Losses of human disease-associated genes in placental mammals.

We systematically investigate whether losses of human disease-associated genes occurred in other mammals during evolution. We first show that genes lost in any of 62 non-human mammals generally have a lower degree of pleiotropy, and are highly depleted in essential and disease-associated genes. Despite this under-representation, we discovered multiple genes implicated in human disease that are truly lost in non-human mammals. In most cases, traits resembling human disease symptoms are present but not deleterious in gene-loss species, exemplified by losses of genes causing human eye or teeth disorders in poor-vision or enamel-less mammals. We also found widespread losses of and  genes, where loss-of-function mutations in humans protect from atherosclerosis. Unexpectedly, we discovered losses of disease genes () where deleterious phenotypes do not manifest in the respective species. A remarkable example is the uric acid-degrading enzyme , which we found to be inactivated in elephants and manatees. While loss in hominoids led to high serum uric acid levels and a predisposition for gout, elephants and manatees exhibit low uric acid levels, suggesting alternative ways of metabolizing uric acid. Together, our results highlight numerous mammals that are 'natural knockouts' of human disease genes.