Developed KASPar assays for selection of low erucic acid are high throughput, robust, and codominant capable to track recessive alleles in their heterozygous state.

Indian mustard [Brassica juncea (L.) Czern and Coss] is an important source of edible and industrial oil in the Indian subcontinent. Its seeds contain nearly 40% oil on a seed weight basis. Similar to other plants, oil in Indian mustard is a complex mixture of triglycerides (>95%) along with diacylglycerols (<5%), phytosterols, etc. Though mustard oil contains seven major fatty acids, erucic acid, a very long chain fatty acid (C22:1) is the predominant fatty acid contributing more than 50% to the total pool of fatty acids. Such levels are undesirable in view of their perceived role in cardiac lipidosis and fatty deposits in skeletal muscles (Vogtmann et al., 1975; Burrows and Tyrl, 2013). This concern precipitated substantive efforts to genetically reduce the erucic acid content in different rapeseed-mustard crops. Canadian rapeseed breeders were first to develop cultivars with low erucic acid (<2%) in B. napus and B. rapa (Downey and Craig, 1964). Combining low erucic acid trait with low glucosinolate content (<30 μmol/g defatted meal) in the meal subsequently aided the development of canola quality rapeseed cultivars. Similar success was achieved in B. juncea with the discovery of low erucic acid (<2%) genotypes (Zem 1 and Zem 2) in Australia (Kirk and Oram, 1981). Only a few varieties and hybrids of canola mustard are currently under cultivation. However, linkage drag continues to be an issue as low erucic acid donors (Zem 1 and Zem 2) and their derivatives are late maturing, small seeded, and possess low seed oil content (Banga et al., 2015). Traditional breeding methods have not been able to eliminate limitations of low oil in the presently available low erucic acid mustard varieties. Transferring the low erucic acid trait into agronomically superior genotypes is difficult, since the trait is double recessive and fatty acid composition is defined by the genotype of the embryo. Each cycle of backcrossing requires one generation of selfing after every backcross. Plants carrying low erucic acid alleles in the self-progenies are selected on the basis of half-seed analysis for the next cycle of recurrent backcrossing to the recurrent parent. A countless number of half seeds must be analyzed in each generation and this procedure is beyond the analytical capacity of most laboratories in mustard-growing countries. Although many rapid screening techniques (e.g., Near-infrared spectroscopy) are now available (Kaur et al., 2016; Sen et al., 2018), these require a large sample size and are not suitable for the analysis of single seeds. The development of functional markers with the ability to differentiate between homozygous and heterozygous states of plants is considered important. It will obviate the wet chemistry analysis for fatty acid profiling and the necessity of selfing after every generation of backcrossing.

The genes involved in the biosynthesis of very long-chain fatty acids (VLCFAs) are well known. Long-chain fatty acids are synthesized de novo with two enzyme systems, acetyl-CoA carboxylase (ACCase) and fatty acid synthase (FAS) complex. ACCase catalyzes the formation of malonyl-acyl carrier protein (ACP) (Harwood, 1996), whereas FAS complex catalyzes the elongation of malonyl-ACP into palmitoyl-ACP with seven rounds of four sequential reactions. Elongation of palmitoyl-ACP results in stearoyl-ACP, which is then saturated by plastidal stearoyl-ACP desaturase to produce oleoyl-ACP. The removal of ACP by acyl-ACP thioesterase (FatA/FatB) marks the termination of elongation in plastids and leads to the formation of free fatty acids such as palmitic, stearic, and oleic acid. Most of the free fatty acids are transported to endoplasmic reticulum, where oleic acid is desaturated and elongated to produce polyunsaturated fatty acid and VLCFA. Fatty acid desaturases (FAD) encoded by FAD genes catalyze the desaturation reaction (Kunst et al., 1992; Miquel and Browse, 1992), whereas β-ketoacyl-CoA synthase (KCS) catalyzes the first step of two rounds of four sequential chain elongation reactions (James et al., 1995). This enzyme is encoded by the fatty acid elongase 1 (FAE1) gene in Brassica. Mutation in the FAE1 genes impairs the activity of KCS enzyme (Katavic et al., 2002). Orthologs of the FAE and FAD genes were also reported in different oilseed Brassica species (Barker et al., 2007; Wang et al., 2015; Tang et al., 2019; Xue et al., 2020). B. juncea (AABB) and B. napus (AACC) possess two homologs of FAE1.1 and FAE1.2 genes. These show more than 99% nucleotide identity. High homologies exist between the gene sequences of the gene from B. juncea, B. rapa, and B. napus (Zeng and Cheng, 2014). Four and three single nucleotide polymorphisms (SNPs) in FAE1.1 and FAE1.2 differentiated low erucic genotypes from high erucic acid ones in B. juncea (Gupta et al., 2004). A total of 26 SNPs caused changes in 13 amino acids in B. rapa (Yan et al., 2015). Deletion of two bases in C homolog in B. napus produced low erucic acid phenotype (Fourmann et al., 1998). Single nucleotide primer extension (SNuPE) assays have been used to track allelic variation in the FAE1 gene (Gupta et al., 2004), but these are expensive and tedious to implement. Cleaved amplified polymorphic sequences were also used to characterize FAE1.1 and FAE1.2 alleles in B. juncea (Saini et al., 2016), but CAPS markers are dominant and these fail to identify recessive alleles in heterozygous state. Recently, Saini et al. (2019) have reported PCR markers based on the sequence variation in the promoter regions of the FAE1. These are gel-based markers. A marker-assisted breeding program requires simple but robust molecular assays that benefit from economies of scale. Kompetitive allele-specific PCR (KASPar) genotyping assay is considered as an ideal approach to hasten the breeding processes through rapid genotyping of thousands of samples at very low error rates (Chen et al., 2010; Kumpatla et al., 2012). In this study, we report the development of KASPar assay for high-throughput genotyping of erucic acid content. These allele-specific assays were developed on the basis of SNPs at positions 735 and 1,476 from A and B homologs of FAE1.1 and FAE1.2. These SNPs were identified on the basis of Sanger sequencing of cloned FAE1.1 and FAE1.2 genes from both the high and low erucic acid accessions of B. juncea and the progenitor species: B. rapa and B. nigra. All these sequences have been submitted to the National Center for Biotechnology Information (NCBI) and included in Supplementary File 1. The practical utility of the designed KASPar assays was established by validating these in a germplasm collection and a segregating backcross progeny varying for erucic acid contents.

Genetic modification of oil quality is now a major breeding objective in mustard. However, most breeding groups are following traditional crop-breeding methods with little application of marker-assisted breeding in India. This is in spite of the availability of many high-density linkage maps and molecular marker systems in the species (Pradhan et al., 2003; Ramchiary et al., 2007; Panjabi et al., 2008; Yadava et al., 2012). KCS enzyme encoded by the FAE1 is involved in one of the step for the synthesis of erucic acid from oleic acid. The mutation in coding region of the FAE1 gene impairs its activity and results in drastic reduction of erucic acid (Katavic et al., 2002). Molecular markers, SNuPE (Gupta et al., 2004), and CAPS (Saini et al., 2016) for tracking the allelic variation for the FAE1 gene, have been developed in B. juncea. Recently, a PCR-based marker system has been developed, which exploits the variation in the promoter region of the FAE1 (Saini et al., 2019). None of these systems allow high throughput and cost-effective genotyping. It is, thus, important to develop breeder friendly and robust marker system.

We independently isolated FAE1.1 and FAE1.2 from four genotypes of B. juncea, two of B. rapa, and one of B. nigra. Cloning and gene sequencing allowed us to identify one copy each of FAE1.1 in B. rapa and FAE1.2 in B. nigra. Both the genes (FAE1.1 and FAE1.2) were also detected in allotetraploid B. juncea as it has been reported earlier (Das et al., 2002; Gupta et al., 2004; Yan et al., 2015). Comparative sequence analysis has also indicated that the coding sequences (CDSs) from the FAE1 are highly conserved with high similarity across the tested Brassica species. A broader genetic conservation of the gene has also been reported in the related Brassicaceae species, Sinapis alba (Zeng and Cheng, 2014). The predicted amino acid sequences of Sinapis alba have shared high homology with cultivated Brassica and Arabidopsis. Our studies have also emphasized the role of point/missense mutations in the coding region to inactivate KCS. Differences in the regulatory region or CDS due to SNPs, Insertion/deletions (InDels), and insertion of transposable elements cause dysfunction in the FAE1 (Roscoe et al., 2001; Katavic et al., 2002; Chiron et al., 2015). In contrast, a deletion of four base pairs in the FAE1 CDS from B. napus has produced a frameshift mutation and led to the formation of non-functional polypeptide with short length (Wu et al., 2008). Serine at 282 positions in the FAE1 is considered essential for elongase activity in B. napus (Katavic et al., 2002). We have also observed transition and transversion types of nucleotide variations at eight positions in FAE1.1 gene sequences of B. rapa. Of these, seven positions have been identified as transition changes and three of which cause amino acid alterations to inactivate KCS. These transition types of alterations have also been reported to cause low erucic acid in B. rapa (Wang et al., 2010). A total of 26 SNPs have caused changes in 13 amino acids and consequently loss of activity of KCS has also been reported in B. rapa (Yan et al., 2015). Whereas for B. juncea, total six nucleotide variations were detected in FAE1.1 gene and all were of transition types. Comparatively, less nucleotide variations in FAE1.1 of allotetraploid B. juncea were identified as compared to its diploid progenitor, B. rapa. Of these variations, four SNPs are common across the ploidy. These SNPs have also been reported earlier by Gupta et al. (2004) in B. juncea. For FAE1.2, nucleotide variations have occurred at the three positions both in diploid, B. nigra and allotetraploid, B. juncea. However, only one of these SNPs caused amino acid alteration that led to the loss of function of KCS. SNPs resulting from transition changes in FAE1.2 gene have also been documented in B. juncea (Gupta et al., 2004). Low sequence diversity was recorded for B homologs of the FAE1 genes in diploid and allotetraploid crop Brassica species. SNPs and InDels in gene or genome are being exploited for marker-assisted breeding (Garcés-Claver et al., 2007).

In this study, we planned to develop high throughput and breeder friendly markers capable of discriminating homozygous and heterozygous individuals in the segregating progenies. Effectiveness of the KASPar assays for high-throughput genotyping has already been documented in many crops including Brassica (Chhetri et al., 2017; Steele et al., 2018; Chen et al., 2021; Fu et al., 2021). The KASPar assays reported for FAE1.1 and FAE1.2 genes in the current communication were validated on large number of diverse genotypes and showed complete associations with erucic acid. High erucic acid B. juncea accessions are homozygous dominant for both the genes in contrast to homozygous recessive condition for low erucic acid genotypes as reported earlier (Gupta et al., 2004). Donskaja with intermediate erucic acid level is homozygous recessive for FAE1.1 and dominant for FAE1.2. This observation has explained the intermediate erucic acid levels reported earlier in East European B. juncea (Kirk and Hurlstone, 1983). The ability of our present marker system to differentiate between high vs. low erucic acid genotypes from the diverse geographic origins (Australia, India, China, East Europe, and Sweden) indicates its wider applicability. This could be largely due to involvement of Zem 1 and Zem 2 as donors for present low erucic acid cultivars of B. juncea (Potts et al., 1999; Bhat et al., 2002). These newly developed KASPar assays are codominant and could discriminate between homozygous and heterozygous states in the BC₁ progeny investigated. The segregation pattern of these assays is in confirmatory with digenic additive inheritance for erucic acid (Bhat et al., 2002; Gupta et al., 2004). The ability of our marker assays to differentiate homozygous/heterozygous state of alleles in inbred/pure lines and in segregating progenies will expedite the process of marker-assisted selection for low erucic acid in B. juncea.

The KASPar assays were developed on the basis of SNP variations in the coding regions of the FAE1 genes. These assays are highly genome specific and can discriminate between heterozygous, double recessive, or double dominant states of a genotype. These can be coupled with a background selection to expedite the recovery of the recurrent parent and reduce linkage drag associated with the trait donor parent. Per sample costs are also low and a large number of samples can be processed in 1 day. Developed KASPar assays are single step, high throughput, robust, codominant, and cost-effective than gel-based markers and can be efficiently used for marker-assisted selection of low erucic acid.

The data set included in this manuscript is appended as Supplementary Material.

GK and SB conceived and coordinated the study. GK, SK, KK, KG, and MS developed gene-specific primers and cloned the genes. KSG, GK, GDK, and JK conducted genotyping. GK and KSG looked after the field experiments and the phenotypic data collection. PC, MB, and GK designed the KASPar assays. KSG and GK wrote the manuscript and SB edited it. All authors read, commented, and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.