Imprinted QTL in farm animals: a fortuity or a common phenomenon? (Genetics)

The analysis of complex multifactorial traits has always been at the forefront in livestock species. Contrary to human and mouse, monogenic traits and disorders are generally easily recognized and removed by culling potential carrier animals. Initially, breeders applied these aspects in the algorithms and programs used to calculate breeding values for these animals, but with the development of detailed linkage maps and high-throughput genotyping systems in the last two decades, it formed the basis for the localization of many loci underlying such complex traits (for a recent overview of identified QTL in pigs and chicken, see http://www.animalgenome.org/QTLdb/ and http://acedb.asg.wur.nl/). Because these traits are quantitative in nature and are being influenced by a large number of genes as well as by environmental factors, they are referred to as quantitative trait loci or QTL. The initial algorithms used to analyze the data from a genome-wide scan, were restricted in considering only Mendelian inheritance and generally a single QTL per individual chromosome. More recently, several groups have started to extend these programs to include epistatic and epigenetic effects as well (Carlborg and Andersson, 2002; Carlborg etal., 2003; Knott etal., 1998; Jeon et al., 1999; Nezer et al., 1999; de Koning et al., 2000). Population structure and the outbred nature of most livestock breeds make them particularly well suited to address these genetic effects, something that is generally not possible in crosses between inbred laboratory rats and mice.


Although the potential implications of genetic imprinting with regard to quantitative traits in farm animals had already been described by de Vries et al. (1994), the interest for epigenetic effects in livestock species was particularly triggered by the identification of the callipyge locus in sheep. The callipyge locus was observed to be segregating in his herd by a sheep breeder in Oklahoma in 1983. The mutation results in an exceptional muscularity, and a subsequent genetic analysis revealed an exceptional behavior of this locus. Although the segregation of the callipyge phenotype clearly indicated an underlying mechanism of imprinting, the fact that homozygous carriers of the callipyge mutation did not express the callipyge phe-notype was an intriguing observation and the effect was referred to by the authors as polar overdominance (Cockett et al., 1996). In subsequent studies, it was shown that the callipyge mutation is located in a potential longe-range control element (LRCE) located between a group of paternally expressed (DLK1 and PEG11) and maternally expressed (GTL2 and MEG8) genes (Freking et al., 2002). The current working hypothesis for the observed polar overdominance at the callipyge locus is that (one of) the paternally expressed genes has a strong positive effect on muscle development and that (one of) the maternally expressed genes is a negative regulator of the paternally expressed gene(s). In this model, the LRCE is a gain-of-function mutation that results in an increase of both the maternally and paternally expressed genes.Although the callipyge mutation clearly affects a quantitative trait, its large effect on the phenotype enabled the analysis of the locus as a monogenic trait rather than a QTL.

The interest with regard to imprinting in relation to QTL was further boosted by the simultaneous independent findings of two research groups that a QTL on the tip of the p-arm of pig chromosome 2 (SSC2), affecting muscle mass and fat deposition, showed clear evidence of being maternally imprinted (Nezer et al., 1999; Jeon et al., 1999). Subsequent research recently led to the identification of a mutation in a regulatory site of the maternally imprinted gene 1GF2 as being the causative mutation underlying the imprinted QTL (Van Laere et al., 2003). The mutation is located in a CpG island located within intron 3 of the 1GF2 gene, which results in the abrogation of the binding of a repressor, resulting in an increased expression of the 1GF2 gene in pig muscle.

Triggered by the findings of the callipyge and 1GF2 gene, de Koning etal. (2000) decided to include a parent-of-origin effect in their statistical model for the genome-wide identification of QTL in a divergent cross between Chinese Meishan with European white pigs. Their results indicated a surprisingly large number of QTL that seemed to be affected by either maternal or paternal imprinting. The paternally expressed QTL on SSC2 in that study was recently (Jungerius et al., 2004) shown to be caused by the same mutation in 1GF2 that was identified by Van Laere etal. (2003). This mutation, however, was not responsible for the imprinted QTL on SSC2 affecting teat number (Hirooka etal., 2001; Jungerius et al., 2004). Additional analysis by these researchers and by others using another similar cross resulted in the further identification of several QTL for different growth-related traits on a number of different chromosomes (Rattink et al., 2000; De Koning etal., 2001; Milan etal., 2002; Desautes etal., 2002; Quintanilla et al., 2002).

Why do so many QTL seem to be affected by imprinting? There are several aspects to consider in dealing with this question (see Article 37, Evolution of genomic imprinting in mammals, Volume 1). One of the reasons that might cause the unexpectedly large number of imprinted QTLs is due to the nature of the traits being studied in these studies, which mainly were related to body composition and growth. The parental conflict hypotheses for imprinting (Sleutels and Barlow, 2002) suggests that there is a battle between male and female based on a tradeoff for the survival of offspring versus that of the survival and fecundity of the mother. Consequently, genes affecting growth might be expected to be affected by imprinting at a higher frequency then the average gene. The fact that the paternally expressed genes identified for callipyge and on SSC2 have a strong positive effect on growth would support this hypothesis. However, this is not the case for all the observed QTL effects. Furthermore, an imprinted QTL controlling susceptibility to trypanosomiasis with no clear relevance to growth was recently identified in mice (Clapcott et al., 2000). The experimental design of the QTL study and, in particular, the family structure of the population being studied has been shown to strongly affect the chance of identifying spurious imprinted QTL (De Koning et al., 2002). In particular, when the number of F1 animals is small or for smaller QTL effects when the founder lines are not fixed for different QTL alleles, spurious detection of imprinted QTL is a serious problem. A skewed distribution of uninformative markers between the male and female parents potentially could also result in the spurious identification of imprinted QTL in these crosses, although this pitfall could be excluded in the pig studies described above. Finally, other phenomena such as epistatic interactions between different genes might further complicate the correct identification of imprinting effects. Recently, Carlborg etal. (2003) have shown that epistatic effects also seem to play a major role in QTL identified for a large number of growth related traits.

Although several of the identified imprinted QTL effects are likely to be spurious effects, the studies described above provide accumulating evidence that imprinting plays a more important role in multifactorial traits than previously anticipated. Furthermore, the identification of the callipyge and 1GF2 mutations provide further insight into the importance of genetic variation within regulatory regions on quantitative traits. For animal breeding practice, the identification of major imprinted loci affecting body composition has several implications. It calls for a revision of the breeding value evaluation methods and breeding strategies that are currently solely based on the assumption of a large number of genes showing Mendelian expression. Detecting the QTL and confirming the mode of inheritance in commercial populations would open important new opportunities for pig-breeding companies. Imprinted QTL for fatness, for example, offers the opportunity to produce crossbred sows that have higher levels of fat reserves (beneficial for their health and reproduction), while their offspring have lower amounts of fat (requested by the consumer).

The results from the QTL studies in farm animals clearly have emphasized the importance of the inclusion of statistical testing for imprinting in the analysis of complex traits not only in animal genetics but in human medical genetics as well. Subsequent detailed molecular analysis and definite proof of the imprinting effects have to await the identification of the underlying genes responsible for the observed QTL effects.

Next post:

Previous post: