Physical mapping of genes in the porcine ovarian transcriptome

A muscle transcriptome analysis identifies positional candidate genes for a complex trait in pig

Muscle tenderness is an important complex trait for meat quality and thus for genetic improvement through animal breeding. However, the physiological or genetic control of tenderness development in muscle is still poorly understood. In this work, using transcriptome analysis, we found a relationship between gene expression variability and tenderness. Muscle (longissimus dorsi) samples from 30 F(2) pigs were characterized by Warner-Bratzler Shear Force (WBSF) on cooked meat as a measurement of muscle tenderness. Gene expression levels were measured using microarrays for 17 muscle samples selected to represent a range of WBSF values. Using a linear regression model, we determined that samples with WBSF values above 30 N could be effectively analysed for genes exhibiting a significant association of their expression level on shear force (false discovery rate <0.05). These genes were shown to be involved in three functional networks: cell cycle, energy metabolism and muscle development. Twenty-two genes were mapped on the pig genome and 12 were found to be located in regions previously reported to contain quantitative trait loci (QTL) affecting pig meat tenderness (chromosomes 2, 6 and 13). Some genes appear therefore as positional candidate genes for QTL.

Combining mouse mammary gland gene expression and comparative mapping for the identification of candidate genes for QTL of milk production traits in cattle

Background

Many studies have found segregating quantitative trait loci (QTL) for milk production traits in different dairy cattle populations. However, even for relatively large effects with a saturated marker map the confidence interval for QTL location by linkage analysis spans tens of map units, or hundreds of genes. Combining mapping and arraying has been suggested as an approach to identify candidate genes. Thus, gene expression analysis in the mammary gland of genes positioned in the confidence interval of the QTL can bridge the gap between fine mapping and quantitative trait nucleotide (QTN) determination.

Results

We hybridized Affymetrix microarray (MG-U74v2), containing 12,488 murine probes, with RNA derived from mammary gland of virgin, pregnant, lactating and involuting C57BL/6J mice in a total of nine biological replicates. We combined microarray data from two additional studies that used the same design in mice with a total of 75 biological replicates. The same filtering and normalization was applied to each microarray data using GeneSpring software. Analysis of variance identified 249 differentially expressed probe sets common to the three experiments along the four developmental stages of puberty, pregnancy, lactation and involution. 212 genes were assigned to their bovine map positions through comparative mapping, and thus form a list of candidate genes for previously identified QTLs for milk production traits. A total of 82 of the genes showed mammary gland-specific expression with at least 3-fold expression over the median representing all tissues tested in GeneAtlas.

Conclusion

This work presents a web tool for candidate genes for QTL (cgQTL) that allows navigation between the map of bovine milk production QTL, potential candidate genes and their level of expression in mammary gland arrays and in GeneAtlas. Three out of four confirmed genes that affect QTL in livestock (ABCG2, DGAT1, GDF8, IGF2) were over expressed in the target organ. Thus, cgQTL can be used to determine priority of candidate genes for QTN analysis based on differential expression in the target organ.