New research shows that cells often randomly deactivate one of a pair of gene copies or alleles, one of which they get from mom, the other from dad. This inactivation may potentially help explain why some children in a family may exhibit certain heritable disorders, whereas others do not.

Researchers at Massachusetts General Hospital (MGH) in Boston found that such disruptions may take place in as many as 1,000 genes in the autosomal genome (the part of our genetic repertoire that excludes sex cells), leading to different outcomes in the structures and levels of the proteins coded by these genes. These arbitrary alterations may, in some cases, clear a pathway for certain diseases such as Alzheimer's to manifest themselves.

"The general principle is that when a gene is turned on by a given cell or cell type that both alleles get expressed," says Harvard Medical School's Andrew Chess, a associate professor of medicine at MGH's Center for Human Genetic Research and co-author of the new study published in Science. "We've now found a large number of examples of an exception to that."

The finding could speak to more heterogeneity between individuals than can be accounted for by basic genetics. Though Chess and his colleagues do not know the mechanism by which alleles are silenced, the discovery that these events—called random monoallelic expression—occur so frequently suggests that epigenetic effects (influences on the activity of a gene that are not due directly to DNA mutations) may play a much more significant role in the development of some human diseases than previously believed.

Occurrences of allele inactivation are not new to researchers, although fewer than 1 percent of the genome undergoes a process called imprinting, in which an allele is transcribed into RNA and translated into a protein only if it comes from a specific parent. For instance, the IGF2 gene, which codes for the hormone insulinlike growth factor 2 in fetal liver cells, is only activated if it comes from the father. A version of monoallelic expression takes place in all female cells, where one of the two X chromosomes is randomly inactivated to eliminate redundancy. Genes that code for olfactory receptors, immunoglobulins (which produce antigens that trigger immune responses) and T cell receptors (that recognize antigens), also occur in populations of cells that express only one of their alleles, which actually results in a greater repertoire of antigens and receptors.

With the new work, the quantity of genes known to be vulnerable to this has ballooned to between 5 and 10 percent of the entire genome.

Chess's team took advantage of DNA microarray technology to survey the activity of certain gene variants in the genome of human B lymphocytes (white blood cells that, like T cells, help fight infection). By cataloguing point mutations in the genetic code, they could decipher when two different alleles were being transcribed from DNA into RNA (the template that provides the recipes to build specific proteins). They focused on a sample of just under 4,000 genes, nearly 400 of which appeared to undergo monoallelic expression.

Of those 400, several had previously been linked with human disorders, such as APP, the gene for amyloid beta precursor protein, which in excess quantities is believed to up the risk of Alzheimer's disease. Although no genes seemed immune to this process, researchers detected an abundance of cell receptor–coding genes in the mix. "The overabundance of receptors and other surface proteins suggests a role for monoallelic expression in each given cell's interactions with other nearby cells," the authors write in the Science paper.

Rolf Ohlsson, group leader of the mammalian epigenetics lab at Sweden's Uppsala University, wrote in an accompanying editorial that the findings may cause a shift in perspective on how scientists view illnesses believed to have a genetic component. "Anyone unfortunate enough to possess the 'wrong' set of monoallelically expressed genes might be susceptible to the earlier onset of a complex disease, such as Alzheimer's," his commentary read. "Considering the interplay among genotype, epigenotype and gene inactivation [these findings] will become more important in understanding developmental mechanisms and the penetrance of diseases in an individual as well as responses to medical treatments."

Next up: Chess says his lab will try to find triggers for monoallelic expression and probe whether genes can produce protein from one or both alleles at different times throughout a person's life span.