Phylogenetic resolution power of microsatellites and various single-nucleotide polymorphism types assessed in 10 divergent chicken populations.

There has been some debate over the question of which types of DNA variation are most appropriate to accurately reconstruct evolutionary events. We compared the capacity of microsatellites (STRs) and various types of single-nucleotide polymorphism (SNP) loci in the chicken genome. The SNP types differ in their location: in exons, introns and promoters. Genetic distances between all possible pairs of 10 populations were calculated for each marker type. STR loci, which are much more polymorphic than are SNPs, are considered to have occurred at recent time compared with old evolutionary events of SNPs. Using structure software, STR loci assigned individuals to their population much more correctly than did any other marker types, whereas SNPs at promoter regions gave the poorest ascription. Furthermore, 29 STR markers were even better than all 152 SNPs together. Ancient evolutionary events that produced genetic differences between the most distant populations such as Red Jungle Fowl and domestic chicken were detected better by exons and introns than by STR loci and promoters. The significant interactions found between marker types and populations suggest that marker types had different phylogenetic histories, possibly related to a different timescale.

Comparative genome analysis with the human genome reveals chicken genes associated with fatness and body weight.

The selection of meat-type chickens (broilers) for rapid growth has been accompanied by excessive fat deposition. In this study, we analysed 53 candidate genes that are associated with obesity and obesity-related traits in humans, for which we found chicken orthologues by BLAST searches. We have identified single nucleotide polymorphisms (SNPs) with significant differences in allele frequencies between broilers and layers in each of the following six candidate genes: adrenergic, beta-2-, receptor, surface (ADRB2); melanocortin 5 receptor (MC5R); leptin receptor (LEPR), McKusick-Kaufman syndrome (MKKS), milk fat globule-EGF factor 8 protein (MFGE8) and adenylate kinase 1 (AK1). To examine associations with fatness and/or body weight, we used birds of extreme phenotypes in F(2) and backcross populations with varying levels of abdominal fat weight per cent (%AFW) and body weight. We then assessed the level of gene expression by real-time PCR. In two genes, ADRB2 and MFGE8, we found significant association with %AFW. The ADRB2 gene was found to have a significantly higher expression in the liver of lean chickens compared with those of the fat individuals. We believe that this approach can be applied for the identification of other quantitative genes.

Biodiversity of 20 chicken breeds assessed by SNPs located in gene regions.

Twenty-five single nucleotide polymorphisms (SNPs) were analyzed in 20 distinct chicken breeds. The SNPs, each located in a different gene and mostly on different chromosomes, were chosen to examine the use of SNPs in or close to genes (g-SNPs), for biodiversity studies. Phylogenetic trees were constructed from these data. When bootstrap values were used as a criterion for the tree repeatability, doubling the number of SNPs from 12 to 25 improved tree repeatability more than doubling the number of individuals per population, from five to ten. Clustering results of these 20 populations, based on the software STRUCTURE, are in agreement with those previously obtained from the analysis of microsatellites. When the number of clusters was similar to the number of populations, affiliation of birds to their original populations was correct (>95%) only when at least the 22 most polymorphic SNP loci (out of 25) were included. When ten populations were clustered into five groups based on STRUCTURE, we used membership coefficient (Q) of the major cluster at each population as an indicator for clustering success level. This value was used to compare between three marker types; microsatellites, SNPs in or close to genes (g-SNPs) and SNPs in random fragments (r-SNPs). In this comparison, the same individuals were used (five to ten birds per population) and the same number of loci (14) used for each of the marker types. The average membership coefficients (Q) of the major cluster for microsatellites, g-SNPs and r-SNPs were 0.85, 0.7, and 0.64, respectively. Analysis based on microsatellites resulted in significantly higher clustering success due to their multi-allelic nature. Nevertheless, SNPs have obvious advantages, and are an efficient and cost-effective genetic tool, providing broader genome coverage and reliable estimates of genetic relatedness.