Defining the role of essential genes in human disease

Abstract

A greater understanding of the causes of human disease can come from identifying characteristics that are specific to disease genes. However, a full understanding of the contribution of essential genes to human disease is lacking, due to the premise that these genes tend to cause developmental abnormalities rather than adult disease. We tested the hypothesis that human orthologs of mouse essential genes are associated with a variety of human diseases, rather than only those related to miscarriage and birth defects. We segregated human disease genes according to whether the knockout phenotype of their mouse ortholog was lethal or viable, defining those with orthologs producing lethal knockouts as essential disease genes. We show that the human orthologs of mouse essential genes are associated with a wide spectrum of diseases affecting diverse physiological systems. Notably, human disease genes with essential mouse orthologs are over-represented among disease genes associated with cancer, suggesting links between adult cellular abnormalities and developmental functions. The proteins encoded by essential genes are highly connected in protein-protein interaction networks, which we find correlates with an over-representation of nuclear proteins amongst essential disease genes. Disease genes associated with essential orthologs also are more likely than those with non-essential orthologs to contribute to disease through an autosomal dominant inheritance pattern, suggesting that these diseases may actually result from semi-dominant mutant alleles. Overall, we have described attributes found in disease genes according to the essentiality status of their mouse orthologs. These findings demonstrate that disease genes do occupy highly connected positions in protein-protein interaction networks, and that due to the complexity of disease-associated alleles, essential genes cannot be ignored as candidates for causing diverse human diseases.

Figure 1. Physiological system analysis of disease genes.

Distribution of all Disease genes (D), Disease Viable genes (DV) and Disease Lethal genes (DL) in different disease classes, according to the physiological system affected. The D set corresponds to all disease genes without separation according to essentiality.

Figure 2. Network representation of protein-protein interaction between proteins chosen from the viable (V), lethal (L), disease viable (DV) and disease lethal (DL).

For clarity, interactions are only displayed in the figure if both interacting partners have the same classification (e.g. DV-DV interactions). However, statistical analysis (Table 2) was performed for all interactions (e.g. proteins of the same classification interacting and proteins of different classifications interacting). The color corresponds to node degree (relative to each network) as indicated for each panel, with the lowest degrees in red and highest degrees in purple. The node degree denotes the number of PPIs for a given gene.

Figure 3. GO analysis of disease genes.

Distribution of Viable (V), Lethal (L), Disease Viable (DV), Disease Lethal (DL), and all disease (D) proteins analysed for cellular localization according to GO terms.

Figure 4. Disease mechanism analysis.

Classification of disease mechanism in the total disease gene set (D, red bars), and Disease Viable gene (DV, green bars) and Disease Lethal gene (DL, blue bars) subsets. Other refers to diseases caused by chromosomal translocations or chimeric proteins.

Figure 5. Mode of inheritance of disease genes.

Proportion of disease genes inherited in an autosomal dominant pattern (AD), autosomal recessive pattern (AR) or X-lined pattern (X) in the total Disease gene set (D, red bars), or Disease Viable gene (DV, green bars) and Disease Lethal gene subsets (DL, blue bars).