Introduction

Although birds and mammals diverged over 300 million years ago, several chromosomal segments of similar gene content are conserved between human and chicken (Burt et al. 1999). Comparative mapping studies have been shown to be very useful to identify such homologous chromosome segments in human and chicken. Recently, detailed comparative maps between human and chicken chromosomes have been published (Nanda et al. 2000; Crooijmans et al. 2001; Buitenhuis et al. 2002; Jennen et al. 2002).

In our group, a bi-directional approach is used to improve the comparative map of chicken and human (Crooijmans et al. 2001; Buitenhuis et al. 2002; Jennen et al. 2002). First, a BAC contig is built, starting from loci whose positions on the chicken genome are known. Second, genes known to be located in the identified syntenic regions in human and mouse are used to map additional genes in these regions. The linkage map of chicken Chromosome (Chr) 15 (GGA15) is 71 cM in size and contains 19 markers (Groenen et al. 2000). Four chicken orthologues of human genes (CRYBB1, CRYBA4, IGL@, MIF), located in the human on Chr 22q11 (HSA22q11), have been mapped to this chicken microchromosome (Schmid et al. 2000).

The aim of this study is to improve the comparative map between GGA15 and HSA22q11 on the chicken genome.

Materials and methods

Chicken Chr 15 BAC clones

The Wageningen chicken BAC library was screened by PCR (Crooijmans et al. 2000) for all microsatellite markers and STS markers within genes located on GGA15. A detailed description of all loci used can be found at the ARKdb farm animal database (http://www.thearkdb.org/ ). This includes microsatellite marker MCW0052 located within the gene IGVPS. Primers corresponding to all other genes mapped to GGA15 (CRYBB1, CRYBA4, MIF) were designed based on database sequences (Table 1). All identified BAC clones were tested for purity by PCR amplification of the marker directly on two single colonies (colony PCR).

Table 1 Characteristics of markers developed in chicken genes
Full size table

Sequencing

BAC-end sequencing and sample sequencing were performed as described by Jennen et al. (2002). Sequences obtained were first analyzed with PREGAP4 of the STADEN software package (Bonfield et al. 1995; http://www.mrc-lmb.cam. ac.uk/pubseq ). The network BLAST client software (blastc13) of the NCBI was used to compare the sequences of good quality reads with sequences deposited in public databases. The BAC-end sequences were also used to develop new STS markers for chromosome walking. Sample sequences and BAC-end sequences, including STS markers have been submitted to GenBank and have been assigned the accession numbers BZ592394–BZ592544.

Mapping of genes

Genes of interest were either mapped to BACs that were already present within known BAC contigs or mapped by SNP typing as described by Buitenhuis et al. (2002). The SNP was first detected in the parents of the Wageningen mapping population (Groenen et al. 1998). A specific restriction enzyme for the SNP was used to map the gene as a PCR-RFLP on one selected family from the Wageningen mapping population.

Analysis of chromosomal rearrangements.

Chromosomal rearrangements were analyzed by using GRIMM (Tesler 2002; http://www-cse.ucsd.edu/groups/bioinformatics/GRIMM ). GRIMM enables the analysis of rearrangements in multichromosomal genomes and provides a new algorithm for analyzing comparative maps for which gene directions are unknown. Gene data sets based on the comparative map between human, mouse, and chicken were used for the calculation of the minimum possible number of rearrangements steps (the multichromosomal distance) between chicken and human, chicken and mouse, and human and mouse. The data sets were used with an unsigned gene order, because the gene orientation in chicken is unknown.

Results

Construction of GGA15 BAC contigs

BAC contigs of GGA15 were constructed around loci known to be located on this chromosome. The Wageningen chicken BAC library was initially screened with 17 markers. One BAC clone per marker was selected for end sequencing. The BAC-end sequences were used to design specific STS markers for chromosome walking. In total, 104 STS markers were designed and 240 BAC clones isolated, which resulted in the construction of nine BAC contigs.

Identification and mapping of genes

Since GGA15 showed conservation of synteny with HSA22q11 (Schmid et al. 2000), chicken orthologues of human genes from HSA22q11-q12 were identified to further increase the number of starting points for chromosome walking. Chicken orthologues from 14 human genes were identified by using a BLAST search with the mRNA sequences of human genes known to be located on HSA22q11-q12. For these 14 genes, STS markers were developed to screen the BAC library (Table 1). Nine genes—ADRBK2, CRYBB2, CRYBB3, GSTT1, PITPNB, RANBP1, SMARCB1, TFIP11, and XBP1—were mapped to BACs that were already present within the BAC contigs of GGA15.

The other five genes—CDC45L, HIRA, PNUTL1, PPIL2, and UFD1L—were mapped to BACs that formed a single contig, which had not yet been assigned to a chromosome. With PCR-RFLP, this BAC contig could also be mapped genetically to GGA15. Restriction enzyme HhaI was used to map the BAC clone bW041F24 positive for locus HIRA on the chicken linkage map. HIRA and, therefore, the complete BAC contig, were mapped close to microsatellite marker MCW0031 (recombination fraction = 0; LOD score = 12.64).

The chicken orthologue of TBX3, which in human is located on HSA12q24, was initially used within another project and, by using PCR-RFLP, was found to be located on GGA15. With restriction enzyme Tsp 509 I, BAC clone bW110C15 positive for locus TBX3 was mapped on the chicken linkage map close to ACW0169 (recombination fraction = 0.04; LOD score = 10.54).

To further increase the number of genes mapped to GGA15, 19 different BAC clones from GGA15 contigs were used for sample sequencing. The sequences obtained by sample sequencing and BAC-end sequencing were compared with sequences in Genbank and with the UMIST Chicken EST sequences (Boardman et al. 2002; http://www.chick.umist. ac.uk ) by using the BLAST algorithm. In total, sequence identity was found to 66 genes, ESTs, and genomic clones from chicken, human, and other vertebrates. The BLAST hits showed homology to 10 sequences from HSA22q11-q12 and to 16 sequences from HSA12q24. Homology to two genes and two anonymous genome segments from HSA3 was also found (Fig. 1a; Table 2), clearly marking a conserved segment. Furthermore, a sample sequence of BAC clone bW086M10 showed homology to a genomic clone from HSA1 (Fig. 1a; Table 2), but did not show homology to any annotated gene. On average, sequence homology with chicken sequences was 96.9%, and with human sequences, 81.5%.

Table 2 BAC sample sequencing hits and accession numbers
Full size table
Figure 1
figure 1

Comparative map of (a) GGA15 and part of HSA12 and 22, and (b) GGA15 and part of MMU5, 10, and 16. Estimated positions for mouse and human are given in cM and/or Mb, according to the Map Viewer from Entrez Genomes (http://www.ncbi.nlm.nih.gov/ ). The loci located on the linkage map of the chicken genome (Groenen et al. 2000) are shown on the left of the vertical bar. The numbers inside the vertical bar of GGA15 represent the relative positions in cM of the chicken loci. On the left of the vertical bar from GGA15 the 9 BAC contigs are shown in solid bars (not to scale). The genes (human name) and genomic clones (human accession number) located on GGA15 are shown on the right site. Genes mapped in this study by sequence comparison using the BLAST algorithm are in bold; genes mapped by chromosome walking are underlined; and genes used in PCR-RFLP studies are in italics.

Full size image

Analysis of chromosomal rearrangements.

GRIMM was used for the calculation of the multichromosomal distance between chicken and human, chicken and mouse, and human and mouse.

Based on the comparative map between human, mouse, and chicken as shown in Fig. 2, the following data sets were generated (format as needed for GRIMM):

  1. 1

    >chicken

  2. 2

    1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 2728 29 30 $

  3. 3

    >human

  4. 4

    8 9 10 11 12 30 16 29 $

  5. 5

    18 4 3 2 1 7 6 5 17 15 14 13 28 27 26 25 24 23 21 $

  6. 6

    19 20 $

  7. 7

    22 $

  8. 8

    >mouse

  9. 9

    5 18 6 7 1 2 3 4 17 $

  10. 10

    13 14 15 $

  11. 11

    19 20 $

  12. 12

    8 9 21 23 24 25 26 27 28 29 30 16 12 11 10 $

  13. 13

    22 $

Figure 2
figure 2

Comparative map of chicken chromosome 15 (GGA15) to human (HSA) and mouse (MMU). A number is assigned to each gene and genomic clone (1–30), which were mapped in all three species. Chromosome segments in which the gene order in all three species is the same, are indicated by block 1–15. Positions of chromosomal rearrangements are indicated by dotted lines, with the chicken gene order as a start. The numbers of the human and mouse chromosomes are shown inside the vertical bars of HSA and MMU respectively.

Full size image

These data sets were used with an unsigned gene order, because the gene orientation in chicken is unknown. The gene order for human and mouse is given per chromosome. Calculations resulted in a multichromosomal distance between chicken and human of 11, between chicken and mouse of 12, and between human and mouse of 6.

Discussion

GGA15 BAC contigs

The linkage map of GGA15 is 71 cM, which is about 1.8% of the total chicken genome (˜4000 cM) (Groenen et al. 2000; Groenen and Crooijmans 2003). On the basis of the physical size of the chicken genome of 1.2 × 109 bp, 1 cM on average corresponds to 300 kb. Although there are some indications that this ratio is somewhat different for macrochromosomes versus microchromosomes, the estimated physical size for GGA15 would be around 21 Mb. From the average insert size of the BACs of 134 kb (Crooijmans et al., 2000) and correcting for the overlap between the different BACs, we calculated that the BACs would cover around 30–35% of GGA15. However, because no markers were identified between microsatellite marker MCW0323 and AFLP marker ACW0169 (distance 22 cM) to screen the Wageningen BAC library, no BAC clones could be found in this region, except for the BAC clone positive for TBX3. Therefore, only the first 49 cM of GGA15 (between ADL0206 and MCW0323) are covered, with BAC clones in nine contigs. For this region the coverage is estimated to be almost 50%. On average, 4.8 BAC clones were obtained per marker, which is comparable to the previously reported number (Crooijmans et al. 2001; Buitenhuis et al. 2002; Jennen et al. 2002).

The beta crystallin gene cluster

In human, CRYBB1 and CRYBA4 form a gene cluster with CRYBB2, CRYBB2P1, and CRYBB3, the beta-crystalline gene cluster. Orthologues of all human members of this gene cluster, except for the pseudogene CRYBB2P1, are present in mouse as a gene cluster on chromosome 5 (MMU5) (Hulsebos et al. 1995a, 1995b). From the comparative map, we expected the chicken orthologues of human CRYBB2 and CRYBB3 to form a gene cluster with chicken CRYBB1 and CRYBA4. The results above confirm our expectation of the four chicken beta crystallin genes to form a gene cluster. Figure 3 shows that all four beta crystallin genes are located on overlapping BAC clones in the same BAC contig; CRYBB1 and CRYBA4 are located on BAC clones bW043G23 and bW093I01, whereas CRYBB2 and CRYBB3 are located on bW031O19, bW058F23 and bW070F04. In chicken, CRYBB1 and CRYBA4 are linked head to head, with 2147 nt of intervening sequence (Duncan et al. 1995). Also in human both genes are tightly linked head to head, with 3890 bp in between (NCBI Map Viewer build30; http://www.ncbi.nlm.nih. gov/ ). It can be expected that chicken CRYBB2 and CRYBB3 are also tightly linked, because the distance between the human genes is 2235 bp.

Figure 3
figure 3

BAC contig of GGA15 constructed around the beta crystallin gene cluster. The BAC contig with its BAC clones and STS markers is shown on the left and the comparative map with human and mouseon the right. Coloured bars give the positions of the genes and genomic clones in human and mouse. Human and mouse map positions are given in Mb.

Full size image

Furthermore, in our study the distance between CRYBB1 and CRYBB2 is equal to one BAC clone (˜134 kb), which is about 10 times smaller than in human. This in not according to the whole-chicken genome size, which is on average three times smaller than the human. Nevertheless, it could be that genome sizes differ more on specific spots.

Chicken versus human and mouse

Although, GGA15 initially showed conservation of synteny only with HSA22, our results also showed conservation of synteny with HSA12. However, from an ancestral point of view, our findings are not at all surprising. Several reports on the evolution of the mammalian genome (Murphy et al. 2001; Haig 1999; O’Brien and Stanyon 1999; Chowdhary et al. 1998) describe two ancestral chromosomes, which are both a combination of human Chrs 12 and 22. These ancestral chromosomes were reconstructed by using chromosome paints and comparative maps of several primates, rodents, and other mammalian species. The first ancestral chromosome is a combination of HSA12p-q and HSA22qtel (12pq-22qtel) and shows conservation of synteny with segments of GGA1 (Murphy et al. 2001; Schmid et al. 2000). The second ancestral chromosome comprises HSA12qtel and HSA22q (22-12qtel) and is syntenic to GGA15 (this paper).

In order to reconstruct the common ancestor of mouse, human, and chicken, we compared the gene order in chicken with the human and mouse maps. This comparison clearly shows a large number of intra- and interchromosomal rearrangements. For example, the BAC clone bW122C13 within the contig of markers ABR0070, MCW0052, and MCW0231 contains the gene DGCR2 and gene cluster IGL@ (including IGVPS), both located in human on HSA22q, and the gene VPS29, which is located in human on HSA12q24 (Fig. 1a). The presence of DGCR2 and IGL@ from HSA22 as well as VPS29 from HSA12 on the same BAC clone suggests that a chromosomal breakpoint is located between IGL@ and VPS29. This chromosomal breakpoint can also be found in mouse, where IGL@ is located on MMU16 and VPS29 is located on MMU5 (Fig. 1b).

Another example of a chromosomal breakpoint can be seen in the BAC contig containing the beta crystallin gene cluster (Fig. 2). Conserved blocks of genes are shown in all three species within this region. In human, the chromosomal breakpoint can be found between the genes CRYBB3 (HSA22q11) and GIT2 (HSA12q24). This breakpoint is absent in mouse, which is in good agreement with the mammalian ancestor 22-12qtel.

Both examples indicate that during evolution the breakpoints occurred after the separation of the mammalian and bird lineages. Moreover, the second example clearly shows that the breakpoint occurred in the human lineage after the human and mouse lineages separated, suggesting that the origin of the chromosome breaks was in human.

A more detailed comparison of the conserved chromosome segments between GGA15 and the human and mouse chromosomes is shown in Fig. 2. The order of the conserved segments is based on a combination of genetic mapping, chromosome walking results, and sequencing. We can identify at least 15 blocks, which is the minimum number of conserved segments between GGA15 and the human and mouse chromosomes. The number of genes/genomic clones per block varies from 1 (blocks 2, 7, 8, 9, 12, 14, and 15) to 7 (block 13). The dotted lines indicate the points where chromosomal rearrangements took place. For analyzing these rearrangements, we used GRIMM. Between human and mouse the lowest multichromosomal distance (6) is calculated. The distances between chicken and human (11) and between chicken and mouse (12) are comparable. These results are in agreement with the fact that human and mouse evolved from a common ancestor and, therefore, are more closely related than either one is to chicken. This is in contrast to the findings of Burt et al. (1999), who looked at the whole chicken genome and concluded that the genomes of chicken and human are more alike than those of mouse and human. However, the conclusion of Burt et al. (1999) was based on a limited number of mapped genes and conserved segments.

Using the same approach as Burt et al. (1999) and assuming that the GGA15 data are representative for the whole chicken genome, we predict the total number of conserved segments in the chicken– human–mouse comparison to be at least 800 and the total number of chromosomal rearrangements to be in the same size order. Crooijmans et al. (2001) estimated, in the comparative mapping study between GGA10 and HSA15, the total number of conserved segments to be at least 600. From these numbers, we estimate the rate of chromosomal change in the chicken lineage to range from 1.6 to 2 rearrangements per Myr since the divergence 300 Myr ago, which is slightly higher than the estimate of 1.5 in the human and mouse lineages (Pevzner and Tesler 2003).

The comparison between the results of Burt et al. (1999) and our results indicates that a higher gene density as well as the exact gene order and orientation are needed to better understand the evolutionary events that took place in the lineage leading to human, mouse, and chicken.