The genome sequences of three basic species [B. rapa (Brara_Chiifu_V3.5), B. nigra (Brana_ NI100_V2) and B. oleracea (Braol_JZS_V2.0)] and three allotetraploids [B. juncea (Braju_tum_V1.5), B. napus (Brana_Dar_V5) and B. carinata (http://brassicadb.bio2db.com/download.html)] of the genus Brassica were downloaded from the Brassica Info (http://www.Brassica.info/) website (Chen et al., 2010; Liu et al., 2014). The sequences obtained for B. rapa, B. nigra, B. oleracea, B. juncea, B. napus and B. carinata were 353140194 bp, 506000232bp, 561157886 bp, 937030072 bp, 850292103 bp, and 1086987601 bp in length, respectively.

The Krait identification tool was employed to search for the presence of SSR motifs in the genomic sequences (Du et al., 2018). The parameters were set as follows: the minimum number of repeat units was 12 single nucleotides, 7 dinucleotides, 5 trinucleotides, 4 tetranucleotides and pentanucleotides, and 4 hexanucleotides. The frequency and length of the searched SSRs were counted and analyzed.

The primer pairs on both sides of the SSR loci were designed using Krait software (Du et al., 2018). The main parameters were set as follows: the primer length was controlled between 18 and 27 bp, with an optimal size of 20 bp; the melting temperature was 58°C to 65°C, with an optimal temperature of 60°C; the GC content was in the range of 30% to 80%; and the predicted PCR product was in the range of 100-300 bp. All other parameters were set as default.

The plant materials used in the experiment were from B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. juncea (AABB), B. napus (AACC), B. carinata (BBCC), Raphanus sativus (RR) and Arabidopsis thaliana (At) specimens. The total genomic DNA was extracted from young frozen leaf tissues using the SDS method. The genomic DNA concentrations (ng/µL) were adjusted to an experimentally specific 100 ng/µL using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, USA).

From the SSR primers designed based on the genomes of the six chosen species, 40 pairs of primers were randomly selected in each of the three diploid species, B. rapa, B. nigra and B. oleracea. To perform PCR amplification, a reaction mixture containing 1 µL (100 ng/µL) of template genomic DNA, 1µL (10 µmol/L) of each primer and 12.5 µL of 2×T5 Super PCR Mix (PAGE) buffer was added, followed by the addition of ddH₂O to 25 µL total volume of the reaction mixture. The PCR assay amplification procedure included pre-denaturation at 94°C for 3 min, then 35 cycles involving denaturation at 94°C for 30 s, annealing at 53°C for 30 s, and extension at 72°C for 30 s, and finally, extension at 72°C for 7 min. The PCR reaction procedure was performed on a BIO-RAD S1000TM Thermal Cycler instrument, and samples were stored at 4°C. The SSR cross-transfer rate refers to the number of the amplified bands obtained from the other seven related species except for itself/total bands×100%.

The PCR product was detected by 12% PAGE following a method modified from “Molecular Cloning: A Laboratory Manual” (Sambrook et al., 2001).

The obtained SSR marker loci were analyzed and calculated using MG2C (http://mg2c.iask.in/mg2c_v2.1/) to locate the physical position of the SSR on each chromosome (Chao et al., 2021). A binary matrix of ‘1’ and ‘0’ was prepared for the SSR marker allele data for all genotypes. The polymorphic information content (PIC) values, gene diversity and heterozygosity were calculated using PowerMarker 3.0 software (Liu and Muse, 2005).

From the genomic sequences of B. rapa, B. nigra, B. oleracea, B. napus, B. juncea, and B. carinata with lengths of 340, 490, 541, 893, 824 and 1085.44 Mb, respectively, we identified 79341, 92089, 125443, 173964, 173604 and 222160 complete mono-nucleotide to hexanucleotide repeat sequence microsatellites with total frequencies of 226, 182, 223.6, 231.31, 235.12 and 204.42 loci per Mb, respectively ( Tables 1 , 2 ).

In the genomic SSRs of the six studied species, the distribution of microsatellite motif lengths was almost identical except in B. nigra (BB); mononucleotide, dinucleotide, trinucleotide and tetranucleotide repeats accounted for a very similar and relatively high proportion, while pentanucleotide and hexanucleotide repeats were relatively uncommon. The mononucleotide repeat motifs were the most abundant of the repeat types, with frequencies largely above 40%, and even close to 50% in B. oleracea ( Figure 2A ).

The type distribution of microsatellite motifs was almost identical in the whole genome sequences of B. rapa, B. nigra, B. oleracea, B. napus, B. juncea, and B. carinata ( Figure 2B ). In other words, the mononucleotide to hexanucleotide motifs making up the major part of the genome sequences of the six Brassica species and those that are scarce were essentially the same. From Figure 2B , it is interesting to find that among the mononucleotide repeat sequences, A has the most repetitive motifs; among the dinucleotide repeat sequences, AT has the most repetitive motifs, followed by AG; among the trinucleotide repeat sequences, AAG has the most repetitive sequences, followed by AAC; among the tetranucleotides AAAT has the most repetitive sequences; among the five and six nucleotides, AAAAT and AAAAAT are also more common than other combinations. Most of the single to hexanucleotide sequences that account for the major motifs contained abundant A\T, while the G\C motifs are all among the scarce motifs. This is in good agreement with previous reports of microsatellites identified in B. rapa, B. oleracea and B. napus. It is also clearly seen that the genomic sequences of B. rapa have a much higher content of A\T relative to C\G.

Among the whole genome sequences of B. rapa (AA), B. nigra, B. oleracea, B. napus, B. juncea, and B. carinata, the distribution pattern of the number of motif repeats of microsatellites is essentially the same, except for B. nigra, where 12 repeats have the highest proportion of all repeats. ( Figure 2C ). At the same time, we can see that the microsatellite abundance decreases significantly as the number of motif repeats increases, with the rate of change being the flattest for dinucleotides, followed by single nucleotide as well as trinucleotide repeats, and more drastic changes can be observed for long repeat motifs. ( Figures 3 , 4 ).

In addition, we compared the corresponding motif lengths ( Figure 5 ), mono- to hexanucleotide microsatellite numbers and motif repeat numbers between the basic species and the two heterotetraploid variants from which they diverged. As can be seen from the figure, the patterns of variation are very similar in the genomic SSRs of the six studied species, but comparisons between them reveal a more similar trend in microsatellite length distribution between B. rapa (AA), B. napus (AACC) and B. juncea (AABB) ( Figure 5A ); among B. nigra (BB), B. juncea (AABB) and B. carinata (BBCC), the microsatellite length distribution trends are more similar between B. nigra (BB) and B. carinata (BBCC) ( Figure 5B ); whereas among B. oleracea (CC), B. napus (AACC) and B. carinata (BBCC), the microsatellite length distribution trends are more similar between B. oleracea (CC) and B. napus (AACC) ( Figure 5C ). However, these differences are not highly significant.

By comparing the number distribution of the three allotetraploid SSR loci in the three subgenomes AA, BB and CC, we can find that the allotetraploid species have significant differences in the number of SSR loci in the genome compared with their basic diploid counterparts ( Table 3 ). It is worth mentioning that the reduction in SSR loci is the highest during the allotetraploidization of B. napus (AACC), which is 31.3% and 37%, respectively, compared with the diploid AA and CC genome. Furthermore, the reduction in SSR loci is intermediate during the allotetraploidization of B. juncea (AABB) L, which is 7.3% and 13.0%, respectively, compared with the diploid AA and BB genome. Moreover, the reduction in SSR loci is the lowest during the allotetraploidization of B. carinata (BBCC) L, which is 1.7% and 6.0%, respectively, compared with the diploid BB and CC genome ( Table 3 ). We also compared the distribution of subgenomic chromosome SSR loci in the three basic species and the three allotetraploids respectively, and found that although the number of SSR loci in most chromosomes showed a downward trend, a few chromosomes showed an increase ( Supplementary Table 1 ). For example, in B. juncea, the A subgenomic chromosomes A01 and A02 increased by 5.55% and 2.60% compared with the basic species, respectively. Moreover, in B. carinata, the B subgenomic chromosomes B01, B03 and B04 increased by 49.53%, 11.17% and 3.13%, respectively ( Supplementary Table 1 ). These results suggest that the cytological and genetic mechanisms of allotetraploid evolution are complex and worthy of further study in the future.

Based on sequencing the whole genome chromosomes of Brassica species in the triangle of U, the characteristics of microsatellites on each chromosome of each of the six considered Brassica species and their physical location were investigated.

The characteristics of length, type and number of repeats on each chromosome of the six Brassica species were consistent with the overall microsatellite characteristics of each species described above. However, the number of microsatellites distributed on each chromosome was extremely heterogeneous. For the basic species B. rapa (AA) ( Supplementary Table 2 ), B. nigra (BB) ( Supplementary Table 3 ) and B. oleracea CC ( Supplementary Table 4 ), the number of microsatellites was the highest on A09 (11214), B02 (13909) and C03 (18286), respectively. In contrast, for the four allotetraploid variants B. napus (AACC) ( Supplementary Table 5 ), B. juncea (AABB) ( Supplementary Table 6 ) and B. carinata (BBCC) ( Supplementary Table 7 ), the microsatellite numbers were the highest on C03 (12561), B02 (12452) and C01 (16036), respectively. This may have occurred because the number of microsatellites is closely related to the length of the chromosomes; the greater the length of a chromosome, the larger the number of its corresponding microsatellites.

In order to explore the exact distribution of microsatellites on each chromosome, the relationship between microsatellites and chromosomes and that between the three basic species and their corresponding allotetraploids should be analyzed more clearly. We used mapping software to locate the physical position of each microsatellite to the corresponding chromosome. The results show that, for the six Brassica species, B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. juncea (AABB), B. napus (AACC), and B. carinata (BBCC), all chromosomes have higher microsatellite frequencies at and near the ends, and they present lower microsatellite frequencies in and near the middle region. This is consistent with previous studies on the location of microsatellites on chromosomes and may correspond to the distribution around the telomeres and the thylakoids. Secondly, the physical distribution of microsatellites across all chromosomes of the six different species of Brassica is highly heterogeneous, suggesting that microsatellites do not occur randomly but their presence is most likely highly correlated with gene function around them. By comparison, we can also find that microsatellite distribution is more concentrated in B. nigra (BB) based on the physical position of microsatellites among the three basic species. Meanwhile, in the three allotetraploids, microsatellite distribution is more concentrated in B. carinata (BBCC), which may be due to the more concentrated distribution of genes on B. nigra (BB) and B. carinata (BBCC). The more concentrated distribution of genes on the latter two species is probably related. Moreover, the high concordance between microsatellites and genes strongly suggests the putative role of microsatellites in regulating genome function and in tagging genes using SSR molecules.

Using Krait software, primer pairs were successfully designed for each of the six species of B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. napus (AACC), B. juncea (AABB) and B. carinata (BBCC), respectively, yielding a total of 52356, 62290, 82984, 111276, 120324 and 144149 primer pairs named in the order of BrSSR00001 ~ BrSSR52356, BniSSR00001 ~ BniSSR62290, BolSSR000001 ~ BolSSR082984, BnaSSR000001 ~ BnaSSR111276, BjuSSR000001 ~ BjuSSR120324, and BcaSSR000001 ~ BcaSSR144149, respectively. The primer sequence, TM value, SSR motif, expected product length, and start/end position on the chromosome for each SSR marker were determined ( Supplementary Tables 8 – 13 ).

In this study, to enrich the SSR marker library of cruciferous crops and confirm the validity of the designed SSR primers, we randomly selected 120 primer pairs ( Supplementary Tables 14 – 16 ) in three basic species and used genomic DNA from eight cruciferous species as DNA template, namely, B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. juncea (AABB), B. napus (AACC), B. carinata (BBCC), R. sativus (RR), and A. thaliana (At). The cross-transferability of SSR markers from B. rapa (40 SSRs), B. nigra (40 SSRs) and B. oleracea (40 SSRs) was assessed and the affinities of the three basic species in forming three allotetraploids were speculated. We used a total of 120 SSRs, 40 from B. rapa, 40 from B. nigra, and 40 from B. oleracea for cross-amplification studies.

For the 40 primer pairs of B. rapa, a total of 310 positive amplifications were made in the eight species, resulting in a total of 325 alleles amplified with a cross-transfer rate of 85.27% ( Table 4 and Supplementary Figure 1A ). This high cross-transfer rate presumes the validity of these SSR markers for the study of the genomes of other species of cruciferous crops. Among these 325 amplified markers, the PIC values ranged from 0.32 to 0.84 with a mean value of 0.73. The gene diversity ranged from 0.41 to 0.86 with a mean value of 0.7461, and the heterozygosity values ranged from 0.21 to 0.70 with a mean value of 0.55.

For the 40 primer pairs of B. nigra, a total of 280 positive amplifications were made across the eight species, resulting in a total of 301 alleles amplified with a cross-transfer rate of 81.33% ( Supplementary Table 17 and Supplementary Figure 1B ). From this high rate of cross-transfer, we speculate that these SSR markers have some validity for genomic studies of other species of cruciferous crops. The PIC values of these 301 amplified markers ranged from 0.33 to 0.73 with a mean value of 0.50. The gene diversity ranged from 0.38 to 0.71 with a mean value of 0.58, and the heterozygosity values ranged from 0.10 to 0.53 with a mean value of 0.41.

For the 40 primer pairs of B. oleracea, a total of 303 positive amplifications were made across the eight species, resulting in a total of 310 amplified alleles with a cross-transfer rate of 73.45% ( Supplemntary Table 18 and Supplementary Figure 1C ). This indicates that the SSR cross-transfer rate of B. oleracea has limitations in terms of its validity for studying the genomes of other species of cruciferous crops. Among these 310 amplified markers, the PIC values ranged from 0.48 to 0.85 with a mean of 0.67. The gene diversity ranged from 0.54 to 0.81 with a mean of 0.64, and the heterozygosity values ranged from 0.0 to 0.71 with a mean of 0.41.

When comparing the polymorphism potential of B. rapa (85.27%), B. nigra (81.33%) and B. oleracea (73.45%), B. rapa was found to have a higher cross-transfer rate (85.27%). In addition, the mean values of PIC, genetic diversity and heterozygosity of B. rapa-derived SSR markers were relatively high, indicating that among cruciferous plant species, B. rapa has better polymorphic potential than B. nigra and B. oleracea. The cross-species transferability has been demonstrated in Brassica crops, while the degree of SSR cross-transfer depends on the evolutionary distance among species (Thakur et al., 2022).

A B-genome specific SSR marker, BniSSR23228, was obtained from 40 selected SSR primers of black mustard ( Supplementary Figure 1B ). After PCR amplification, polyacrylamide gel electrophoresis, cloning verification screening and sequence alignment were carried out to validate the existence of this specific SSR marker ( Figure 6 ). The validity experiment results indicated that BniSSR23228 possessing (AAGGA)₃ sequence characteristics located in chromosome B3 with a total length of 97 bp ( Supplementary Figure 2 ). Subsequently, this molecular marker can effectively screen the B genome of Brassica, and can be used for variety and parent identification, introgression line tracking, genetic diversity analysis, and marker-assisted breeding.

Krait software was used to search for 79341, 92089, 125443, 173964, 173604, and 222160 SSR loci from the whole genome sequences of six species of Brassica, namely, B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. napus (AACC), B. juncea (AABB), and B. carinata (BBCC), respectively. The frequency of SSR occurrences (average SSRs per Mb) was 226 loci/Mb, 182 loci/Mb, 223.6 loci/Mb, 231.31 loci/Mb, 235.12 loci/Mb and 204.42 loci/Mb, respectively. In a previous report, the PERL5 script MIcroSAtellite (MISA; http://pgrc.ipkgatersleben.de/misa/) was employed for the genomes of B. rapa, B. oleracea and B. napus to obtain 140998, 229389 and 420991 SSR markers, respectively (Shi et al., 2014). Using Karit enabled SSR identification and subsequent primer design in less time than MISA, and the long microsatellites identified in this way were more polymorphic and useful. Among the SSR single nucleotide sequences identified in the six species, B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. napus (AACC), B. juncea (AABB), and B. carinata (BBCC), the A sequence repeats were present in 29518 (42.5%), 31308 (34.00%), 57629 (45.94%), 73785 (42.5%), 77106 (44.32%), and 89382 (40.23%) single nucleotides, respectively. They were therefore regarded as the most important of such repeats, while the C sequence was less represented in single nucleotides. This result is consistent with previous studies performing SSR analysis of the whole grapevine genome (Cai et al., 2009). For the dinucleotide repeat type AT, the number of repeats were 13915 (20.04%), 22769 (24.72%), 27267 (21.74%), 37598 (21.66%), 29808 (17.13%) and 54026 (24.32%) respectively. This is therefore considered as the most significant of this replicate type, consistent with previous studies using SSR analysis of the whole B73 maize genome (Qu and Liu, 2013). The frequency of single nucleotide repeats was the highest of all repeat types, a result that differs from previous SSR analysis studies using MISA in B. rapa, B. oleracea and B. napus (Shi et al., 2014). Moreover, Shi et al. study results showed the microsatellite frequencies of Brassica, Arabidopsis and other angiosperm species were significantly negatively correlated with both their genome sizes and transposable elements contents (Shi et al., 2013). Qin et al. (2015) investigated the evolutionary regularities of SSRs during the evolution of plant species and the plant kingdom by analysis of twelve sequenced plant genome sequences. The results showed that, SSRs not only had the general pattern in the evolution of plant kingdom, but also were associated with the evolution of the specific genome sequence.

Probably, the deviation may be due to differences in the SSR software algorithms, parameter settings and the original databases used. In addition, our data indicated that the allotetraploidization process resulted in a significant reduction in SSR loci in the three subgenomes AA, BB and CC. The reasons may be partial gene-dominated chromosomal homologous recombination and rearrangement during the evolution of basic diploid species into allotetraploids (Song et al., 2021). Meanwhile, there are a large number of transposable elements (TE) in the Brassica species in the triangle of U genome (Cai et al., 2022), and TE insertion seems to result in chromosomal translocation, leading to the reduced number of SSR loci in the three subgenomes AA, BB and CC during the process of allotetraploidization. Of course, further experiments are required to prove this hypothesis. Shi et al. (2014) carried out microsatellite characterization based on genome-wide and marker development in three recently sequenced Brassica crops, and suggested that the distribution pattern of microsatellites may be conserved in the genus Brassica. This view was reinforced by the use of Krait to identify SSR signatures of six species of Brassica, suggesting that the distribution patterns of microsatellites are likely to be conserved in all Brassica species. Thus far, the comprehensive identification, characterization and primer development of SSRs for six Brassica species in the triangle of U have not been carried out. However, in this study, based on the complete whole genome sequences of six Brassica species in the triangle of U, not only were the SSRs of each variety comprehensively analyzed, but also the differences between the Brassica crops in the triangle of U were compared and a comprehensive primer design for the SSRs was carried out. To our knowledge, this is the first report to identify the SSR loci and design the SSR primers based on the complete whole genome sequences of six Brassica species in the triangle of U together. These markers will act as a powerful tool for future genomic and genetic studies of Brassica cruciferous crops in the near future.

High transferability has been reported for SSRs of different plant species, such as “Chiifu” of B. rapa, that is, 95% of its SSRs could amplify a fragment of other species (Wang et al., 2011). Thakur et al. (2018) study result indicated 100% cross-transferability was obtained for B. juncea and three subspecies of B. rapa with 124 Brassica-derived SSR loci assayed, while lowest cross-transferability (91.93%) was obtained for Eruca sativa. The average % age of cross-transferability across all the seven species was 98.15%. In addition, 47% of EST-SSR markers developed from B. rapa, B. oleracea, and B. napus were transferable to six Brassica species (An et al., 2011). Sim et al. randomly selected 41 SSR markers of thistle and alfalfa, and found that the transferability was 53% to 71% in the leguminous plant (alfalfa) and 33% to 44% in the non-leguminous plant (thistle). About 57% of cereal EST-SSRs could also be amplified in ryegrass (Sim et al., 2009). Additionally, about 60% of EST-SSR markers from barley could be amplified in wheat and rye (Castillo et al., 2008). Cui et al. (2005) used 69 pairs of SSR primers of non-heading Chinese cabbage in eight varieties of Brassica crops, and found that the transferability amplification rate was 49.3% to 85.5% and that 33% of the SSR primers in the inter-specific hybrids of Brassica presented abundant diversity.

Based on the 1176 SSR-containing ESTs in cabbage, a total of 978 primer pairs have been successfully designed and assessed by validation of the amplification on two inbreed lines (Chen et al., 2010). Subsequently, the results indicated that the developed SSRs from ESTs of B. oleracea were valid and practicable in marker-assisted selection and QTL analysis in cabbage (Su et al., 2015). Some useful information about SSR and sequence analysis in Brassica crops can also be obtained on the website of Brassica DB database (http://Brassica.bbsrc.ac.uk/).

In this work, the functional utility of SSR markers derived from B. rapa (AA), B. nigra (BB) and B. oleracea (CC) was evaluated by analyzing their cross-transferability among B. rapa (AA), B. nigra (BB), B. oleracea (CC), B. juncea (AABB), B. napus (AACC), B. carinata (BBCC), R. sativus (RR), and A. thaliana (At). From our results, it was inferred that the cross-transferability of SSR markers from B. rapa (AA) showed higher potential than those from B. nigra (BB) and B. oleracea (CC) among these eight species, with cross-transferability rates of 85.27%, 81.33% and 73.45%, respectively. In fact, enriching other varieties with SSR markers alleviates the hassle of the expansion and development process and can facilitate the genetic improvement of new varieties by the genomes of superior varieties. Our findings suggest that genomic SSR markers with high transferability can be used for different Brassica species and even non-Brassica species. Therefore, these genomic SSR markers with clear location and uniform nomenclature system have high potential to be more widely used in several fields, such as gene localization, genetic mapping, evolutionary analysis, molecular marker-assisted breeding, and provide marker materials for genetic and comparative genomics analysis to further introduce some important agronomic traits into other superior Brassica species lacking these traits.

Alien chromosome additions have been used to link species-specific characteristics to particular chromosomes (Kapoor et al., 2011). The plasticity of the Brassica genome and existence of natural amphiploids have made it possible to develop several alien chromosome additions by dissecting the B. rapa, B. oleracea, and B. nigra genomes (Chevre et al., 1996; Gu et al., 2009; Li et al., 2013). Compared with the single species-specific SSR marker obtained by comparison between two species developed in previous studies (Gu et al., 2009; Li et al., 2013), the specific SSR developed in this study, BniSSR23228, has been verified among Brassica species in the triangle of U and their closely related species, radish and Arabidopsis. The results revealed that it is a B-genome specific SSR marker, therefore has more significant B-genome specificity and more extensive application value in the future.