Gene duplication and fragmentation in the zebra finch major histocompatibility complex

Abstract

Background: Due to its high polymorphism and importance for disease resistance, the major histocompatibility complex (MHC) has been an important focus of many vertebrate genome projects. Avian MHC organization is of particular interest because the chicken Gallus gallus, the avian species with the best characterized MHC, possesses a highly streamlined minimal essential MHC, which is linked to resistance against specific pathogens. It remains unclear the extent to which this organization describes the situation in other birds and whether it represents a derived or ancestral condition. The sequencing of the zebra finch Taeniopygia guttata genome, in combination with targeted bacterial artificial chromosome (BAC) sequencing, has allowed us to characterize an MHC from a highly divergent and diverse avian lineage, the passerines.

Results: The zebra finch MHC exhibits a complex structure and history involving gene duplication and fragmentation. The zebra finch MHC includes multiple Class I and Class II genes, some of which appear to be pseudogenes, and spans a much more extensive genomic region than the chicken MHC, as evidenced by the presence of MHC genes on each of seven BACs spanning 739 kb. Cytogenetic (FISH) evidence and the genome assembly itself place core MHC genes on as many as four chromosomes with TAP and Class I genes mapping to different chromosomes. MHC Class II regions are further characterized by high endogenous retroviral content. Lastly, we find strong evidence of selection acting on sites within passerine MHC Class I and Class II genes.

Conclusion: The zebra finch MHC differs markedly from that of the chicken, the only other bird species with a complete genome sequence. The apparent lack of synteny between TAP and the expressed MHC Class I locus is in fact reminiscent of a pattern seen in some mammalian lineages and may represent convergent evolution. Our analyses of the zebra finch MHC suggest a complex history involving chromosomal fission, gene duplication and translocation in the history of the MHC in birds, and highlight striking differences in MHC structure and organization among avian lineages.

Figure 1

Schematic diagram highlighting results of BAC clone assembly and annotation, FISH mapping, and evolutionary comparisons. For zebra finch, genes within boxes are linked in a single BAC contig. Contigs within dashed ovals are linked by known location within a single BAC but the order is uncertain. BACs that map to the same chromosome via FISH mapping in are within a solid oval (see also Figure 4 for FISH mapping results). For chicken boxes represent MHC-B and MHC-Y regions. For Xenopus boxes represent sequenced BACs whose chromosomal organization is unknown. For clarity, not all genes of the MHC are shown.

Figure 2

Genomic map of the Chicken MHC - B complex after Shiina et al. [20] compared to two sequence zebra finch Class I clones. While KIFC and MHC Class I were identified in a single BAC, no orthologs of the intervening chicken genes were found in zebra finch. An MHC Class I gene was not found in the TAP containing zebra finch clone despite the proximity of these genes in the chicken MHC. Following the chicken naming scheme, class I MHC genes in chicken are denoted BF1 and BF1, and class IIB genes are denoted BLB1 and BLB2. Genes targeted in the BAC screening are marked with arrows.

Figure 3

Sequence conservation and alignment diagram using Zpicture. Zebra finch BAC 157B03 and previously sequenced cosmid clone (rwcos3) from red-winged blackbird [48] were compared highlighting regions of sequence conservation. The Y axis in each panel represents the percent similarity. Exons (blue boxes), UTRs (yellow boxes) and intergenic regions are based on FGENESH predictions, and repeats (green boxes) are predicted by Zpicture [47] (using Repeatmasker). Regions of sequence similarity (brown boxes) not only include the Class IIB gene, but also the zinc finger-like sequences identified. Gene names are based on best BLAST hits. The ordering of genes is based on the zebra finch BAC assembly and is not necessarily the same in the red-winged blackbird.

Figure 4

FISH mapping of BAC clones. A) Single color FISH mapping of TGAC-157B03 reveals extensive cross-hybridization across chromosomes. Similar results were observed for other Class II clones presumably as a result of their high repeat content. B) Lack of cohybridization between Clones TGAC-102M22 and a known chromosome 22 BAC indicates that TGAC-102M22 is not on chromosome 22 as indicated by the genome assembly. C) Dual color FISH of TGAC-86I22 (red) and TGAC-167E04 (green) indicating cohybridization of these clones, a result also supported by sequence analysis. These clones were assembled together, and contain g-filamin, TNXB, TAP1 and TAP2 genes. D) Clones TGAC-102M22 (red) (contains MHC Class I, FLOT, TUBB, KIFC and DAXX) and TGAC-86I22 map to different chromosomes. Key components of the classical MHC therefore map to different chromosomes in the zebra finch genome.

Figure 5

RFLP/Southern Blot of 10 captive zebra finches. Individuals 1 to 7 are from a captive American population and individuals 8 to 10 are from a Swedish population. The left panel shows the banding patterns using a Class I probe and the right panel shows the results using a Class II probe. Results from Class I analysis suggest a minimum of two loci whereas Class II probes indicate a very large lumber of loci (mean number of bands = 19 +/- 4.6, range: 12 to 27).

Figure 6

Comparison of gene density across three avian lineages and the human HLA region. Estimates from zebra finch are based on two BAC assemblies (TGAC-102M22 and TGAC-167E04/TGAC-86I22) containing 11 predicted genes.

Figure 7

Long terminal repeat (LTR) content in avian MHC regions. Chicken (AB268588), Quail (AB078884.1), and Blackbird (AF328738) sequences from Genbank are compared with sequenced zebra finch BACs.

Figure 8

Phylogenetic analysis and selection on MHC Class II sequences. A) Phylogenetic relationships among passerine MHC Class II exon 2 and 3 sequences. Four sequences with open reading frames were found in the zebra finch genome. The remaining sequences are from GenBank. The root of the tree was placed at a divergent zebra finch lineage (TAGU 2) based on a larger analysis in which non-passerine sequences were included. TAGU 1 to 4 correspond to loci 1 to 4 in Table 1. B) Predicted amino acid sequences of the second exon of four apparently functional zebra finch MHC Class IIB genes. Stars represent sites showing evidence of selection in passerine birds. Note the correspondence between sites showing evidence of selection in passerines and the predicted peptide binding region in humans.