A Mutation in the MYBL2-1 Gene Is Associated with Purple Pigmentation in Brassica oleracea

Anthocyanins are well-known antioxidants that are beneficial for plants and consumers. Dihydroflavonol-4-reductase (DFR) is a key gene of anthocyanin biosynthesis, controlled by multiple transcription factors. Its expression can be enhanced by mutations in the negative regulator of anthocyanin biosynthesis myeloblastosis family transcription factor-like 2 (MYBL2). The expression profiles of the DFR gene were examined in 43 purple and green varieties of Brassica oleracea L., Brassica napus L., Brassica juncea L., and Brassica rapa L. MYBL2 gene expression was significantly reduced in purple varieties of B. oleracea, and green varieties of B. juncea. The MYBL2 gene sequences were screened for mutations that can affect pigmentation. Expression of the DFR gene was cultivar-specific, but in general it correlated with anthocyanin content and was higher in purple plants. Two single nucleotide polymorphysms (SNPs) were found at the beginning of the DNA-binding domain of MYBL2 gene in all purple varieties of B. oleracea. This mutation, leading to an amino acid substitution and the formation of a mononucleotide repeat (A)8, significantly affects RNA structure. No other noteworthy mutations were found in the MYBL2 gene in green varieties of B. oleracea and other studied species. These results bring new insights into the regulation of anthocyanin biosynthesis in genus Brassica and provide opportunities for generation of new purple varieties with precise mutations introduced via genetic engineering and CRISPR/Cas.


Introduction
The genus Brassica includes species distributed worldwide, such as B. oleracea, B. napus, B. juncea, and B. rapa. Brassica vegetables are economically important dietary products rich in phenolic compounds, carotenoids, vitamins C, and E [1]. Purple varieties of Brassica sp. are the most rare and beneficial for health due to the presence of anthocyanins. These phenolic compounds can be used as natural dyes and colorants in food due to the high thermal stability [2]. Anthocyanins are also associated with increased stress tolerance [3].
The anthocyanin biosynthesis pathway is well-studied; however, the specific genes responsible for purple pigmentation in Brassica sp. have not been discovered yet. The genetic regulation of anthocyanin accumulation is traditionally studied in ornamental plants, such as petunia, phalaenopsis, and snapdragon [4][5][6][7].
The key gene of anthocyanin biosynthesis, encoding dihydroflavonol-4-reductase (DFR), is controlled by multiple transcription factors and micro RNAs. Low expression of this gene, which catalyzes the synthesis of leucoanthocyanidin, is suggested to be the bottleneck, preventing anthocyanin synthesis in plants [8]. PAP1 (AN2), PAP2, TT8, TTG1, MYB2, and Delila are known as its positive regulators. GLABRA2, HAT1, CPC, MYBL2, MYB1, and MYB57 were reported to downregulate the expression of the DFR gene. Three types of transcription factors: MYB, MYC/basic Helix-Loop-Helix (bHLH), and WD40repeat (WDR) proteins usually act in MBW complexes. These complexes bind to the DFR promoter and regulate the transcription of DFR gene to alter anthocyanin biosynthesis [9].
In phalaenopsis, DFR and MYB transcripts were present in the purple rather than the white sectors of the flower; however, the expression level of other structural and regulatory genes was the same in both sectors [4]. Zoysiagrass cultivars with green and purple spikes and stolons differed by the expression level of Anthocyanidin synthase and DFR genes [10]. In black-skinned pomegranate, the DFR gene expression was higher, and AN2 gene expression was lower than in samples with white skin [11]. In the white petunia, the introduction of an exogenous DFR gene resulted in a pink-flowered phenotype [12]. In general, in white flowers the expression of structural anthocyanin genes is undetectable [4,13]. However green varieties of Brassica sp. produce some anthocyanins, their content is by several orders of magnitude lower than in purple varieties [14]. Unlike petunia, which has three different DFR genes to control pigmentation in different organs, most Brassicaceae have only one functional gene and usually accumulate anthocyanins in all tissues [15]. Overexpression of the exogenous AtDFR gene in B. napus increased DFR transcript levels, anthocyanin accumulation in the shoots, and salt tolerance [3]. Purple pigmentation may appear or decrease in response to stress, lighting intensity, changes in nitrogen concentration and other factors. This suggests that in green Brassica sp. DFR expression is not blocked absolutely, but is controlled via a quantitative regulatory system [16].
Some structural and regulatory genes of Brassicaceae can be solely responsible for purple pigmentation or its absence [17]. Mutations such as indels can affect anthocyanin pigmentation. A mutation in GLABRA3 gene contributed to a 98% decrease in the DFR transcript content and anthocyaninless phenotype in Arabidopsis thaliana [8]. Insertion of a natural transposon in a TT8 gene of B. rapa contributed to the loss of pigmentation in the seeds [14]. Mutation of the promoter activated the expression of the BoMYB2 gene in purple cauliflower [18].
Mutations identified in the coding sequence of the DFR gene itself also can disrupt the functions of the protein [19,20]. Deletion of the DFR gene in purple ornamental kale resulted in a loss of pigmentation [21].
The negative regulator, MYBL2, which encodes a R3-MYB-related protein, appears to make the greatest contribution to changes in DFR gene expression and anthocyanin pigmentation in many plant species. MYBL2 expression appears to be dependent on light intensity and temperature [22,23]. The MYBL2 transcription factor has a DNA-binding MYB domain and a repressive domain. In A. thaliana the minimal repression domain of MYBL2 consists of six amino acids (TLLLFR) at the carboxyl terminus [24]. Possible inhibition mechanisms of this negative regulator have not yet been clarified. It either binds to the bHLH protein and prevents the formation of MBW complexes or replaces it in the complex [25][26][27].
It has been confirmed that silencing of the MYBL2 gene in A. thaliana promoted DFR expression [26,27]. Unlike DFR gene, MYBL2 might have several orthologs in Brassica sp. [28]. For example, MYBL2-1 mRNA was not detectable in purple B. oleracea; however MYBL2-2 was expressed. This suggests that only MYBL2-1 is closely associated with anthocyanin production in B. oleracea; however, the second ortholog probably did not retain this function. Anthocyanin hyperaccumulation in these plants resulted from either the promoter substitution or deletion of the MYBL2-1 gene [25]. Data on the role of MYBL2-1 gene in B. rapa are contradictory. It has been suggested that a 100-bp insertion in the third exon of MYBL2-1 gene is associated with purple pigmentation in this species [17,29]. However, MYBL2-1 gene is highly expressed in another purple variety of B. rapa and therefore may serve as a positive regulator [30].
However it is clear that the MYBL2 gene plays a key role in the regulation of anthocyanin biosynthesis, data on mutations that reliably induce anthocyanin accumulation in genus Brassica are very limited. Most of the experiments were carried out on the model plant A. thaliana, which has a single MYBL2 gene copy. The studied B. oleracea and B. rapa varieties were of Asian origin; however, purple Brassica vegetables are supposed to be native to the Mediterranean region of Europe [29]. In B. napus and B. juncea this gene has never been studied. MYBL2 has not yet been subjected to CRISPR/Cas editing, although it appears to be a perfect target gene, responsible for phenotypic changes [9].
To figure out if any mutations in the MYBL2 gene can be associated with pigmentation in 43 purple and green varieties of B. oleracea L., B. napus L., B. juncea L., and B. rapa L., expression profiles of DFR and MYBL2 genes and sequencing of the MYBL2 gene were performed. Anthocyanin content was measured in green and purple varieties. The possibilities to introduce precise mutations to the MYBL2 gene via CRISPR/Cas editing were evaluated.

Expression Profiles of the DFR Gene in Brassicaceae
Visible purple pigmentation in B. oleracea was not directly related to the expression of the DFR gene ( Figure 1). In general, purple varieties of the same cultivar produced more mRNA of the DFR gene than green ones, but there was a great difference between cultivars. Purple kohlrabi (Figure 1(20-22)) did not differ significantly from several green varieties of headed cabbage by the expression of DFR gene (Figure 1(6,7)); however, the phenotypic differences were clearly visible. The primary leaf veins, petioles, and stems were the most pigmented organs in B. oleracea, and not the leaf lamina.
In several varieties of headed cabbage and Brussels sprout ( Figure 1(2,3,8)) the DFR gene was expressed at the level of the reference gene (70-90%); however, in cauliflower and kohlrabi it only reached 20-30% of the reference gene expression. Our results indicate that the DFR gene is not decisive for the formation of anthocyanin pigmentation in all varieties of cabbage; therefore, other genes may be involved.
B. rapa was the only species where mRNA content of the DFR gene was higher than those of the reference gene ( Figure 2 (38)(39)(40)). The cotyledons of B. rapa were also the most pigmented among all studied Brassica species. On the contrary, B. napus with pigmented roots had green leaves with unremarkable expression level of the DFR gene ( Figure 2 (33,34)). One of the B. juncea varieties was supposed to accumulate anthocyanins (Figure 2(28)); however, its leaves were not visibly pigmented and demonstrated low expression level of the DFR gene, but not as low as in green varieties. In green cabbage this gene was expressed at a certain level ( Figure 1); however, in green B. juncea and B. rapa its mRNA content was close to zero ( Figure 2 (31,32,42,43)).
Our results indicate that expression profiles of the DFR gene in Brassicaceae are speciesspecific and even cultivar-specific.

Expression Profiles of the MYBL2 Gene in Brassicaceae
However DFR gene expression in several purple varieties of B. oleracea was no higher than in green ones, MYBL2 mRNA content was very low and did not exceed 10% in all purple varieties (Figure 1 (1,4,21). In green varieties of B. oleracea MYBL2 gene expression was also cultivar-specific. In headed cabbage and kohlrabi it was the highest (40-50% of the reference gene), but in Brussels sprout and cauliflower it was almost as low as in purple plants (Figure 1(8-19)). Nevertheless, in these two cultivars a strong negative correlation was observed between the expression levels of DFR and MYBL2 genes (correlation coefficients = −0.78 and −0.8, respectively); however, in headed cabbage and kohlrabi coefficients were lower (−0.5 and −0.65).
Interestingly, in three purple varieties of B. juncea, the expression of MYBL2 gene was significantly higher than in green varieties ( Figure 2(26-28)), which was unexpected of a putative negative regulator of anthocyanin biosynthesis. No correlation was found between the expression levels of DFR and MYBL2 genes in this species.
The content of MYBL2 mRNA in purple B. rapa and B. oleracea was comparable, except for the Mizuna Red cultivar (Figure 2(39)). The level of DFR expression in this cultivar was also lower than in Ruby little mermaid and Mizuna purple ( Figure 2 (38,40)). The correlation coefficient between the expression levels of the two studied genes in B. rapa was −0.62.  pigmented roots had green leaves with unremarkable expression level of the DFR gene ( Figure 2 (33,34)). One of the B. juncea varieties was supposed to accumulate anthocyanins (Figure 2(28)); however, its leaves were not visibly pigmented and demonstrated low expression level of the DFR gene, but not as low as in green varieties. In green cabbage this gene was expressed at a certain level ( Figure 1); however, in green B. juncea and B. rapa its mRNA content was close to zero ( Figure 2 (31,32,42,43)).

Total Anthocyanin Content
To reveal how DFR and MYBL2 gene expression levels affect the accumulation of anthocyanins, their content was measured in green and purple varieties.
The content of anthocyanins in the studied plants ranged from 6 to 411 mg/100 g of dry weight (DW). In general, this parameter positively correlated with the DFR gene expression level (correlation coefficient = 0.83). Among groups, only in B. napus and headed cabbage correlation between DFR mRNA and anthocyanin content was weaker (0.4 and 0.66). Correlation with MYBL2 gene expression was negative, but very weak. It was notable only in kohlrabi (−0.92), Brussels sprout (−0.69), and B. rapa (−0.75). These results are indicative of the direct involvement of the DFR gene in anthocyanin biosynthesis and regulatory role of the MYBL2 gene, which may differ depending on the plant variety and exogenous factors.
Green varieties contained on average 14 mg/100 g of anthocyanins, while in purple varieties anthocyanin content was 10 times higher. These results are consistent with the literature data [10]. However, there were some exceptions. For example, in the rutabaga Gera (Figure 2(34)), which was supposed to have pigmentation only in roots, the leaves contained no less anthocyanins than the purple varieties of B. oleracea. The expression of the DFR gene in B. napus was also higher than in most of the anthocyaninless varieties of other species. This indicates that the regulation of anthocyanin biosynthesis in B. napus may be more complex due to ploidy and multiple copies of the gene.
In purple-headed cabbages Kalibos and Mars (Figure 1(1,4)) DFR mRNA content was the same as in green varieties. Indeed, anthocyanin content in these plants was also low, and MYBL2 expression was higher than in other purple varieties of B. oleracea, but much lower than in green-headed cabbages.
The highest anthocyanin content as well as the highest DFR gene expression was observed in B. rapa. As far as these species have thin, delicate leaves, they may contain less debris and more anthocyanins per unit weight. The involvement of the MYBL2 gene in anthocyanin biosynthesis in these species is very likely. On the contrary, no noticeable effect of the MYBL2 expression on the accumulation of anthocyanins was observed in B. juncea.
To determine if the negative regulator of anthocyanin biosynthesis MYBL2 is related to the DFR expression profiles and anthocyanin content, we subjected the samples of each species and cultivar to sequencing.

Sequencing of the MYBL2 Gene
However it was earlier reported that purple varieties of B. oleracea lacked BoMYBL2-1 coding sequences [25], amplification was successful in all studied samples. A 100-bp insertion, associated with purple pigmentation in B. rapa [29], also was not detected.
We discovered two completely new SNPs in a MYBL2 sequence of all analyzed purple varieties of B. oleracea, which affected the beginning of the DNA-binding MYB domain, resulting in the replacement of two amino acids (from KESN to KKNN), compared with B. rapa (Figure 3). The mutation was common for all cultivars of B. oleracea; however, it was not detected in any of the green varieties, or in B. juncea, B. napus, and B. rapa. In green varieties of B. oleracea there was also one SNP in this site, compared with the MYBL2 gene of B. rapa, resulting in an amino acid change (GAA into CAA -KESN into KQSN); however, it was outside of the DNA-binding domain. Therefore, the mutation probably did not affect DNA binding. Other species did not have any mutations within this site. Close to the carboxyl terminus there was also one SNP, specific for purple cabbage (GTG into GTT); however, it did not result in an amino acid substitution.
It should be noted that there were occasional heterozygous mutations in green varieties of cabbage ( Figure 3C); however, in purple varieties there were almost none which means they have homozygous alleles of MYBL2 gene.
There were other differences between the varieties, which cannot be directly associated with anthocyanin pigmentation. For example, a 30 bp long insertion (GTAATGTGAATGT-TATTTTTTTTGCAAAAA) was observed in the noncoding sequence of the MYBL2 gene in purple and green B. oleracea, compared with B. rapa.
Therefore, we were unable to detect any noteworthy mutations in purple varieties of B. juncea, B. napus, and B. rapa, which can be associated with hyperpigmentation. napus, and B. rapa. In green varieties of B. oleracea there was also one SNP in this site, compared with the MYBL2 gene of B. rapa, resulting in an amino acid change (GAA into CAA -KESN into KQSN); however, it was outside of the DNA-binding domain. Therefore, the mutation probably did not affect DNA binding. Other species did not have any mutations within this site. Close to the carboxyl terminus there was also one SNP, specific for purple cabbage (GTG into GTT); however, it did not result in an amino acid substitution.

Discussion
As a result of a present study, we report a mutation in the beginning of a DNA-binding site of a MYBL2 gene of purple B. oleracea for the first time. According to the NCBI database, AAAGAAAGCAAC into AAAAAAAACAAC (KESN into KKNN) mutation is unique and has never previously been found. The presence of the same mutation in all purple varieties of B. oleracea strongly suggests that it is directly associated with anthocyanin pigmentation.
There are several possible mechanisms by which this mutation can result in phenotypic changes. First of all, it is a mononucleotide (A) 8 repeat (SSR) in the coding region, which can lead to gene inactivation, changes in function and phenotype [31]. SSRs can serve as binding sites for regulatory elements, may form altered DNA secondary structures, and affect gene translation. For example, translational repression of the MYBL2 gene by MiR858 was shown to enhance anthocyanin biosynthesis in seedlings of A. thaliana [32]. SSRs often cause DNA slippage during DNA replication [33]. This mutation also creates the KKXX dilysine motif, which promotes the relocation of proteins from the Golgi apparatus to the endoplasmic reticulum [34].
The mutation affects the secondary structure of RNA (Figure 4), which plays a central role in post-transcriptional regulation of gene expression and affect recognition by ribosomes. U-and A-rich sites may correlate with the recognition of endonucleases for regulating alternative polyadenylation [35].
In all purple varieties of B. oleracea expression level of the MYBL2 gene was lower than 10%; however, in green varieties it can be as high as 50% (Figure 1). Variability in the mRNA content of this gene was not so pronounced in other plant species (Figure 2). It is interesting that strong correlation between the expression of MYBL2 and DFR genes was also observed in B. napus, which is indicative of the special role of MYBL2 gene in the C genome. In B. oleracea and B. napus this gene negatively affected the expression of DFR, such as in A. thaliana [26,27].  In all purple varieties of B. oleracea expression level of the MYBL2 gene was lowe than 10%; however, in green varieties it can be as high as 50% (Figure 1). Variability in the mRNA content of this gene was not so pronounced in other plant species (Figure 2) It is interesting that strong correlation between the expression of MYBL2 and DFR genes was also observed in B. napus, which is indicative of the special role of MYBL2 gene in the C genome. In B. oleracea and B. napus this gene negatively affected the expression o DFR, such as in A. thaliana [26,27].
No significant mutations were detected in the MYBL2-1 gene sequence in purple varieties of B. juncea, B. napus, and B. rapa, which suggests that DFR gene expression in these species can be regulated differently. It has been shown that six amino acids (TLLLFR) at the carboxyl terminus are obligatory for the repression activity of the MYBL2 transcription factor in A. thaliana [24]. However, in B. rapa there is a substitution of the last amino acid from R to Q (CAG). In B. juncea MYBL2 expression was even higher in purple varieties, which suggests that, such as in B. rapa, this gene can also serve as a positive regulator of anthocyanin biosynthesis [30]. Therefore, in these species MYBL2 may work through a different mechanism of repression activity or have none a all. In our study the reverse primer for gene amplification covered half of the sequence encoding this repression domain. Successful amplification indicated that in purple varieties of B. rapa and B. juncea it was present.
This also suggests that the absence of the PCR product of MYBL2-1 gene in purple No significant mutations were detected in the MYBL2-1 gene sequence in purple varieties of B. juncea, B. napus, and B. rapa, which suggests that DFR gene expression in these species can be regulated differently. It has been shown that six amino acids (TLLLFR) at the carboxyl terminus are obligatory for the repression activity of the MYBL2 transcription factor in A. thaliana [24]. However, in B. rapa there is a substitution of the last amino acid from R to Q (CAG). In B. juncea MYBL2 expression was even higher in purple varieties, which suggests that, such as in B. rapa, this gene can also serve as a positive regulator of anthocyanin biosynthesis [30]. Therefore, in these species MYBL2 may work through a different mechanism of repression activity or have none at all. In our study the reverse primer for gene amplification covered half of the sequence encoding this repression domain. Successful amplification indicated that in purple varieties of B. rapa and B. juncea it was present.
This also suggests that the absence of the PCR product of MYBL2-1 gene in purple varieties of cabbage [25] may be due to primer design or contamination of the DNA with polyphenols, which can inhibit PCR reaction. There can be significant differences between European and Asian varieties of cabbage. All cultivars of B. oleracea involved in our study may have the same European (possibly Mediterranean) origin, because none of them were of Asian breeding. Some varieties, such as Kalibos, have been known in Europe for more than 200 years.
Interestingly, varieties of B. oleracea, such as Brussels sprout and kohlrabi, had different levels of MYBL2 and DFR genes expression and anthocyanin content. In purple and green kohlrabi the difference between MYBL2 mRNA content was dramatic; however, in green Brussels sprout MYBL2 mRNA content was rather low. On the contrary, there was a great difference in DFR expression between green and purple cultivars of Brussels sprout. This may be explained by the fact that DFR is not the only gene controlled by MYBL2, and MYBL2 is not the only anthocyanin biosynthesis repressor in Brassica. In B. oleracea the content of anthocyanins increases with maturation, which indicates the important role of regulatory genes such as MYBL2 in this process. The loss of MYBL2 activity in A. thaliana resulted in an increase in mRNA content of leucoanthocyanidin dioxygenase (LDOX) and, to a lesser extent -flavanone 3β-hydroxylase (F3H) and chalcone isomerase (CHI) structural genes [22]. Similar to DFR, LDOX is regulated by MBW complexes [36]. LDOX is involved in anthocyanin biosynthesis pathway downstream of the DFR, promoting the transformation of leucocyanidin to cyanidin [37]. F3H and CHI are involved upstream and are able to provide more substrate for DFR.
It has been demonstrated that an insertion in the coding region of LDOX gene resulted in a complete loss of pigmentation in pomegranate [38]. In A. thaliana a single aminoacid substitution in LDOX resulted in a pale brown seed color and lack in anthocyanin accumulation [36]. LDOX gene of Reaumuria trigyna complemented reduced proanthocyanin and anthocyanin levels in the Arabidopsis loss of function mutant [39]. Transgenic A. thaliana overexpressing RtLDOX2 accumulated more anthocyanins and flavonols only under abiotic stress, but also increased primary root length, biomass accumulation, and stress resistance [40]. There are several functional copies of endogenous LDOX gene in the genomes of Brassica sp., which complicates the investigation of their impact on the anthocyanin accumulation. Therefore, they remain understudied in these species. Consequently, LDOX may be a gene of interest in purple cabbage cultivars with low expression of DFR gene, such as kohlrabi.
We were unable to detect any significant mutations in MYBL2-1 gene of B. juncea, B. napus, and B. rapa, including those previously described [25,28]. This suggests that other genes may be responsible for anthocyanin accumulation in studied varieties. There are a lot of candidate genes, including not only transcriptional repressors CPC, LBD, and GLABRA2, but also activators such as PAP1, PAP2, MYB1, and MYB2 [9]. For example, a mutation in the promoter region activated the expression of BoMYB2 gene in purple B. oleracea [18]. Whole genome sequencing might be a more suitable method for further studies of B. juncea, B. napus, and B. rapa purple varieties.
Knowledge of beneficial mutations can allow fast generation of new purple varieties via traditional selection, genetic engineering, and CRISPR/Cas. Precise editing in the mutation site of the MYBL2-1 gene can allow to prove if two SNPs in the beginning of the DNA-binding domain are enough to induce athocyanin hyperaccumulation not only in B. oleracea, but also in other Brassica species.
To evaluate the possibility of CRISPR/Cas editing, we examined the mutation site in the beginning of the DNA-binding domain for the presence of the PAM sequences. There are several types of Cas nucleases, which recognize different PAM sequences [41]. The site of interest is AT-rich; therefore, the most common nuclease SpCas9 is poorly applicable (Figure 5a). Two of the three possible gRNAs close to the target site are inefficient, according to CRISPOR software calculations. The second most used nuclease Cas12a also is not appropriate for introducing the desired mutation, since the predicted double strand breaks are quite far from the target site (Figure 5b).
Modified nucleases such as iSpyMacCas9 [42], recognizing A-rich PAM sequence 5 -NAAR, can be a better choice to introduce desired mutations in the beginning of the DNA-binding domain of a MYBL2 gene (Figure 5c). Modified nucleases such as iSpyMacCas9 [42], recognizing A-rich PAM sequence 5′-NAAR, can be a better choice to introduce desired mutations in the beginning of the DNA-binding domain of a MYBL2 gene (Figure 5c).
Prime editing [43] might be the most suitable method to check if this particular mutation cause anthocyanin pigmentation in B. oleracea. With the use of elongated guide RNA, containing a donor sequence, it is possible to introduce the precise point mutations, identical to those in naturally bred purple cabbage. Nevertheless, according to the previous reports, many types of mutations in the MYBL2 gene, which are much easier to obtain, are applicable for generation of new purple varieties [25][26][27]29]. Any of the described nucleases can be used with multiple gRNAs to ensure complete excision of the target region or the whole gene. However complete excision may not be recommended as far as it can affect other functions of MYBL2-1 gene, such as trichome initiation and brassinosteroid signaling [26,44], absence of this gene copy did not seem to affect fitness of purple B. oleracea [25]. Therefore, attempts for introduction of the two desired SNPs are of little value to agriculture.
The discovery and introduction of point mutations, ensuring anthocyanin hyperaccumulation, can help to produce new purple varieties of economically important Brassicaceae crops. Brassica vegetables can be considered a functional food that can be used directly or as a dietary supplement [45]. Altering competition for substrate between flavonol synthase and dihydroflavonol-4-reductase by creating mutations in the MYBL2 gene may also increase stress tolerance and nutritional value of new purple varieties.

Plant Materials
We analyzed 43 varieties of four economically important species of Brassica, including 22 varieties that were supposed to be purple, according to the manufacturer. B. oleracea was represented by 25 varieties, including purple-headed cabbage (Kalibos, Ludmila, Ruby, Mars, and Firebird), purple Brussels sprout (Garnet bracelet and Bunch of grapes), purple cauliflower (Gardener's dream and Purple ball), and purple kohlrabi (Madonna, Violetta, and Vienna purple), as well as green varieties of headed cabbage (Royal Vantage and Moscow Late), Brussels sprout (Rosella, Hercules, and Sapphire), cauliflower (Bird's milk, Baby, Flame star, Green bunch, and Alpha), and kohlrabi (Vienna white, Gulliver, and Picante). Anthocyanin pigmentation was observed in all varieties that were supposed to be purple, according to the manufacturer.
Five purple varieties of B. juncea (Vitamin, Red velvet, Freckle, Miracles in the sieve, and Red giant) and two green varieties (Wavelet and Vigorous) were used in this study. However, Freckle lacked anthocyanin pigmentation. Among five varieties of B. napus, rutabaga Krasnoselskaya and Gera were supposed to accumulate anthocyanins in roots; Prime editing [43] might be the most suitable method to check if this particular mutation cause anthocyanin pigmentation in B. oleracea. With the use of elongated guide RNA, containing a donor sequence, it is possible to introduce the precise point mutations, identical to those in naturally bred purple cabbage. Nevertheless, according to the previous reports, many types of mutations in the MYBL2 gene, which are much easier to obtain, are applicable for generation of new purple varieties [25][26][27]29]. Any of the described nucleases can be used with multiple gRNAs to ensure complete excision of the target region or the whole gene. However complete excision may not be recommended as far as it can affect other functions of MYBL2-1 gene, such as trichome initiation and brassinosteroid signaling [26,44], absence of this gene copy did not seem to affect fitness of purple B. oleracea [25]. Therefore, attempts for introduction of the two desired SNPs are of little value to agriculture.
The discovery and introduction of point mutations, ensuring anthocyanin hyperaccumulation, can help to produce new purple varieties of economically important Brassicaceae crops. Brassica vegetables can be considered a functional food that can be used directly or as a dietary supplement [45]. Altering competition for substrate between flavonol synthase and dihydroflavonol-4-reductase by creating mutations in the MYBL2 gene may also increase stress tolerance and nutritional value of new purple varieties.

Plant Materials
We analyzed 43 varieties of four economically important species of Brassica, including 22 varieties that were supposed to be purple, according to the manufacturer. B. oleracea was represented by 25 varieties, including purple-headed cabbage (Kalibos, Ludmila, Ruby, Mars, and Firebird), purple Brussels sprout (Garnet bracelet and Bunch of grapes), purple cauliflower (Gardener's dream and Purple ball), and purple kohlrabi (Madonna, Violetta, and Vienna purple), as well as green varieties of headed cabbage (Royal Vantage and Moscow Late), Brussels sprout (Rosella, Hercules, and Sapphire), cauliflower (Bird's milk, Baby, Flame star, Green bunch, and Alpha), and kohlrabi (Vienna white, Gulliver, and Picante). Anthocyanin pigmentation was observed in all varieties that were supposed to be purple, according to the manufacturer.
Five purple varieties of B. juncea (Vitamin, Red velvet, Freckle, Miracles in the sieve, and Red giant) and two green varieties (Wavelet and Vigorous) were used in this study. However, Freckle lacked anthocyanin pigmentation. Among five varieties of B. napus, rutabaga Krasnoselskaya and Gera were supposed to accumulate anthocyanins in roots; however rutabaga Novgorodskaya and Child love, as well as fodder cabbage Veha, were supposed to be anthocyaninless. Visible purple pigmentation was not detected in the leaves of any of these varieties. We also studied three purple varieties of B. rapa (Ruby little mermaid, Mizuna Red, and Mizuna purple) and three green varieties (Impulse, The little mermaid, and Mizuna green).
Green and purple varieties of each cultivar of B. oleracea and other studied species were subjected to Sanger sequencing of the MYBL2 gene. Namely, green varieties Moscow late, Sapphire, Flame star, Gulliver, Vigorous, Novgorodskaya, Child love, and Impulse as well as purple varieties Ludmila, Garnet bracelet, Gardener's dream, Purple ball, Madonna, Vienna purple, Vitamin, Krasnoselskaya, Gera, and Mizuna red were examined.
Plants were grown at 20 • C under 10,000 lux and a 16:8 h day:night photoperiod, generated by LED grow light. True leaves were used for anthocyanin content measurement. Other experiments were carried out on cotyledons, homogenized using FastPrep-24 bead beating grinder and lysis system (MP Biomedicals, Santa Ana, CA, USA).

Real-Time Quantitative PCR
RNA was extracted from 50 mg of the plant tissue using Lira reagent (Biolabmix, Novosibirsk, Russia). cDNA was synthesized with OT-M-MuLV-RH kit (Biolabmix, Novosibirsk, Russia) using oligo(dT) primer.

Total Anthocyanin Content
Anthocyanin pigment concentration was measured by the spectrophotometric pH differential method using a Perkin Elmer LS 55 Luminescence Spectrometer (Perkin Elmer, Waltham, MA, USA) and expressed as cyanidin-3-glucoside equivalents. Anthocyanins in Brassicaceae are known to be cyanidin derivatives. Extracts from 15 mg of dried leaf tissue in pH 1.0 and pH 4.5 buffers were prepared and analyzed according to Lee et al. [49]. Measurements were performed in three technical and three biological replicates in 96-well plates.

Statistical Analysis
Three samples of each variety were used for DNA and RNA extraction. Primers were designed using NCBI Primer BLAST tool [46]. RT-qPCR was carried out for three replicates of each sample. Results of RT PCR were assessed by 2 −∆∆CT method. Sequencing was performed using forward and reverse primers for each sample. Sequences were aligned to the reference sequence of the MYBL2-1 gene of B. rapa (JN379102) and analyzed using SnapGene software. Functional domains were analyzed via InterPro [50]. Search for gRNAs was performed in CRISPOR web tool [51]. RNA structure was predicted by RNAfold web tool [52]. For all experiments, means and standard deviation (p < 0.05) were compared by