Deconstructing Molecular Phylogenetic Relationship Among Cultivated and Wild Brassica Species

Abstract

Brassica represents an agriculturally important and diverse group of oilseed crops with a long evolutionary history. Various molecular markers played an important role in understanding origin and evolution of Brassica species. In present research both Single Nucleotide Polymorphisms (SNPs) and Simple Sequence Repeats (SSRs) developed form Brassica juncea were used to find out the phylogenetic relationship between various cultivated and wild Brassica species. A total of 88 SSR and 58 SNP markers were found to be functional across 38 genotypes belonging to ten different taxon groups. The polymorphic markers were able to group the genotypes into three different clusters and showed relatedness among different genomes based on genetic distance. The transferability of these markers serves the purpose of their quick use in cultivar identification, diversity and phylogenetic analysis in those orphan crops species where no or less genomic information is available.

Introduction

Brassicaceae is a highly diverse family that comprises of agriculturally important species including model plant (Arabidopsis thaliana), highly cultivated oilseed crops (Brassica napus, B. rapa, B. carinata and B. juncea), widely utilized vegetable species (B. oleracea), eatable roots (Raphanus sativus) and various herbs (B. nigra and Sinapis alba). This broad array of diversity is mainly because of genetic manipulations caused due to natural evolution of polyploids and amphidiploid hybrids, and traditional breeding including selection followed over the time (Edward et al. 2011). The genus Brassica has a long evolutionary history consisting of amphidiploid hybrids that are formed by the interspecific crosses between diploid progenitors.

The Brassicas had long suffered a narrow diversity bottleneck, and cross species incompatibility have led to limited use of related species for the improvement of Brassicas. Most of the interspecies and intergeneric hybrids between Brassica and its related species were carried out to transfer important traits from related wild species to some cultivated Brassica species. Among the Brassica spp., B. juncea, also known as Indian mustard, is a dominant species grown on Indian subcontinent for seed oil. It is an amphidiploid (2n=4x=36; AABB) thought to have developed through hybridization of its progenitors followed by the natural doubling of chromosomes (Nagaharu, 1935). Among the three genomes (A, B and C), the AA genome has most variation in chromosome size and morphology and therefore most of the variation in B. juncea can be attributed due to AA genome (Kulak et al. 2002).The diversity present in Brassica spp. governs its usability for breeding purpose and further improvement of the cultivated crop. A plethora of molecular markers had been reported to be used in analysis of the molecular diversity and population genetics studies. Among different marker system used, SSRs (Simple Sequence Repeats) and SNPs (Single Nucleotide Polymorphisms) are more promising in relating the patterns of closeness in different Brassica species (Sudan et al. 2016, Li et al. 2019). As such, huge emphasis has been laid on the development of genic SSR and SNP markers since they find primary application in high-resolution genetic map construction, association mapping, genetic diagnostics, genetic diversity analysis, phylogenetic analysis, and characterization of genetic resources (Singh et al. 2018). A large number of studies had been reported in which SSR markers have been developed in some Brassica spp. and their cross transferability had been explored in other Brassica spp. (Singh et al. 2016, Thakur et al. 2018, Li et al. 2019). However, no attempts had been made to explore the efficiency of cross species transferability of SNP as well as SSR markers from B. juncea to wild Brassica spp. The transferability of these markers serves the purpose of their quick use in cultivar identification, diversity and phylogenetic analysis in those orphan crops species where no or less genomic information is available.

In the present study, we used both SNP and SSR markers developed from Brassica juncea and evaluate them for population, genetic diversity and evolutionary relationship studies in wild Brassica spp.

Materials & Methods

2.1 DNA extraction and dd-RAD library preparation

Young leaves from Brassica genotypes and wild Brassica relatives were used for the isolation of high quality genomic DNA using SGS buffer method (Sudan et al. 2017). A total of six B. juncea genotypes (Pusa tarak, Urvashi, RSPR 1, Zem 1, Donskaja 4 and EC 287711) were used for the dd-RAD library preparation using the modified protocol (Yang et al. 2016). The pooled dd-RAD library was then sequenced through high throughput Illumina HiSeq 2000 using 100 bp paired-end reads. The ddRAD-seq reads were subjected to a pipeline analysis using CLC Genomics workbench to obtain a high quality SNP set from genic regions. A subset of SNP markers from B. juncea were then used to check the efficiency of cross-species transferability of markers in 32 Brassica relatives (Table 1). Out of 32 genotypes, 14 belongs to S. alba, five from B. rapa, four from R. sativus, two each of Eruca vesicaria, Eruca sativa and B. oleracea and one each from B. napus, B. tournefortii and Brassicoraphanus. The seeds were procured from Department of Genetics and Plant Breeding, SKUAST- Jammu, India and Plant Gene Resources, Agriculture and Agri-Food, Canada.

Table 1. List of genotypes used in the present study

*NS- not sequenced

2.2 SNP development and genotyping

A subset of 61 genic SNPs was selected with a view of taking 3-4 markers from each chromosome. Using the flanking sequences of SNP, forward, reverse and iPLEX universal extension primer were designed using AgenaCX assay design suite V2.0 software. The forward, reverse and iPLEX universal extension primers were diluted in the concentration of 100µM, 100µM and 500µM respectively and were used for genotyping cultivated and wild Brassica species using Sequenom’s MassARRAY manufacturer protocol. The genotyping calls were evaluated using MassARRAY TYPER 4.0 software.

2.3 SSRs development and genotyping

The 100 SSR markers that were developed across the B. juncea A-genome (unpublished data) were also used to study genetic variation in wild Brassica relatives. For SSR marker analysis, a two-step thermal amplification profile was followed during PCR amplification. The first 10 cycles were carried out at T_m-1 and next 25 cycles were carried out at T_m-3. The amplified products were then resolved on 3.5% agarose gel containing ethidium bromide by using a 100 bp ladder. Bands were scored according to the presence of different alleles at same locus.

2.4 Diversity and Population structure analysis

The molecular data from SNP and SSR genotyping were arranged in a matrix format and used for various parameters of genetic diversity analysis using Powermarker v3.51 (Liu and Muse, 2005). The molecular data was also used for phylogenetic tree construction using Darwin5 software (Perrier and Jacquemoud-Collet, 2006). The data was used for calculation of dissimilarity coefficient, which in turn was used for tree construction using unweighted neighbour-joining method. Bootstrap value for this tree was determined by re-sampling loci at 1000 times. For the optimum value of K, the membership coefficient from STRUCTURE were integrated into CLUMPP software to generate a Q matrix (Jakobsson and Rosenberg, 2007) which is then used to draw a bar plot using DISTRUCT software (Rosenberg, 2004).

Results and Discussions

3.1. Across genome transferability of SNP and SSR markers

Both SSR and SNP markers were used on related species of Brassica juncea from Brasicaceae family to understand the genetic relationship between different taxon genomes based on the ability of markers to detect corresponding loci in related species. A total of 88 SSR and 58 SNP markers were found to be functional across 38 genotypes belonging to ten different taxon groups (Supplementary Table 1). The SNPs used in the current study have revealed high cross-genome transferability of more than 59% across the species. Interestingly, the SNPs developed from amphidiploid genome AABB were found to be highly transferable to distantly related species genomes such as EE, CC, RR and SS. The highest transferability (93.17%) was observed in case of close relatives - B. napus (AACC) and B. oleracea (CC); and lowest (63.8%) was observed for R. sativus. For B. tournefortii the cross transferability was observed to be 70.68%; although, Prakash and Narain (1971) designated B. tournefortii as D genome due to low cross compatibility, hybrid sterility and very little gene flow. When the SNP markers from individual genome of B. juncea (A and B) is taken into account, the B-genome SNP loci show more transferability to wild Brassica species as compared to A-genome derived markers where none of the two genomes are available (Fig. 1). A-genome derived markers from B. juncea shows high rate of transferability in all those species where A genome is present suggesting a high conservancy of this genome during the evolution period.

The SSR markers from AA-genome were found to be highly transferable to the related genomes of different taxon included in the current study. The extent of transferability ranged from 78% to 100% (Supplementary Table 1). The high rate of SSRs transferability reveal highly syntenic genomes in the most widely used taxon of the Brasicaceae family. The percent cross-transferability obtained in present investigation is in accordance with some recent studies (Thakur et al. 2018, Singh et al. 2018). However when the transferability potential of both the markers were compared, the SSRs seems to be far better in studying the population genetics in orphan crops species by exploiting multi-allelism among the genomic resources of cultivated species. Moreover, as the SNP loci were selected from the genic region, so some of these regions might be highly conserved for a particular species.

3.2. Genome-wise allelic patterns

The number of alleles detected by SSR markers ranged from 89 alleles (DD genome; B. tournefortii) to 152 alleles (SS genome; S. alba). The SSR markers used in the present study have been developed using sequence information from diploid progenitor (B. rapa) contributing AA genome to B. juncea. Out of 88 SSR markers, thirty one (31) markers detected private alleles among six taxon groups with the most private alleles among genotypes from SS (Sinapis alba) and EE (Eruca species) taxons. Taxons with AA, CC, AACC and DD genomes did not carry any private allele. Detection of highest number of 16 private alleles among SS taxon indicate either a selection pressure experienced by it or it could be due to it being a descendent of ‘nigra’ lineage.

The number of alleles detected by 41 SNPs among 10 taxon groups ranged from 35 (DD genome; B. tournefortii) to 75 (SS genome; Sinapis alba). Like SSRs, the SNP markers too detected highest number of alleles in DD genome of B. tournefortii; but contrary to high number of private alleles in SSRs, there were only three SNPs that were reported to have detected private alleles among SS (Sinapis alba) and RR (Raphnus sativus).

3.3. Molecular genetic diversity

Out of total functional SNP markers, only 41 with an average missing data less than 40% were considered for diversity analysis due to software requirement. As a result of biallelic nature of SNPs, a total of 82 alleles were amplified. The minor allele frequency ranged from 0.013 to 0.48 with an average of 0.235. The gene diversity and heterozygosity value was also able to identify the variability among the genotypes. The gene diversity ranged from 0.026 to 0.496 and heterozygosity level of markers ranged from 0.029 to 0.810 (Supplementary Table 2). The PIC (Polymorphism Information Content) value ranges from 0.025 to 0.375 with an average of 0.244 (Figure 1).

The 88 polymorphic SSR markers were able to amplify 252 alleles among 38 genotypes and in many cases, the amplified PCR product is different from the expected size as obtained in case of species having A progenitor genome. The AA-genome SSR markers were used for the amplification of corresponding loci from AA-genome containing amphidiploid species B. juncea (AABB) and B. napus (AACC). The SSR analysis revealed that the alleles for nearly 74% SSRs were amplified in different size (80-400bp) range among the two amphidiploid genotypes. The results indicated that present day A-genome in these two amphidiploids is diverse from each other both at non-coding (Thakur et al. 2018) and coding sites. This might be due to the fact that the AA-genome in these amphidiploids had evolved under different selection pressure after originating from the parental progenitor species (B. rapa). The number of alleles detected at each locus ranged from one to five with an average of 2.97 alleles per locus, with a size range of 80-400 bp reflecting a wide variation among repeat regions of different alleles. The gene diversity ranged from 0.027 to 0.694 and heterozygosity level of markers ranged from 0.026 to 0.833 (Supplementary Table 3). The PIC (Polymorphism Information Content) value ranged from 0.027 to 0.782 with an average of 0.478 (Figure 2). The average SSRs PIC value is more than the SNPs which relies that SSRs are more informative as compared to SNP markers. Moreover, the high PIC values are also contributed due to SSRs being multi-allelic while SNPs are almost always bi-allelic in nature. The diversity within a collection of germplasm depends upon the degree of relatedness, origin of individual genotypes and the types of markers used to estimate allelic information at different loci. The SSR markers obtained from non-coding DNA tend to uncover higher genetic diversity than SNP markers that are obtained from highly conserved coding regions of a genome. However, SNPs from conserved regions are more likely to be involved in phenotype causal relationship.

3.4 Phylogenetic relationship and Population Structure

In order to see the efficacy of these two markers in determining the genetic distance between various species of Brassicacaeae, a dendrogram based on unweighted neighbor joining method was constructed. Both the markers were able to grouped 38 genotypes into three major clusters {SS, (A, B, C, E, T) and (R, AR)} depending upon the difference in their genome composition. The cluster I, II and III represent SS genome group (S. alba); A-, B-, C-, E-, T- genome group and R- genome group genotypes respectively. The clustering indicated the ability of these molecular markers to form grouping of related genotypes from a genome with high level of accuracy. In case of SNP markers, the cluster I consists of genotypes from S. alba (SS) and cluster III contains genotypes from R. sativus (RR) and Brassicoraphanus (AARR) (Figure 1c). As such the genetic distance of these three species is far away from the core species. S. alba seem closest to B. napus (0.472) and quite distinct from B. juncea (0.563) and B. rapa (0.548) (Supplementary Table 4). Brassicoraphanus was formed from the combination of two genome (AA and RR), but it shows more genetic closeness to R. sativus(RR) (0.243) when compared to AA genome species i.e. B. rapa (0.423) and B. juncea (0.510). The cluster II consists of genotypes from E. sativa/vesicaria, B. juncea, B. tournefortii, B. oleracea, B. napus and B. rapa. E. sativa/vesicaria was found to be closer to B. juncea (0.41) than to B. rapa (0.424), B. napus (0.432) and B. oleracea (0.514). Interestingly, B. tournefortii (TT genome) that showed low cross transferability of markers, was tend to found closer to B. juncea (0.432) than to B. oleracea (0.543) and other Brassica spp. SSR markers also showed nearly the same clustering pattern (Figure 1c) but however the genetic distance between the species was large when compared to SNP markers (Supplementary Table 5). As the SNP markers in the present study was derived from genic region and SSRs were mostly obtained from non-genic region, so the SNP loci shows more conserve nature due to low mutation rate in the genic region as compared to non-genic region. Population structure estimated using STRUCTURE V2.3.4 software under the Hardy-Weinberg Equilibrium also clustered 38 genotypes into three (SSR makers) and seven (SNP markers) groups based on the maximum likelihood and delta K (ΔK) (Figure 1d and 2d).

Conclusion

This is the first report of the use of both SSR and SNP markers derived from B. juncea to be used in wild Brassica for their phylogenetic relationship with the core species. Interestingly, the markers from both A and B genome are able to show amplification in diverse genomes such as E, R and T. This clearly indicates that the various species in the family Brassicaceae are closely related and share some unique sequences that remain conserved during the course of evolution. The present study also provides various molecular markers that would be useful in orphan Brassica crops for their various genetic studies including fine mapping and association analysis.

Declarations

Funding

The present work was supported by the funds received as research grant (BT/PR3946/AGR/2/839/2011 from 2013 to 2016) from Department of Biotechnology, Government of India, New Delhi.

Conflicts of Interest

The authors declare that they have no competing interests.

Availability of data and material

Supplementary data to this article is available under appendix.

Code Availability

Not applicable

Authors’ contributions

RS design the experiment; JS and RS performed the experiments and written the manuscript; RM and RKS helped in the analysis of various parameters of genetic diversity.

Ethics approval

Not applicable

Consent for publication

All the authors agreed to publish in this journal and in present form.

Acknowledgements

We would like to thank Dr. Manmohan Sharma of School of Biotechnology, SKUAST-Jammu for generously sharing aliquots of SSR markers.

References

1. Edwards D, Batley J, Parkin I, Kole C (2011) Genetics, genomics and breeding of oilseed Brassicas. CRC Press.
2. Jakobsson M, Rosenberg NA, (2007) CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23(14):1801-1806.
3. Kulak S, Hasterok R, Maluszynska J (2002) Karyotyping of Brassica amphidiploids using 5S and 25S rDNA as chromosome markers. Hereditas 136:144–150.
4. Li P, Su T, Yu S, Wang H, Wang W, Yu Y, Zhang D, Zhao X, Wen C, Zhang F (2019) Identification and development of a core set of informative genic SNP markers for assaying genetic diversity in Chinese cabbage. Horticulture, Environment, and Biotechnology 60(3):411-425.
5. Liu K, Muse SV (2005) PowerMarker: an integrated analysis environment for genetic marker analysis. Bioinformatics 21(9):2128-2129.
6. Nagaharu U (1935) Genome analysis in Brassica with special reference to the experimental formation of B. napus and peculiar mode of fertilization. Japan. J. Botany7:389–452.
7. Perrier X, Jacquemoud-Collet JP (2006) DARwin software http://darwin.cirad.fr/darwin
8. Prakash S, Narain A (1971) Genomic status of Brassica tournefortii Gouan. Theoretical and Applied Genetics 41(5):203-204.
9. Rosenberg NA (2004) DISTRUCT: a program for the graphical display of population structure. Molecular Ecology Notes 4(1):137-138.
10. Singh BK, Choudhary SB, Yadav S, Malhotra EV, Rani R, Ambawat S, Pandey A, Kumar R, Kumar S, Sharma HK, Singh DK (2018) Genetic structure identification and assessment of interrelationships between Brassica and allied genera using newly developed genic-SSRs of Indian Mustard (Brassica juncea L.). Industrial Crops and Products 113:111-120.
11. Singh BK, Mishra DC, Yadav S, Ambawat S, Vaidya E, Tribhuvan KU, Kumar A, Kumar S, Kumar S, Chaturvedi KK, Rani R (2016) Identification, characterization, validation and cross-species amplification of genic-SSRs in Indian Mustard (Brassica juncea). Journal of Plant Biochemistry and Biotechnology 25(4):410-420.
12. Sudan J, Khajuria P, Gupta SK, Singh R (2016) Analysis of molecular diversity in Indian and Exotic genotypes of Brassica juncea using SSR markers. Indian J. Genet. 76(3):361-364.
13. Sudan J, Raina M, Singh R, Mustafiz A, Kumari S (2017) A modified protocol for high-quality DNA extraction from seeds rich in secondary compounds. Journal of Crop Improvement 31(5):637-647.
14. Sudan J, Singh R, Sharma S, Salgotra RK, Sharma V, Singh G, Sharma I, Sharma S, Gupta SK, Zargar SM (2019) ddRAD sequencing-based identification of inter-genepool SNPs and association analysis in Brassica juncea. BMC Plant Biology 19(1):1-15.
15. Thakur AK, Singh KH, Singh L, Nanjundan J, Khan YJ, Singh D (2018) SSR marker variations in Brassica species provide insight into the origin and evolution of Brassica amphidiploids. Hereditas 155(1):6.
16. Yang GQ, Chen YM, Wang JP, Guo C, Zhao L, Wang XY, Guo Y, Li L, Li DZ, Guo ZH (2016) Development of a universal and simplified ddRAD library preparation approach for SNP discovery and genotyping in angiosperm plants. Plant Methods 12(1):39.