Structural and functional analysis of a 0.5-Mb chicken region orthologous to the imprinted mammalian Ascl2/Mash2-Igf2-H19 region

Abstract

Previous studies revealed that Igf2 and Mpr/Igf2r are imprinted in eutherian mammals and marsupials but not in monotremes or birds. Igf2 lies in a large imprinted cluster in eutherians, and its imprinting is regulated by long-range mechanisms. As a step to understand how the imprinted cluster evolved, we have determined a 490-kb chicken sequence containing the orthologs of mammalian Ascl2/Mash2, Ins2 and Igf2. We found that most of the genes in this region are conserved between chickens and mammals, maintaining the same transcriptional polarities and exon-intron structures. However, H19, an imprinted noncoding transcript, was absent from the chicken sequence. Chicken ASCL2/CASH4 and INS, the orthologs of the imprinted mammalian genes, showed biallelic expression, further supporting the notion that imprinting evolved after the divergence of mammals and birds. The H19 imprinting center and many of the local regulatory elements identified in mammals were not found in chickens. Also, a large segment of tandem repeats and retroelements identified between the two imprinted subdomains in mice was not found in chickens. Our findings show that the imprinted genes were clustered before the emergence of imprinting and that the elements associated with imprinting probably evolved after the divergence of mammals and birds.

Figure 1.

Overview of the ASCL2/CASH4-IGF2 region on chicken Chromosome 5 and the orthologous region on mouse Chromosome 7. The locations of the chicken BAC clones that were sequenced in this study are shown as horizontal bars. The DDBJ/GenBank/EMBL accession numbers of the sequences are indicated in parentheses.

Figure 2.

Overall structure and gene arrangement of the chicken region. Arrangement of genes (boxes) is illustrated at the top of each panel. Horizontal arrows or arrowheads above or below the genes indicate directions of transcription. Upward arrowheads indicate positions of CpG islands and tandem repeats. Observed/Expected ratio of CpG dinucleotide per 100 bp and G+C content per 100 bp are graphically shown under the gene arrangement.

Figure 3.

Sequence comparison between chickens and mice from CD81/TAPA1 to IGF2 (A) and from MRPL23/L23MRP to LSP1 (B). Percent identity plot analysis was done using the PipMaker software using the chicken sequence as a reference. The order and arrangement of the chicken genes are shown at the top. The dots and short horizontal bars correspond to the segments of sequence conservation. The regions corresponding to the exons and introns of the chicken genes are shown in green and yellow, respectively. Conserved sequences in the noncoding regions, including those in introns, are circled.

Figure 3.

Sequence comparison between chickens and mice from CD81/TAPA1 to IGF2 (A) and from MRPL23/L23MRP to LSP1 (B). Percent identity plot analysis was done using the PipMaker software using the chicken sequence as a reference. The order and arrangement of the chicken genes are shown at the top. The dots and short horizontal bars correspond to the segments of sequence conservation. The regions corresponding to the exons and introns of the chicken genes are shown in green and yellow, respectively. Conserved sequences in the noncoding regions, including those in introns, are circled.

Figure 4.

Structure and expression of chicken PHEMX/TSSC6. (A) RT-PCR analysis of chicken PHEMX/TSSC6 in whole chick embryos and adult tissues. GAPD was used as a housekeeping control. Splicing variants, revealed by sequencing of the PCR products, are shown (isoforms 1–6). (B) Nucleotide sequence and predicted amino acid sequence of chicken PHEMX/TSSC6 cDNA (isoform 1; GenBank accession no. AB101639). Arrowheads indicate the positions of exon–intron boundaries. (C) Alignment of the predicted amino acid sequences of human (GenBank accession no. AF125569), mouse (GenBank accession no. AJ279791), and chicken PHEMX/TSSC6. Amino acids conserved in more than two species are shaded.

Figure 4.

Structure and expression of chicken PHEMX/TSSC6. (A) RT-PCR analysis of chicken PHEMX/TSSC6 in whole chick embryos and adult tissues. GAPD was used as a housekeeping control. Splicing variants, revealed by sequencing of the PCR products, are shown (isoforms 1–6). (B) Nucleotide sequence and predicted amino acid sequence of chicken PHEMX/TSSC6 cDNA (isoform 1; GenBank accession no. AB101639). Arrowheads indicate the positions of exon–intron boundaries. (C) Alignment of the predicted amino acid sequences of human (GenBank accession no. AF125569), mouse (GenBank accession no. AJ279791), and chicken PHEMX/TSSC6. Amino acids conserved in more than two species are shaded.

Figure 5.

Structure of chicken MRPL23/L23MRP. (A) Nucleotide sequence and predicted amino acid sequence of chicken MRPL23/L23MRP cDNA (GenBank accession no. AB101640). Arrowheads indicate the positions of exon–intron boundaries. (B) Alignment of the predicted amino acid sequences of human (GenBank accession no. BC027710), mouse (GenBank accession no. U84902), and chicken MRPL23/L23MRP.

Figure 6.

Structure of chicken LSP1. (A) Nucleotide sequence and predicted amino acid sequence of chicken LSP1 cDNA (GenBank accession no. AB101641). Arrowheads indicate the positions of exon–intron boundaries. (B) Alignment of the predicted amino acid sequences of human (GenBank accession no. NM002339), mouse (GenBank accession no. NM019391), and chicken LSP1.

Figure 6.

Structure of chicken LSP1. (A) Nucleotide sequence and predicted amino acid sequence of chicken LSP1 cDNA (GenBank accession no. AB101641). Arrowheads indicate the positions of exon–intron boundaries. (B) Alignment of the predicted amino acid sequences of human (GenBank accession no. NM002339), mouse (GenBank accession no. NM019391), and chicken LSP1.

Figure 7.

Imprinting status of chicken IGF2 (A), ASCL2/CASH4 (B), INS (C), and the control genes PHEMX/TSSC6 (D) and TH (E). Biallelic expression was revealed in informative embryos and embryonic membranes including the yolk sac by RT-PCR and sequencing. The SNPs used to distinguish the parental alleles were 1242A/G for IGF2 (Yokomine et al. 2001), 29365A/G for ASCL2/CASH4, 106015G/A for INS, 11747A/G for PHEMX/TSSC6, and 86744G/T for TH. In B and D, the data for the other strand are shown. Three additional SNPs confirmed the biallelic expression of ASCL2/CASH4 (data not shown).

Figure 8.

Bisulfite methylation analysis of the chicken IGF2 region homologous to the 54-bp core of the mouse DMR2. CpG dinucleotides are underlined. The mouse CpG sites are numbered according to Murrell et al. (2001). Filled circles indicate methylated sites and open circles unmethylated sites. There is no indication for allelic difference in DNA methylation.