undergone functional diversification through the acquisition of distinct regulatory mechanisms, potentially enabling specialized roles in different aspects of plant development in B. juncea.

3.4. Insect Herbivory Induces Differential Expression of BjuMYB29s

To characterize the defense-related functions of BjuMYB29s, we analyzed their expression patterns following beet armyworm (Spodoptera exigua) infestation in B. juncea. Third-instar larvae were allowed to feed on mature leaves, with tissue samples collected at 0, 1, 6, 12, 36, and 72 h post-infestation for RT-qPCR analysis (Figure 5A). All four BjuMYB29 homologs showed significant induction by insect feeding, though with distinct temporal expression profiles. BjuA10.MYB29 exhibited the most robust and sustained response, showing continuous upregulation throughout the 72-h experimental period. In contrast, BjuA03.MYB29 and BjuB08.MYB29 displayed biphasic induction patterns with peaks at 6 and 72 h, while BjuB02.MYB29 showed strong activation at 6, 12, and 72 h post-infestation.

To validate these findings, we examined the response of BjuA10.MYB29 using ProBjuA10.MYB29-GUS transgenic Arabidopsis plants. Histochemical staining revealed substantially enhanced GUS activity in the leaves, particularly in the veins, after 36 h of feeding by second-instar larvae compared to untreated controls (Figure 5B), confirming the insect-induced activation observed in RT-qPCR analysis. The consistent induction patterns across both experimental systems strongly suggest that BjuA10.MYB29 plays a central role in coordinating B. juncea’s defense responses against insect herbivores, with other BjuMYB29 members potentially contributing to specific temporal phases of the defense cascade.

3.5. Overexpression of BjuA03.MYB29 and BjuA10.MYB29 Confers Enhanced Resistance to Beet Armyworm in Arabidopsis

To investigate the defensive functions of BjuMYB29s, we selected BjuA03.MYB29 and BjuA10.MYB29 as representatives based on their sequence and protein characteristics, and generated transgenic Arabidopsis lines constitutively expressing these homologs under the control of the CaMV35S promoter (BjuMYB29s-OX). To facilitate the comparison and assessment of overexpression levels in heterologous transgenic plants, a primer pair (denoted as MYB29 in Table S1) was designed to target a highly conserved region shared by AtMYB29, BjuA03.MYB29, and BjuA10.MYB29 for RT-qPCR analysis. Quantitative RT-qPCR analysis verified successful transgene expressions in multiple independent lines (Figure 6A), revealing variable expression levels among transformants.

Insect bioassays demonstrated that beet armyworm larvae feeding on BjuMYB29s-OX plants showed significantly impaired growth compared to wild-type controls (Figure 6B). Despite occasional experimental variability due to larval escape, reproducible results were obtained from at least one transgenic line per BjuMYB29 homolog. Importantly, we observed a dose-dependent relationship between transgene expression levels and larval growth inhibition, with higher-expressing lines (e.g., BjuA03.MYB29-OX#4 vs. #5) exhibiting more pronounced anti-herbivore effects. After seven days of feeding, larvae reared on high-expressing BjuMYB29s-OX plants (BjuA03.MYB29-OX#4 and BjuA10.MYB29-OX#12) displayed visibly reduced body size compared to controls (Figure 6C). These findings provide compelling evidence that BjuMYB29 proteins function as positive regulators of plant defense against insect herbivory.

3.6. BjuA03.MYB29 and BjuA10.MYB29 Enhance Glucosinolate Production via Activating the Expression of GSL Biosynthetic Genes

To elucidate the molecular mechanisms underlying BjuMYB29-mediated insect resistance, we conducted comprehensive analyses of GSL profiles and biosynthetic gene expression in representative BjuA03.MYB29-OX and BjuA10.MYB29-OX transgenic Arabidopsis lines. HPLC quantification revealed that both BjuA03.MYB29-OX and BjuA10.MYB29-OX lines exhibited significant increases in aliphatic GSLs, particularly 4-methylsulfinylbutyl GSL (4MSOB), compared to wild-type plants (Figure 7A,C). A clear gene-dosage effect was evident across transgenic lines: BjuA03.MYB29-OX line #4, which exhibited higher transgene expression, showed correspondingly greater aliphatic GSL accumulation than line #5. Similarly, the two BjuA10.MYB29-OX lines (#9 and #12) displayed GSL levels that aligned with their respective expression levels. This consistent correlation between transgene expression and metabolite accumulation underscores the dose-dependent effect of BjuMYB29s on aliphatic GSL biosynthesis. In contrast, BjuA03.MYB29 and BjuA10.MYB29 differentially regulated indolic GSL accumulation. While BjuA10.MYB29-OX lines showed elevated levels of indol-3-ylmethyl GSL (I3M), BjuA03.MYB29-OX lines displayed either reduced or unchanged I3M content (Figure 7B,D). These metabolic changes suggest that while BjuMYB29s universally enhances aliphatic GSL biosynthesis, their effects on indolic GSL are isoform-specific. The robust accumulation of aliphatic GSLs, especially 4MSOB, likely represents the principal mechanism underlying the enhanced insect resistance observed in transgenic plants.

To further elucidate the transcriptional regulation of GSL biosynthesis by BjuMYB29s, we analyzed the expression profiles of key GSL biosynthetic genes in transgenic Arabidopsis lines. RT-qPCR analysis revealed distinct regulatory patterns between BjuA03.MYB29 and BjuA10.MYB29 (Figure 8). BjuA10.MYB29 functioned as a broad-spectrum activator, significantly upregulating all tested genes involved in both aliphatic (MAM1, MAM3, CYP79F1, CYP83A1, SOT16, SOT18, and AOP2) and indolic (CYP79B2) GSL biosynthesis. In contrast, BjuA03.MYB29 exhibited selective activation, specifically enhancing only MAM1, MAM3, SOT16, and AOP2 expressions while maintaining wild-type levels of CYP79F1, CYP83A1, SOT18, and CYP79B2.

These differential gene expression patterns correspond with the observed metabolic profiles, where both transcription factors increased aliphatic GSLs but showed distinct effects on indolic GSL accumulation. The comprehensive activation of biosynthetic genes by BjuA10.MYB29 explains its broader impact on GSL profiles, while the more targeted regulation by BjuA03.MYB29 accounts for its specific enhancement of aliphatic GSL production. These results demonstrate that BjuMYB29 family members have evolved distinct regulatory specificities in controlling GSL biosynthesis, with BjuA10.MYB29 acting as a global regulator and BjuA03.MYB29 functioning as a selective activator of specific branches in the aliphatic GSL pathway.

3.7. Direct Transcriptional Regulation of Glucosinolate Biosynthetic Genes by BjuA03.MYB29 and BjuA10.MYB29

Our findings demonstrate that heterologous expression of BjuMYB29s in Arabidopsis enhances insect resistance through glucosinolate accumulation. This effect appears to be mediated by transcriptional activation of biosynthetic genes, suggesting a direct regulatory relationship between BjuMYB29 proteins and glucosinolate pathway components that warrants further investigation.

As an allotetraploid species rich in aliphatic GSL, B. juncea possesses a complex glucosinolate biosynthetic network. We focused our analysis on key genes representing three critical biosynthetic phases: MAM family genes for side-chain elongation, SOT family genes for core structure formation, and AOP family genes for side-chain modification. Phylogenetic analysis identified an expanded gene complement in B. juncea, including 9 MAM, 10 SOT, and 9 AOP homologs (Figure S1). Promoter analysis revealed MYB recognition sites (CCGTTG) and MYB binding sites (CAACTG) in several candidate genes (BjuB02.MAM3, BjuB03.SOT16, BjuB05.SOT16, BjuA03.AOP2, and BjuB05.AOP2), while BjuA03.MAM2-1 contained only MYB-related elements (Figure S2). These six genes were selected for dual-luciferase assays to test direct regulation by BjuA03.MYB29 and BjuA10.MYB29.

We generated effector vectors expressing BjuA03.MYB29 or BjuA10.MYB29 and reporter vectors containing the LUC gene driven by six selected promoters (Figure S3). Dual-luciferase assays in tobacco showed significantly enhanced LUC activity for all reporter constructs except BjuA03.MAM2-1Pro when co-expressed with either BjuA10.MYB29 (Figure 9) or BjuA03.MYB29 (Figure S4). Consistent results were obtained through LUC/REN ratio quantification by RT-qPCR (Figure 9 and Figure S4). These data demonstrate that BjuA03.MYB29 and BjuA10.MYB29 directly activate transcription from promoters containing MYB recognition sites and binding sites (BjuB02.MAM3, BjuB03.SOT16, BjuB05.SOT16, BjuA03.AOP2, and BjuB05.AOP2), but not from BjuA03.MAM2-1 which lacks these elements.

4. Discussion

4.1. Evolutionary Expansion of MYB Regulators in Brassica Species

Advances in genomic resources have revolutionized our understanding of gene family evolution in Brassica species. While early studies relying on Arabidopsis sequences reported limited gene copies, current genome assemblies reveal substantially expanded MYB families. For instance, whereas only four MYB28 homologs were previously identified in B. juncea [36], we now recognize six. Leveraging these improved resources, we successfully identified and experimentally confirmed four MYB29 homologs in B. juncea: BjuA03.MYB29 and BjuA10.MYB29 from the A subgenome, and BjuB02.MYB29 and BjuB08.MYB29 from the B subgenome.

This gene family expansion originated from an ancestral whole-genome triplication event ~13–17 MYA, which established multiple paralogous copies in diploid Brassica species [37]. Subsequent hybridization and genomic reorganization during allopolyploid formation further amplified gene copy numbers [38,39], resulting in pronounced expansion of GSL biosynthesis-associated genes compared to Arabidopsis [14,26,27,28,29,30,31]. Our systematic analysis reveals remarkable evolutionary divergence among these expanded MYB genes. Most strikingly, despite their close phylogenetic relationship in Arabidopsis, MYB29 and MYB76 experienced dramatically different fates—MYB29 expanded to four homologs in B. juncea while MYB76 was completely lost. This pattern contrasts with the consistent retention of MYB28, MYB34, and MYB51 homologs. The differential evolutionary outcome between these closely related genes suggests that MYB29 underwent functional specialization that favored its retention, while MYB76 became dispensable in the Brassica lineage.

These findings collectively demonstrate that the differential retention and duplication of MYB genes during polyploid evolution have fundamentally shaped the diversity of GSL biosynthesis pathways in Brassica species [21,22,23,24,25]. The expanded MYB29 family provides a genetic foundation for regulatory innovation in GSL biosynthesis, representing a key evolutionary adaptation that enables fine-tuned control over specialized metabolic pathways.

4.2. Expression Divergence of MYB29 Genes in Allotetraploid Brassica juncea

This study provides compelling evidence for expression divergence among MYB29 homologs in allotetraploid B. juncea, offering new insights into the evolutionary trajectory of regulatory genes following polyploidization. While MYB29 is a key GSL biosynthesis regulator alongside MYB28, its expression dynamics in polyploid Brassica species remain poorly characterized, with previous studies limited to AtMYB29 and partial characterization of a single homolog in B. rapa [16,40]. In contrast, MYB28 homologs in B. juncea have been well documented, showing A-subgenome-specific expression bias that disproportionately contributes to aliphatic GSL biosynthesis [36].

We found that the four BjuMYB29 homologs exhibit distinct spatiotemporal expression profiles (Figure 3 and Figure 4), with tissue-preferential expression in stems or roots. This divergence likely originated from the ancient WGT event, as homologs from the same fractionated subgenome showed greater expression similarity than those from the same basic subgenome. Unlike BjuMYB28 homologs, which are predominantly expressed in the leaf lamina [36], BjuA03.MYB29 and BjuA10.MYB29 were highly expressed in leaf veins, suggesting that MYB28 and MYB29 may orchestrate different aspects of GSL distribution within leaves, potentially contributing to specialized defense strategies in different leaf tissues.

Under stress conditions, the BjuMYB29 homologs further displayed differential response dynamics. Consistent with homoeolog behavior in other polyploids [41], they showed varied induction patterns upon insect herbivory. Transcriptomic data indicated that larval feeding specifically upregulates BjuA10.MYB29 [29], which was corroborated by our data (Figure 5). BjuA10.MYB29 responded with rapid, sustained induction, implying a primary role in defense initiation, while the other three homologs exhibited delayed or intermittent induction, suggesting complementary functions.

These expression patterns reflect a hallmark of polyploid genomes. Homoeologous genes often show organ-specific expression and differential stress responses [42,43], and polyploidization broadly reshapes gene expression patterns [44,45,46,47]. We propose that this regulatory divergence has facilitated the evolution of a layered defense system in B. juncea, fine-tuning glucosinolate-mediated resistance through spatiotemporal specialization of homologous genes. This functional diversification beyond genetic redundancy underscores the adaptive innovation enabled by polyploidy in Brassica species.

In addition to transcriptional divergence, alternative splicing (AS) provides another mechanism for functional specialization among homologous genes [48,49]. In this study, we identified two AS isoforms of BjuA10.MYB29, including a truncated variant BjuA10.MYB29.2 that shows specific expression in inflorescences. This finding suggests that AS may further enhance the functional differentiation of MYB29 homologs in B. juncea.

4.3. Subfunctionalization of BjuMYB29s in the Regulation of Glucosinolate Biosynthesis

Polyploidy represents a major evolutionary mechanism driving speciation and phenotypic diversification in plants [50,51]. In allopolyploids, duplicated genes typically undergo distinct evolutionary trajectories—some are lost or silenced, while others develop functional specialization through neofunctionalization or subfunctionalization [52,53]. A particularly intriguing case in Brassica species involves the loss of MYB76, one of the three core transcription factors regulating aliphatic GSL biosynthesis in Arabidopsis. While MYB28 and MYB29 were retained and expanded, MYB76 was completely lost during Brassica evolution. The evolutionary drivers behind this specific gene loss, and whether and how its functions were compensated, have remained unexplored. Our study provides compelling evidence that subfunctionalization among the four MYB29 homologs in B. juncea may effectively compensate, at least partially, for the loss of MYB76. Specifically, BjuA10.MYB29 appears to have inherited the distinctive regulatory function of Arabidopsis HAG2/MYB76, enhancing both aliphatic and indolic GSL accumulation [16]. This functional compensation represents an elegant evolutionary solution where the regulatory portfolio of a lost gene has been assimilated by its expanded homologs. Meanwhile, BjuA03.MYB29 maintains the characteristic HAG3/MYB29 function, primarily enhancing aliphatic GSLs while suppressing indolic GSLs [16].

This functional specialization pattern extends across Brassica species. Studies in B. oleracea demonstrate MYB29-mediated enhancement of aliphatic GSLs [54,55], while in B. napus, Bna.HAG3.A3/MYB29 modulates root aromatic GSL variation [56]. The root-predominant expression of BjuA03.MYB29 and BjuB08.MYB29 suggests their potential involvement in aromatic GSL regulation, revealing another dimension of functional specialization. Structurally, variations in the C-terminal regions of BjuMYB29 proteins may underlie their functional divergence. Future investigation of BjuB02.MYB29 and BjuB08.MYB29, combined with comparative analysis across Brassica species, will further elucidate how sequence variation translates to regulatory specialization in GSL regulation.

For crop improvement applications, the functionally diversified BjuMYB29 homologs provide valuable molecular tools for metabolic engineering. To date, genes involved in GSL biosynthesis have been extensively characterized in Arabidopsis and Brassica species [12,57], with some already serving as effective targets for modifying GSL profiles [58]. Building on this knowledge, our identification of functionally specialized BjuMYB29 homologs enables precise manipulation of GSL pathways. These insights establish a robust foundation for developing improved mustard varieties with optimized defense and nutritional properties through targeted selection of specific MYB29 homologs.

5. Conclusions

This study advances our understanding of GSL regulation beyond the Arabidopsis model by demonstrating how polyploid evolution in B. juncea has driven functional diversification of the MYB29 transcription factor family. We identified four MYB29 homologs that exhibit distinct spatiotemporal expression patterns and differential responses to biotic stress, reflecting evolutionary subfunctionalization following whole-genome triplication. More significantly, we demonstrated that while both BjuA03.MYB29 and BjuA10.MYB29 enhance aliphatic GSL accumulation and insect resistance, they exert opposing effects on indolic GSL biosynthesis, revealing a regulatory divergence that enables independent control of distinct GSL pathways. These findings provide new molecular insights for mustard crop improvement. The functional specialization among MYB29 homologs offers unique opportunities for precise manipulation of GSL profiles—whether through selective enhancement of defense traits or balanced improvement of both resistance and quality characteristics. The natural variation in these regulatory genes represents a valuable genetic resource for future breeding strategies aimed at developing improved mustard cultivars with optimized agronomic and nutritional traits.