Integrating genomic analysis with the genetic basis of gene expression: preliminary evidence of the identification of causal genes for cardiovascular and metabolic traits related to nutrition in Mexicans

Abstract

Whole-transcriptome expression profiling provides novel phenotypes for analysis of complex traits. Gene expression measurements reflect quantitative variation in transcript-specific messenger RNA levels and represent phenotypes lying close to the action of genes. Understanding the genetic basis of gene expression will provide insight into the processes that connect genotype to clinically significant traits representing a central tenet of system biology. Synchronous in vivo expression profiles of lymphocytes, muscle, and subcutaneous fat were obtained from healthy Mexican men. Most genes were expressed at detectable levels in multiple tissues, and RNA levels were correlated between tissue types. A subset of transcripts with high reliability of expression across tissues (estimated by intraclass correlation coefficients) was enriched for cis-regulated genes, suggesting that proximal sequence variants may influence expression similarly in different cellular environments. This integrative global gene expression profiling approach is proving extremely useful for identifying genes and pathways that contribute to complex clinical traits. Clearly, the coincidence of clinical trait quantitative trait loci and expression quantitative trait loci can help in the prioritization of positional candidate genes. Such data will be crucial for the formal integration of positional and transcriptomic information characterized as genetical genomics.

Figure 1.

The numbers of transcripts detectable above background; most genes were detectable in ≥2 tissue types.

Figure 2.

Ingenuity Pathway Analysis functional class assignments for correlated transcripts. Ingenuity Pathway Analysis A is a powerful curated database and analysis system for understanding how proteins work together to effect cellular changes. We use this system to classify the correlated sets by gene function. The top 10 functional classes for each focal transcript are shown. There is considerable consistency in functional assignment across tissues. GHRL, ghrelin; TNF, tumor necrosis factor; PPARG, peroxisome proliferator-activated receptor gamma

Figure 3.

Selected functional comparisons of the top 10% (gray bars) and bottom 10% (black bars) of transcripts ranked by intraclass correlation coefficient. Bar height indicates relative enrichment for genes of a given functional class.

Figure 4.

The intraclass correlation coefficient (ICC) is positively correlated with the proportion of total variance due to cis variants: r = 0.07, P = 1.1 × 10⁻7.