br Acknowledgements br Introduction In Arabidopsis thaliana

Acknowledgements
Introduction In Arabidopsis thaliana (L.) Heynh, there are at least four copies of the caseine kinase-like gene Ck1, which is involved in peptide phosphorylation in mammals [1]. These genes are distributed on GS-9620 australia 1, 2, 4, and 5. A similar number of copies for Ck1 have been inferred by Southern blot analysis in Brassica[2]. Little is known about their origin and functionality, and even less on the Ck1 homologs in other Brassicaceae, such as Brassica oleracea (L.). Although A. thaliana and B. oleracea belong to the same family, they are classified in different tribes [3] and their level of divergence has been estimated to be approximately 28 million years [4]. The main contrasting difference between the genomes of Brassica and Arabidopsis is size, measured by DNA content and chromosome number. The B. oleracea genome is approximately four times larger and has almost twice the chromosome number than the Arabidopsis genome [5]. In spite of its reduced size, the Arabidopsis genome is extensively duplicated [5], [6], [7]. Therefore, the existence of numerous copies for Ck1 is not surprising. A. thaliana is a model species for comparative mapping, but gene duplications make difficult to do orthology assignment of specific members of a gene complex into other species. B. oleracea has also a complex genome largely duplicated and it is believed to be itself a paleopolyploid resulting from ancient hybridization events [8]. So far, two copies of Ck1 have been described in B. oleracea, BoCk1a, and BoCk1b. These two genes are orthologous to Arabidopsis, AtCk1/4, and AtCk1/1, respectively [9]. In this paper, we compare the sequences of the duplicate members of four Ck1 homologs in A. thaliana and five homologs in B. oleracea. This comparison allowed assignment of orthology and paralogy of these duplicated genes within and between the two species. On the basis of their relationship and existing hypothesis on the origin of the Brassica genome, we present a possible scenario leading to the evolution of these sequences.
Materials and methods
Results A total of five copies of Ck1 were detected in B. oleracea from screening the BAC library, BoCk1a, BoCk1b, BoCk1c, BoCk1d, and BoCk1e. Four of these were complete copies, whereas the homolog BoCk1d was truncated at the 3′ end. We compared the 5′ ends for all five Brassica and four Arabidopsis genes. The complete coding regions and 3′ ends of the four Brassica and four Arabidopsis homologs were also compared. Truncated BoCk1d was also aligned to the rest of the homologs at the first nine introns, since the truncation occurred in exon 10.
Discussion The difference in exon numbers separates the Ck1 homologs of both Arabidopsis and Brassica into two major groups. The event leading to the absence of intron 11 resulting from the fusion of exons 11 and 12 took place before the separation of the lineages leading to the tribes Brassiceae and Arabidae. This event, plus phylogenetic analysis of sequence data (Fig. 4), served to assign orthology of AtCk1/5 to BoCk1c, BoCk1d, and BoCk1e. Our results indicate that BoCk1c, BoCk1d, and BoCk1e are paralogs that may have arisen via gene duplications following the separation of Brassica from Arabidopsis. We do not know whether these genes are on the same or different chromosomes, but probably they are not duplicated in tandem, considering that they were harbored by different non-contiguous BAC clones. BoCk1c differs from the other homologs in this cluster, and may represent a pseudogene, for the following reasons: it has a much larger intron 7 than the rest of the homologs; exons 11 and 14 are smaller in BoCk1c than in AtCk1/5 and BoCk1e (these exons were not available in BoCk1d for analysis). Its promoter region has lower sequence similarity than that observed between BoCk1d and BoCk1e. Phylogenetic analyses of exon nucleotide and inferred amino acid sequences supported the orthology of AtCk1/4 and BoCk1a, and of AtCk1/1 and BoCk1b, as previously reported by Quiros et al. [9]. The latter group also shared an inflated size of intron 2. The chromosome 1 segment, harboring AtCk1/1, is partially duplicated on chromosome 3, whereas the chromosome 4 segment carrying AtCk1/4 is partially duplicated on chromosomes 2 and 5 [9].