Summary

Genetic variation between individuals plays a significant role in many different phenotypic traits. Considerable advances have been made in our ability to define the relationship between phenotype and genotype, and to map by linkage or association the genetic loci responsible for, or contributing to, a given phenotype. In many cases, however, the causative variant(s) are unknown, notably in common multifactorial traits, where fine mapping genetic associations and assigning causality to a specific functional genetic variant remains the exception rather than the rule.

The analysis of the genetics of gene expression described in this chapter goes an important way along the path to facilitating such work. To treat gene expression as a quantitative trait and apply the powerful genetic analytical approaches based on linkage and association seems in retrospect to be an almost obvious thing to do, but that is true of many notable scientific insights. In contrast to analysis of genetic variation in the context of the whole organism, the relative proximity in mechanistic terms of underlying genetic variation at the DNA level to gene expression within a cell might be expected to reduce the noise inherent to phenotypic analysis and increase the chances of success. We have seen how gene expression varies between and within populations, is a heritable trait, and has been successfully mapped, for transcript originating from a given gene, to specific loci by linkage and association. The high throughput micro-array technologies available for gene expression combined with high density SNP genotyping data have made the 'genetical genomics' approach increasingly popular and successfully applied to both the analysis of model organisms such as yeast and mice, as well as to human populations using lymphoblastoid cell lines and primary cells.

The synergy between analysis of the genetics of gene expression and genetic susceptibility to disease was clearly demonstrated in asthma where robust evidence of genome-wide association with disease at chromosome 17q21 was complimented by genome-wide expression profiling and mapping of gene expression by association using dense SNP genotyping (Dixon et al. 2007; Moffatt et al. 2007). The most strongly disease associated SNPs were also strongly associated with gene expression, in this case of ORMDL3. The genome-wide datasets for expression and SNP genotyping established by this and other studies represent important resources for future use by investigators and are publically available in searchable formats. The increasing application to primary human cells and tissues of the genetical genomics approach should further advance our understanding of the relationship between genetic variation and gene expression, which are often highly context-specific. Such studies, involving thousands of individuals often in family pedigrees, are enormously powerful and confirm the high heritability and important contribution of genetic variation to expression traits (Goring et al. 2007; Emilsson et al. 2008).

Multiple eQTLs are typically found to contribute to a particular gene expression trait, with individually modest effect sizes. Local and distant regulatory variants are implicated with likely c/s-acting effects most frequently identified among highly significant trait associations. Distant, likely trans-acting effects have been more challenging to resolve. Allele-specific gene expression has proved a powerful tool to compliment genetic analysis of gene expression, again with evidence of heritability and being a relatively common occurrence among non-imprinted autosomal genes. Application of genome-wide analysis to the analysis of allele-specific gene expression using either relative transcript abundance or RNA polymerase II loading should prove powerful approaches.

The increasing sophistication of our approaches to the genetics of gene expression, notably technological advances allowing the discrimination of the remarkable diversity that exists at the level of alternatively spliced transcript isoforms, should further advance our ability to define regulatory genetic variants. It is clearly essential to more accurately define and quantify the myriad of splice isoforms that are present across a range of tissues if we are to understand the relationship of gene expression to genetic variation. Similar approaches at a protein level will also prove highly informative and may be more relevant to the whole organism phenotype.

Major challenges remain, however, in ascribing functional mechanisms to specific regulatory variants. The different approaches discussed in this chapter are of great value but determination of a direct effect of the genetic variant requires complimentary testing in experimental systems, whether by manipulation in a model organism or in human cells (Chorley et al. 2008). For example, advances in technologies for transfecting DNA into cells allow for reporter gene and other DNA constructs to compare the effects of different variants directly on gene expression and have been widely used in this field (Rockman and Wray 2002). The consequences of sequence variation can also be predicted by bioinfor-matics with a sophisticated set of tools available for such analyses, for example based on predicted effects of variants on transcription factor binding sites. Direct assays of protein-DNA interactions are also possible both in vitro using the electrophoretic mobility shift assay and in vivo using ChlP.

Context remains paramount in such studies, as it does in the analysis of the genetics of gene expression. Only by analysis of genetic variation in an appropriate cell type and relevant conditions to the phenotype of interest will functionally important regulatory genetic variation be likely to be found, as exemplified by studies of genetic variation and globin gene expression. The themes of genetic variation and gene expression are continued in the next chapter where the remarkable diversity at the MHC is reviewed - a region highlighted by the genome-wide mapping of gene expression discussed in this chapter, and of genome-wide association studies for a range of autoimmune and infectious diseases (Section 9.3).

Was this article helpful?

0 0

Post a comment