Brassica genomics: a complement to, and early beneficiary of, the Arabidopsis sequence.

Those studying the genus Brassica will be among the early beneficiaries of the now-completed Arabidopsis sequence. The remarkable morphological diversity of Brassica species and their relatives offers valuable opportunities to advance our knowledge of plant growth and development, and our understanding of rapid phenotypic evolution.

Beyond Arabidopsis

Less than 20 years after DNA-level studies [1] highlighted the small DNA content and simple organization of the genome of Arabidopsis thaliana - an ephemeral weed-like member of the Cruciferae family - an essentially complete sequence of its genome has been announced [2]. This, the first complete genome sequence for a flowering plant, provides the means to investigate in unprecedented detail the common features that plants share with one another and with other life forms, as well as the unique features that adapt particular taxa to diverse life histories (such as the largely sessile nature of plants). Many botanists are especially interested in studying how the interaction of natural and human selection has enabled a small subset of plants to be 'domesticated' to provide food, feed, and fiber. Such investigations should be greatly facilitated by the availability of plant genome sequences.

The economic and morphological importance of Brassica

Crops of the genus Brassica (tribe Brassiceae), which are in the same taxonomic family as Arabidopsis thaliana, are widely used in the cuisine of many cultures, due, in part, to the many choices of edible forms in the genus. Economically, Brassica is loosely categorized into oilseed, vegetable, and condiment crops. B. napus, B. rapa (formerly campestris), B. juncea, and B. carinata provide about 12% of the worldwide edible vegetable oil supplies [3]. B. oleracea and B. rapa, the so-called 'cole crops', comprise many of the vegetables in our daily diet. Several of these vegetables have extreme morphological characteristics. Examples of such morphologies include the enlarged inflorescence of cauliflower (B. oleracea subspecies botrytis) and broccoli (B. oleracea subspecies italica); the enlarged stem of kohlrabi (B. oleracea subsp. gongylodes) and marrowstem kale (B. oleracea subspecies medullosa); the enlarged root of turnip (B. rapa subspecies rapifera); the enlarged and twisted leaves of Pak-choi (B. rapa subspecies chinesis) and Chinese cabbage (B. rapa subspecies pekinesis); and the enlarged single apical bud of cabbage (B. oleracea subspecies capitata) or the many axillary buds of Brussels sprout (B. oleracea subspecies gemmifera) [4]. Finally, the seed of B. nigra is utilized as a condiment - mustard. Brassica species are a valuable source of dietary fiber, vitamin C, and other possible salubrious factors such as anticancer compounds [5]. Estimates of the economic importance of Brassica species are conservative, because several cole crops such as collards are cultivated primarily for local or home use, but are nonetheless a dietary mainstay in low-income communities where other fresh vegetables can be prohibitively expensive.

The current state of Brassica genomics

The genomes of diploid Brassica species are 3-5 times the size of the Arabidopsis genome, ranging from 0.97 pg/2C (468 Mb/1C; where C is haploid DNA per nucleus) for B. nigra to 1.37 pg/2C (662 Mb/1C) for B. oleracea. The amphidiploid genomes, which have two sets of chromosomes from each parent species, range from 2.29 pg/2C (1,105 Mb/1C) to 2.56 pg/2C (1,235 Mb/1C) [6,7]. Relationships among the three diploid species - B. rapa (syn. rapa; 2n = 20, genome AA), B. nigra (2n = 16, genome BB) and B. oleracea (2n = 18, genome CC) - and three amphidiploids - B. napus (2n = 38, genome AACC), B. juncea (2n = 36, genome AABB) and B. carinata (2n = 34, genome BBCC) - are well known [8]. Rapid-cycling strains of B. oleracea have been developed that have a life cycle as short as that of Arabidopsis thaliana, making them easier to study [9].

Neutral DNA polymorphisms are common among con-specific genotypes of most Brassica species, and at least 15 molecular maps have been produced, using at least 900 different publicly available Brassica and Arabidopsis DNA probes (which are each used with similar efficacy). Virtually all the probes hybridize to two or more loci, even in diploid Brassica species; however, only a subset of loci segregates for allelic variation in any one mapping population. The subset of genetically mapped loci can, therefore, be used to infer the locations of many additional sequence tagged sites (STSs) [10].

The comparative organization of the chromosomes of Brassica and Arabidopsis is well studied. On the basis of 186 corresponding loci, about 19 chromosome structural rearrangements differentiate B. oleracea and A. thaliana orthologs, suggesting that chromosomal tracts of about 20-25 cM remain largely colinear in the two taxa [10]. Microsynteny studies involving comparative genetic and physical mapping of specific chromosome segments have shown largely conserved gene order in Arabidopsis and Brassica, but some disruption in gene content by deletions or insertions [11,12]. Furthermore, duplicated partial gene clusters are commonly found in both Arabidopsis and Brassica species [13,14,15,16]. Comparative sequencing has permitted orthology assignment of some duplicated segments in Arabidopsis and Brassica species [17].

Expectations of comparative Arabidopsis-Brassica genomics

Comparative data help to highlight how tools from Arabidopsis might be used to identify genes that may directly account for Brassica quantitative trait loci (QTLs). An example of this from one of our labs involves the flowering-time genes. In B. rapa, a major QTL for flowering time was found in a region homologous to the top of chromosome 5 in Arabidopsis where several flowering-time genes are located [18]. After backcrossing, the B. rapa locus segregated as a discrete character, and, by comparative fine mapping, it was shown to correspond to the Arabidopsis flowering-time gene flc [19]. In B. oleracea, 15 QTLs controlling flowering time were aligned to Arabidopsis genomic regions containing flowering-time genes [20]. Ten of these QTLs occurred in regions corresponding to several segments of the Arabidopsis genome containing genes that affect flowering: the top of chromosome 5, which contains 7 flowering genes (tfl1, flc, tfl2, co, fy, art1 and emf1); three regions of chromosome 1 containing efs, fha, and gi; and a region of chromosome 3 containing hy2 and vrn1. Similarly abundant sets of Arabidopsis candidates have been found for Brassica QTLs that affect leaf and overall plant architecture [21]. The Arabidopsis genome sequence provides a valuable resource for identifying and further evaluating sets of candidate genes that may account for the genetic control of complex traits in Brassica.

New research opportunities from Brassica genomics

Comparative genomics has advanced the discovery and understanding of conserved genes, but most sequencing projects have also revealed many rapidly evolving 'orphan' genes of unknown function and evolutionary history. Such rapidly evolving genes in Drosophila, mammals, and several other species are important for reproductive success, cell-cell recognition, and cellular response to pathogens [22,23]. Because expressed sequence tag (EST) databases for Brassica remain small (as of 27 January 2001, a search of the National Center for Biotechnology Information (NCBI) databases [24] for 'Brassica' reveals only 5,868 matches), the abundance and possible importance of rapidly evolving genes in accounting for key differences between Brassica and Arabidopsis, or among different Brassica species and genotypes, remain unknown.

Naturally occurring Brassica genetic variants may prove useful for new investigations into plant development, particularly through the study of the effects of alleles that lack variation in Arabidopsis. Many Brassica phenotypes are under much more complex genetic control than similar phenotypes in Arabidopsis. For example, the discovery that the joint effects of two mutations in Arabidopsis, CAULIFLOWER and APETALA1, could produce a plant with a curd-like inflorescence [25] is in contrast to the much more complex genetic control of curd architecture reported for Brassica [20]. The high level of duplication in the Brassica genome [10] may permit mutations in loci that are under tight selective constraints in Arabidopsis. Although early hints that there may be large-scale chromatin duplications in Arabidopsis [26,27] have been borne out by the complete sequence [2,28,29,30], gene loss or divergence has been substantial; most Brassica genes, however, are represented in two to three or even more copies. Investigations into the expression and function of redundant genes in Brassica species should shed light on the role of genome replication in the phenotypic divergence of the genus.

Brassica species also provide the opportunity to study rapid genome changes associated with polyploidy. Each of the amphidiploid species - B. napus, B. juncea and B. carinata - can be resynthesized by hybridizing the diploid species and then doubling the chromosomes [31]. This results in completely homozygous polyploid lines that should not segregate, but variation in DNA restriction fragments [32] and phenotype [33] has been observed among progeny derived from self-pollination of single resynthesized polyploids. As Brassica and Arabidopsis genes share, on average, 87% sequence identity [16], it may be possible to survey large numbers of Brassica genes for changes in expression using DNA microarrays based on Arabidopsis gene sequences. Comparisons of newly derived polyploids with their exact diploid progenitors should provide insight into the consequences of polyploidy on plant evolution.

Brassica is distinguished from many other plant taxa by its remarkable propensity to evolve new morphological variants rapidly, as is evident from the inter-fertility among common vegetables such as cabbage, cauliflower, broccoli, Brussels sprouts, and kohlrabi. In much the same manner that the dog genome project promises to contribute to the identification of genes conferring morphological differences among mammals [34], Brassica genomics might provide new insights into the size and shape of plants. To this end, the complete sequence of Brassica's close relative, Arabidopsis thaliana, should prove to be a powerful tool.