Understanding the genetic control behind the seed color is essential for breeding rapeseed genotypes with increased protein and reduced fiber contents. Two independent dominant genes, Br1 and Br3, are responsible for seed coat color. Dominance at the Br1 locus results in brown seeds, while the dominant Br3 locus with a homozygous recessive Br1 locus results in yellow-brown seeds. Both loci in a homozygous recessive condition produce yellow seeds (Stringam, 1980).

In B. napus, three genes are responsible for the seed coat color expression (Rahman et al., 2010). The seed coat colors are black to dark brown, light brown, dark yellow, light yellow and yellow. The genes BSNap1 and BSNap2 control black/brown seed coat color, while the gene DKYSNap3 influences dark/light yellow seed coat color (Rahman et al., 2010). Homozygous recessive condition across all three genes results in yellow seeds. Studies on the seed coat color inheritance in F₂, BC₁ (Back Cross 1) and the DH (Double Haploid) population revealed the epistatic interactions between the dominant gene for yellow seed color and two independently segregating dominant genes for black color seeds (Liu et al., 2008).

Quantitative trait loci (QTLs) mapping identified two QTLs responsible for seed coat color, which accounted for 46% and 30.9% variation (Liu et al., 2006). Further discoveries of QTLs for seed coat color, protein, and fiber contents concluded that many genes control these traits and are highly influenced by the environment. For example, temperatures can affect the yellow and black-seeded rapeseed varieties. High temperatures might disrupt pigment synthesis, resulting in yellow-seeded genotypes (Deynze et al., 1993).

The discovery of QTLs significantly affects marker-assisted selections (MAS) and genomic selections. Identification of molecular markers associated with QTL of desirable traits facilitates MAS. This enables early selection, pyramiding favorable QTLs, and reduces the duration of the breeding cycle to get favorable phenotypes with improved performance. However, the QTL analysis for seed quality traits in rapeseed requires trials in multiple environments to minimize the environmental influence and determine stable QTLs across environments. Only upon validation of such major and consensus QTLs, they can be recruited in MAS programs.

As mentioned above, SSPs have immense potential in the food processing industry. However, breeding for favorable genotypes requires understanding the trait’s genetic basis. SSP content is a quantitative trait that involves complex gene regulation and biosynthesis. The SSPs, cruciferin and napin are coded by multiple genes, particularly 9-12 genes involved in the cruciferin biosynthesis (Wanasundara, 2011), and 10 genes are responsible for the napin biosynthesis (Josefsson et al., 1987). SSP content within the seed also depends on post-translational processes and the transfer of assembled SSP into storage vacuoles (So and Duncan, 2021). Several QTLs are associated with SSP content. QTLs Nap-1, Nap-2, Nap-3, Cru-1, Cru-2, Nap/Cru-1, Nap/Cru-2, and Nap/Cru-3 were identified in the double haploid populations of winter oilseed rape (Schatzki et al., 2014). Moreover, transcription factors such as Abscisic acid insensitive 3 (ABI3), Fusca 3 (FUS3), Leafy Cotyledon 1 (LEC1) and Leafy Cotyledon 2 (LEC2) are involved in the SSP regulation (Vicient et al., 2000; Kroj et al., 2003; So and Duncan, 2021). Hence, understanding the genetic control and regulatory mechanisms behind SSPs is crucial for modern breeding strategies to enhance SSP content for improved food processing applications.

Classical mutation breeding techniques have often focused on increasing rapeseed oil yield and quality (Röbbelen, 1990). Nonetheless, mutation breeding also has immense potential for reducing ANFs (Harloff et al., 2012) and increasing the protein fractions in RSM. Double haploid mutant lines produced from seeds or microspores are a promising option for producing successful mutations. The National Research Council of Canada (NRC) utilized this approach and recently reported a mutant line with a 7% increase in protein content (National Research Council Canada, 2023; available at: https://nrc.canada.ca/en/stories/boosting-canola-crop-value-mutation-breeding).

Similarly, transposon element (TE) insertions provide a biological means of inducing plant mutations. Transposons are mobile elements in plants that were once considered junk DNA and are now a valuable tool for breeders to create favorable plant variations. TE insertions in the coding regions will lead to loss-of-function mutations. In contrast, insertion in intron regions can result in new isoforms through exonization, alternative splicing, truncation, or a combination of these mechanisms (Dubin et al., 2018). Moreover, TE insertions can potentially disrupt enhancers or regulatory promoter elements outside the transcribing regions, thereby altering the transcriptional level (Dubin et al., 2018). TE activation can be achieved through toxins that inhibit DNA methylation, and stress-mediated TE activation has also been reported in different crops (Ito, 2022), for example, in vitro regeneration stress results in TE element activation in rice (Hirochika et al., 1996) and tobacco (Hirochika, 1993). However, TE-mediated mutagenesis requires robust methods to induce TE activation and transgenerational inheritance of TE insertions for further breeding implications (Kirov, 2023).

TE insertions in rapeseed have improved silique length and seed weight and reduced pod shattering. CACTA-like TE (3.9 kb) inserted in the upstream of BnaA9.CYP78A9 increased the expression of the loci, resulting in elongated siliques (Shi et al., 2019). Similarly, long-term repeat retrotransposon insertion (4803 bp) in the promotor of BnSHP1.A9 repressed its expression and conferred resistance to pod shattering (Liu et al., 2019). However, such transposons-mediated mutagenic studies have yet to be conducted to improve seed meal quality traits.

Mutation breeding could be essential as it avoids legislative pressure and regulations. Despite its potential, mutation breeding faces challenges due to a low frequency of mutation occurrence, unspecific mutations causing undesirable mutations, and safety hazards associated with the use of dangerous physical and chemical mutagenic agents.

Genomics in breeding enables the exploration of various advanced tools to obtain accurate population genotypic data. Genome sequencing technologies provide high throughput genome sequencing data with high accuracy and reliability. The recent advances in the discovery of single-nucleotide polymorphism (SNP) markers through genome-wide association studies (GWAS) improved the efficiency of correlating the phenotypic and genotypic data and selection efficiency for precision breeding. Brassica napus 60K Illumina Infinium™ SNP array allows multiplex SNP genotyping, enabling simultaneous surveys of thousands of SNPs and increasing the efficiency of the genotypic data (Wang et al., 2014; Mason et al., 2017; Zhou et al., 2021).

AM approach facilitates the identification of markers closely linked to the genes of interest. Unlike linkage mapping, which relies on recombination events between the gene and the marker in a bi-parental cross, AM utilizes all the recombination events that would have happened in the past in the population (Singh and Singh, 2015). The post-GWAS era offers new approaches for processing GWAS data to extract meaningful information by identifying and functionally characterizing candidate genes using bioinformatics and reverse genetics approaches. Targeting-induced local lesions in genomes (TILLING) approaches functionally characterize candidate genes identified via GWAS in mutagenized populations. The Eco-TILLING approach is used in germplasms or natural populations (Gupta et al., 2019). GWAS studies have been successfully utilized for several traits in rapeseed, including disease resistance, flowering-related traits, salt tolerance, aluminum tolerance and silique length (Helal et al., 2021; Roy et al., 2021; Wang et al., 2021; Zhang et al., 2022; Zhou et al., 2022; Starosta et al., 2024; Zhang et al., 2024). Similarly, GWAS has been successfully applied to rapeseed for various seed quality traits like fiber traits, ANFs, oil, and amino acid content (Wang et al., 2015; Wei et al., 2019; Bhinder et al., 2022; Tan et al., 2022; Zhao et al., 2022). However, studies focusing on seed storage proteins seem to be lacking (So and Duncan, 2021). Even though GWAS has identified many QTLs for essential traits in rapeseed, this technique faces challenges such as poor reproducibility and false positives, indicating the need for further improvement in this crop.

Traditional breeding focuses on phenotypic selection for genetic improvement. However, recent developments in genomics led to the identification of several QTLs useful for marker-assisted selection (MAS). MAS facilitates an indirect selection based on markers linked to the target traits. These markers are highly reliable, unaffected by the environment, and non-specific to plant growth stages, promoting early selection and reducing the breeding cycle. Combining MAS with speed breeding techniques can reduce the breeding cycles, thus promoting the development of new varieties in short periods. However, focusing solely on genetic composition might mislead the selection of agronomically poor phenotypes. The phenotypic selection combined with MAS is always preferred to achieve all the agronomic and quality traits.

The trait of yellow seeds was confirmed to be quantitative after intense research by various scientists (Liu et al., 2006; Fu et al., 2007; Chao et al., 2022a). The QTL discovery for fiber content and its correlation with the yellow seeds paved the way for genetic manipulation (Badani et al., 2006). The fiber content of the seeds compromises acid detergent fiber (ADF), neutral detergent fiber (NDF), and acid detergent lignin (ADL). Hemicellulose, cellulose, and lignin predominate in NDF, while the ADF comprises cellulose and lignin. ADL represents the nondigestible lignins in the seeds (Behnke et al., 2018). Altering these fractions could be beneficial for animal husbandry.

Different DNA-based molecular marker systems have been used for MAS to improve these quality traits. The yellow-seeded trait associated with 8 RAPD markers co-segregated with a single major gene (Pigment 1) responsible for the yellow seed coat color trait (Somers et al., 2001). Other markers, including restricted fragment length polymorphism (RFLP), amplified fragment polymorphism (AFLP), cleaved amplified polymorphic sequences (CAPS), sequence related amplified polymorphism (SRAP), sequence characterized amplified region (SCAR), and simple sequence repeat (SSR) markers (Deynze et al., 1995; Liu et al., 2006; Rahman et al., 2010; Zhang et al., 2010b; Huang et al., 2016) were also identified for breeding yellow seeds and low fiber content. Pleiotropy arises when a single gene or genetic locus affects multiple phenotypic traits. Pleiotropic effects between the target traits are advantageous as they facilitate breeding. Selecting one QTL might give the desired phenotype for the other targeted trait.

The QTL analysis by various scientists ( Table 7 ) clearly states that QTL at A09 is a major QTL for seed color and fiber traits (Liu et al., 2013; Miao et al., 2019; Chao et al., 2022a). The QTL at C05 is a major QTL for fiber and protein traits (Yusuf and Möllers, 2024). Focusing on these loci could provide desirable phenotypes. An apparent pleiotropic effect exists between seed fiber traits, plant height, and protein content (Wurschum et al., 2012; Liu et al., 2013; Miao et al., 2019). Additionally, a negative correlation exists between oil content and fiber characteristics, indicating that reducing the fiber content could improve the oil content. However, enhancing protein profiles will decrease the oil content due to their negative correlation (Yu et al., 2016; Chao et al., 2017, Yusuf and Möllers, 2024).

The delivery of the CRISPR/Cas9 system into the host requires a vector. Agrobacterium-mediated delivery of sgRNA and Cas9 protein is widely deployed in genome editing. Agrobacterium tumefaciens, a gram-negative soil bacterium, is responsible for causing gall growth in plants. It possesses a Ti plasmid, which makes it a perfect biological tool for vector constructs. The T-DNA region of a Ti plasmid is approximately 15–30 kb in size (Rahman et al., 2023). The DNA insert size is also limited in this type of gene transfer, thus limiting the possibility of the multiplexing techniques. The transfection process is mediated by inserting the gene of interest by replacing the T-DNA region and co-cultivating the bacteria with the explant. Successful Agrobacterium-mediated genome editing was done in rapeseed, targeting seed coat color and tannins (Cheng et al., 2023; Li et al., 2024).

However, transformation efficiency highly depends on the transformation protocols, the nature of the explant, co-cultivation periods, and the markers used in the transformation (Rahman et al., 2023). The random integration of T-DNA from bacterial plasmids into the host raises regulatory concerns, affecting the marketability of the products due to reduced consumer acceptance ( Figure 3 ). T-DNA components such as Cas9, gRNA, and selection markers can be effectively removed by selfing or backcrossing with wild-type to obtain edited progeny that is free of foreign DNA.

Another promising alternative for eliminating the foreign DNA integration with the host genome is RNPs, which ensures transgene-free genome editing ( Figure 3 ). This technique eliminates the need for a vector to deliver the sgRNA sequence and the protein while still being able to manipulate the DNA. Unlike Agrobacterium-mediated CRISPR/Cas9 editing, RNP-mediated editing contains Cas9 enzyme coupled with guide-RNA. The first step in RNP-mediated editing is isolating the protoplasts, followed by the transfection process, usually using polyethylene glycol (PEG) electrophoresis or biolistics for RNP delivery. The protoplasts lack cell walls, thereby facilitating the entry of RNPs into the host cells. The use of RNP complexes can reduce off-target effects and the risk of random DNA integration compared to plasmid-based delivery methods. This is due to the transient nature of RNPs in cells, which limits the time for unintended editing (Woo et al., 2015). However, carefully selecting guide RNAs can further minimize off-target effects in both Agrobacterium-mediated and RNP-based gene editing (Young et al., 2019).

Successful RNP-mediated genome editing using the biolistic approach has been shown in various crops (Svitashev et al., 2016; Liang et al., 2017; Banakar et al., 2020). Plasmid and RNP-based gene editing using protoplasts has been successfully demonstrated in rapeseed (Li et al., 2021; Sidorov et al., 2022). However, genomic instability (change in ploidy level due to endoreduplication and aneuploidy) associated with protoplast regeneration has been reported in various crops (Newell et al., 1984; Olin-Fatih, 1996; Sheng et al., 2011; Fossi et al., 2019).

As mentioned before, ANFs play a vital role in plant defense. The complete elimination of ANFs by knocking out the target genes through genome editing may be detrimental to the survival of the plants in some cases. So, reducing their levels instead of eliminating them could keep the defense system intact. Promoter editing ( Figure 4 ) could be used to target ANF profiles by altering the gene expression. The promoter regulates the gene expression and is located upstream of the region to be transcribed (Porto et al., 2014). It is a region where RNA polymerase binds to initiate transcription.

Promoter editing begins with assessing the secondary metabolites/ANF compounds that are present in the crop. This is followed by studying the metabolic pathways, enzymes, and compounds involved. Bioinformatics and omics approaches are essential for identifying the genes involved in these pathways and for pre-determining the candidate genes and their promoter regions that must be targeted to alter the ANF profiles. Once the editing is successful, plant phenotyping is essential for the morphological and agronomical assessment of the crop. Screening the phenotypes for disease resistance is also vital to determine that the defense system is not compromised.

Successful promotor editing has been done in various crops (Cui et al., 2020; Huang et al., 2020; Zeng et al., 2020; Liu et al., 2021a; Song et al., 2022b). However, Agrobacterium-mediated editing has been followed in promoter editing, which raises biosafety concerns. In this regard, RNP-based promoter editing would be an alternative approach to producing transgene-free mutants to overcome legislative issues.

The candidate genes that could be considered for genome editing to alter the ANF profiles are discussed below ( Table 8 ).

MYB28 is a transcription factor responsible for activating genes involved in aliphatic GSL biosynthesis. Downregulating MYB28 might reduce the aliphatic GSL biosynthesis. RNAi-mediated knockdown of the MYB28 decreased aliphatic GSL profiles, but the indolic GSLs remained unchanged (Gigolashvili et al., 2007). MYB28, MYB76, MYB29 activate genes in the aliphatic GSL biosynthetic pathways, such as MAM3, CYP79F1, and CYP83A1, without affecting genes involved in the indolic GSL pathway. Conversely, MYB34 and MYB51 activate genes for the indolic GSL biosynthesis (Gigolashvili et al., 2007). The downregulation of these transcription factors might affect genes such as CYP79B2 and CYP83B1 and prevent their conversion from tryptophan to GSLs. However, this might negatively affect the plant’s auxin homeostasis due to the interconnected nature of these pathways.

Moreover, the role of GSLs in defense must be addressed. Knocking down all transcription factors might harm the plants due to their interconnection with phytohormones, defense, growth and development. Thus, manipulating GSL profiles in specific tissues could be a better approach to maintaining the plant defense intact.

Biosynthesis and accumulation of GSLs show a distinct spatiotemporal pattern (Jorgensen et al., 2015) facilitated by glucosinolate transporters (GTR) transporting GSLs to various tissues. GTR proteins belong to the family of nitrate transporter 1/peptide transporter (NRT1/PTR). GTR1, GTR2, and GTR3 were recently identified, and elaborate information on GSLs mobilization and accumulation in sulfur-rich cells (S-cells) for storage in A. thaliana was described by (Schuster et al., 2006; Nour-Eldin et al., 2012). The manipulation of the GSL profiles via their transporters could be a better option. The gtr1gtr2 double knockout mutant failed to transport GSLs to seeds, resulting in a tenfold accumulation of GSLs in source tissues such as leaves and silique walls. The absence of aliphatic and indole GSLs in the gtr1gtr2 double knockout mutant seeds indicates their role in phloem loading and long-distance GSL transport. Thus, disrupting these transporters effectively reduces GSLs in the seeds without compromising plant defense.

This was successfully demonstrated in Brassica juncea (Mann et al., 2023). Knocking out GTR1 and GTR2 reduced the seed GSL content in B. juncea from 146.09 μmoles/g DW to 6.21 μmoles/g DW without compromising the defense (Mann et al., 2023). Similarly, the knockout in Camelina sativa reduced the seed GSL content up to 85%-88% (Hölzl et al., 2023). A clear picture of the GSL transporter is given in Figure 5 (Xu et al., 2023; Sanden et al., 2024). They have reported that the GTRs function as importers, and UMAMIT (Usually Multiple Amino Acids Move in And Out Transporter) functions as an exporters; together, they function as GSL transporters.

UMAMIT are localized in the plasma membrane and export GSLs into the apoplast from the source tissue. In contrast, GTRs in phloem companion cells import GSLs into the phloem, facilitating the transport to sink tissues. At the phloem unloading zone, GSLs are exported by UMAMIT to the apoplast, and the GTRs import the GSLs into the seed. Both transporters transport all three types of GSLs: aliphatic, aromatic and indolic. Targeting UMAMIT genes impairs GSL export and prevents deposition in the sink tissues. UMAMIT29, UMAMIT30, and UMAMIT3 are exporters found in the Arabidopsis thaliana (Xu et al., 2023). The umamit29 umamit30 umamit31 triple mutants exhibited low seed GSLs. In the seeds, mutant alleles of UMAMIT29 showed an 80% reduction in total GSL levels. Thus, UMAMIT29 could be the optimal target gene for reducing the GSL export.

Reducing phytic acid content in rapeseed is essential for improving the nutrient’s bioavailability. Phytic acid is synthesized through lipid-dependent and lipid-independent myo-inositol pathways. PMT5 and PIP5Ks play a vital role in the lipid-dependent biosynthesis pathway, whereas ITPK is pivotal in the lipid-independent biosynthesis pathway. SAC8 acts as a phosphatase and regulates phosphoinositide (Zhong and Ye, 2003). PMT5 (Polyol/monosaccharide transporter) is involved in transporting myo-inositol, glycerol, and several hexoses and pentoses, including ribose (Klepek et al., 2005). Thus, regulating the transporter could reduce the phytic acid content in the seeds. However, a major limitation lies in the multiple roles of the transporter. The knockdown mutants might result in poor seed development since PMT5 transports other compounds like glycerol and ribose. Thus, knockdown of the biosynthesis genes could be a better solution.

Phosphatidylinositol 4-phosphate 5-kinases (PIP5Ks) are a group of kinases that catalyze the production of phosphatidyl inositol (4,5)-bisphosphate (Van den Bout and Divecha, 2009). The ITPK gene family in B. napus holds four paralogues, namely BnITPK1, BnITPK2, BnITPK3, and BnITPK4. Ethyl methanesulphonate (EMS)-induced mutant lines targeting six genes in the phytic acid biosynthesis pathway were done by (Sashidhar et al., 2020a). However, due to the polyploid nature of rapeseed, the desired phenotype carrying all the mutant alleles requires laborious crossing. To overcome these challenges, BnITPK1 and BnITPK4 were knocked out by using CRISPR, resulting in a 35% phytate reduction (Sashidhar et al., 2020b). Similar phytate reduction has also been reported in soybean (Song et al., 2022a) and wheat (Ibrahim et al., 2022) using the CRISPR/Cas9 system.

Sinapine and lignin biosynthesis are interconnected and produced through the phenylpropanoid pathway (Boerjan et al., 2003; Milkowski and Strack, 2010). This interconnection provides a strategic advantage to alter both profiles by editing specific genes involved in biosynthesis. The enzymes involved in sinapine biosynthesis are UDP glucose sinapate glucosyltransferase (SGT), sinapoyl glucose choline sinapoyl transferase (SCT), sinapoyl glucose L-malate sinapoyl transferase (SMT) and sinapoyl choline (sinapine) esterase (SCE) (Milkowski and Strack, 2010). RNAi-mediated silencing was done in B. napus for the UGT84A9 gene, which disrupted the amounts and nature of the phenylpropanoid end products (Wolfram et al., 2010). Similarly, RNAi-mediated BnREF1 suppression resulted in reduced levels of sinapate esters, conjugated monolignols, dilignols, and trilignols and altered the minor conjugates of ferulate, caffeate, and 5hydroxyferulate (Mittasch et al., 2013).

The polymerization of monolignols like p-coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol is essential for lignin biosynthesis (Xie et al., 2018). The monolignols are synthesized in the cytoplasm and transported to the cell wall, where they are polymerized into lignin, creating the p-hydroxyphenyl (H), guaiacyl (G) and syringyl (S) units (Boerjan et al., 2003). Targeting genes involved in this biosynthetic pathway, such as quinate hydroxycinnamoyl transferase (HCT), caffeic acid O-methyltransferase1 (COMT1), ferulate-5-hydroxylase, BGLU46, BGLU45, cinnamyl alcohol dehydrogenase, phenyl ammonia lyase 4 and cinnamoyl-CoA reductase 1 could reduce the lignin contents. The mutant obtained through CRISPR/Cas9 editing for caffeic acid O-methyltransferase 1 reduced lignin content by 14% and 34% increase in higher fermentable glucose recovery rate than the wild-type (Lee et al., 2021).

Additionally, cellulose synthase and mannose-1-phosphate guanylyltransferase 1 are involved in cellulose and hemicellulose biosynthesis, respectively (Lukowitz et al., 2001; Hu et al., 2018; Bhinder et al., 2022). This type of genetic modification improves the use of rapeseed in biofuel production by altering its fiber fraction, which comprises lignin, cellulose and hemicellulose. These compounds are involved in plant defense and maintain cell wall integrity to provide mechanical support to the plants. In this regard, promoter editing could be used to alter their levels rather than knocking down the genes completely, to maintain the essential structural integrity and defense while optimizing biosynthetic pathways.

Recent efforts have been focused on altering SSPs to improve the meal quality of rapeseed. Unlike napin, a cysteine-rich SSP in rapeseed, the SSP of Brazil nut is rich in methionine. A chimeric gene with a methionine-rich 2S SSP coding sequence with soybean lectin promoter was used to improve the methionine content in rapeseed (Guerche et al., 1990). A similar attempt was made using a phaseolin promoter, resulting in an increase of 33% in the methionine content (Altenbach et al., 1992).

Improving the transport of the amino acids to the seed (sink) is another possible strategy. In arabidopsis, CAT6, AAP1, AAP2 and AAP8 have been reported to transport the amino acids from the source to the sink. CAT6 preferentially is known to transport neutral and cationic amino acids (Hammes et al., 2006). The mutant aap1 reduced total nitrogen (N) but showed a substantial increase in free amino acids in the seed coat and endosperm. This indicates that AAP1 is involved in the import of amino acids into the embryo (Sanders et al., 2009). AAP2 is involved in the phloem loading of amino acids; the aap2 mutant showed decreased amino acids in phloem sap and increased N content, leading to more branches (Zhang et al., 2010a). Meanwhile, AAP8 is crucial in transporting amino acids into the endosperm to nourish the embryo during early embryogenesis, with a high affinity towards acidic amino acids (Schmidt et al., 2007).

Altering a particular transporter could increase specific amino acid content. For instance, upregulating the expression of AtCAT6 could improve the transport of cationic amino acids like lysine and neutral amino acids like methionine and cysteine. Thus, increasing the expression of transporters in seeds could enhance the levels of essential amino acids such as lysine and methionine, which are often limited in plant-based diets (Yang et al., 2023).

Silencing cruciferin genes in rapeseed increased SSP cysteine, lysine, and methionine levels (Kohno-Murase et al., 1995). Conversely, silencing the napins in rapeseed resulted in higher levels of cruciferins and lower amounts of cysteine and lysine (Kohno-Murase et al., 1994). The inhibition of one type of SSP results in the accumulation of free amino acids, which can be directly allocated to synthesizing another SSP (Kohno-Murase et al., 1995). This suggests that protein profiles can be altered by targeting the specific genes involved in the biosynthesis of SSPs. However, these alterations did not affect the seed’s overall protein or oil content (Rolletschek et al., 2020). Thus, targeting SSP biosynthesis genes could modify the proportion of cruciferin and napins and improve their implications in food processing industries.