Evolution of the Beckwith-Wiedemann syndrome region in vertebrates

Abstract

In the animal kingdom, genomic imprinting appears to be restricted to mammals. It remains an open question how structural features for imprinting evolved in mammalian genomes. The clustering of genes around imprinting control centers (ICs) is regarded as a hallmark for the coordinated imprinted regulation. Hence imprinted clusters might be structurally distinct between mammals and nonimprinted vertebrates. To address this question we compared the organization of the Beckwith Wiedemann syndrome (BWS) gene cluster in mammals, chicken, Fugu (pufferfish), and zebrafish. Our analysis shows that gene synteny is apparently well conserved between mammals and birds, and is detectable but less pronounced in fish. Hence, clustering apparently evolved during vertebrate radiation and involved two major duplication events that took place before the separation of the fish and mammalian lineages. A cross-species analysis of imprinting center regions showed that some structural features can already be recognized in nonimprinted amniotes in one of the imprinting centers (IC2). In contrast, the imprinting center IC1 is absent in chicken. This suggests a progressive and stepwise evolution of imprinting control elements. In line with that, imprinting centers in mammals apparently exhibit a high degree of structural and sequence variation despite conserved epigenetic marking.

Figure 1.

Schematic map of orthologous genes in different species. Shown are maps of the human, chicken, zebrafish, and Fugu gene syntenies across the BWS region. The map is not to scale. Black bars indicate regions that are spanned by assembled genomic shotgun sequences and by BAC clones. For the remaining regions only shotgun assembled sequences were available. Interruptions of the horizontal lines indicate long distances between the genes. The chicken BWS region is present on two BAC contigs (contig 1: GenBank accession nos. BX640540, BX640401, contig 2: GenBank accession nos. BX649221, BX649222, BX640404, AP003796, AP003795, BX663531). The sequence contig in zebrafish is derived from five BAC sequences (GenBank accession nos. AL928843, AL929208, AL928880, BX001047, AL928628). The Fugu Igf2, Th, and Nap1l4 genes were also found in a cosmid sequence (GenBank accession no. AL021880).

Figure 2.

Organization of BWS genes and their paralogs in human and Fugu. Gene organization is shown as schematic maps of the human chromosomes and Fugu scaffolds (not to scale; for precise positions on the human chromosomes see Supplemental Table 3). Interruptions of the vertical gray lines representing human chromosomes indicate distances longer than 4 Mb between genes. Human chromosomes (Hs) are labeled by their numbers, as are Fugu sequence scaffolds (Fr). Initially selected genes on human chromosome 11p15.5 (see Fig. 1) are boxed, and their paralogs are shown in black. Additional genes are labeled in gray.

Figure 3.

Sequence conservation in Kcnq1 intron 10 in vertebrates. (A) Multiple alignments of Kcnq1 intron 10: the genomic human sequence was taken as reference sequence and compared to the genomic galago, cow, mouse, bat, armadillo, chicken, and zebrafish sequences. Before alignment, repetitive elements were masked using RepeatMasker software. Aligned regions are shown in green, highly conserved elements in red (>70% identity, >100 bp length). NICE1–NICE4 are highly conserved in all analyzed mammals. NICE1 and NICE4 are conserved in chicken (>60% identity, >100 bp length). The position of the IC2 CpG island in the human sequence is indicated by the CpG island plot above the multiple alignment. The CpG island plot shows CpG islands that fulfil the definition of a CpG island (length >200 bp, G+C content >50%, CpG_observed/CpG_expected >0.6, http://www.ebi.ac.uk/emboss/cpgplot/). The given scale bar is related to the human sequence. (B) The distributions of CpG islands in Kcnq1 in different vertebrate species. In pairwise alignments, the vertebrate sequences were used as reference sequences and the human sequence as second sequence. Scale bars are related to the reference sequence in each alignment. (C) Arrangements of repeated conserved sequence motifs in the putative IC2 in mammals. The consensus sequences of conserved motifs are listed. Segments that are conserved in overlapping motifs are underlined. The arrangements of these motifs in the different species are shown by different triangles. Motif MD was identified by Mancini-DiNardo et al. (2003). For the identified motifs the following numbers of mismatches to the consensus sequence were allowed: motif A, three mismatches; motifs MD1 and A2, two mismatches; motif A1, one mismatch; motif MD, six mismatches; CCAAT boxes, no mismatches. In some species the analyses were extended to regions flanking the CpG islands that are highlighted by gray bars. For the mouse and human sequences, the transcriptional start sites of Kcnq1ot1 (Du et al. 2004) are depicted by broken arrows indicating that the 3′extension of the transcript is not known. In mouse and human, location of restriction sites (No, NotI; As, AscI; Ea, EagI; Ec, EcoRI) that have been used for characterization of the IC2 in other studies (Du et al. 2003, 2004; Mancini-diNardo et al. 2003) are indicated.