The University Record, April 26, 1993

Discovered: some genes defy laws of inheritance

By Michael Harrison
Medical Center Public Relations

For almost 100 years, scientists have believed that we inherit working copies of genes from each parent and that when one of the two copies malfunctions, it can lead to illness.

New research at the Howard Hughes Medical Institute (HHMI) shows that some genes defy these laws of inheritance. For these genes, it’s normal for only one copy to work, and two active copies may cause disease—a radical concept in genetics, says principal researcher Andrew Feinberg, who is associate professor of internal medicine and human genetics and an HHMI associate investigator.

Feinberg and his colleagues have identified genes that remember which parent they came from and either function or turn themselves off, depending on their sexual origin. This gene gender memory, called imprinting, has been identified in other species but never before demonstrated in humans, he says.

They also found one gene that seems to trigger childhood tumors when copies from both parents work. Together, these discoveries hint that someday it may be possible to treat cancer by switching one gene back to its normal silent state.

Their work, published in the April 22 issue of Nature, proves for the first time that some genes behave differently depending on whether they came from a male or a female, regardless of the sex of the individual carrying the gene.

“We found one gene that works only if it was inherited in a sperm cell. If the same gene came from an egg, it is told to be silent for the life of the individual,” Feinberg says. This gene makes a product called insulin-like growth factor 2 (IGF2) and, Feinberg and research investigator Shirley Rainier discovered, loses its imprinting in a common type of childhood kidney cancer called Wilms’ tumor. The female-derived gene that is supposed to be silent was working in the cancer cells.

“This is a completely novel human genetic alteration,” Feinberg says. “While it’s widely known that genes malfunction when they have information inserted or deleted, no one has ever shown that two complete, functional genes have harmful effects.”

Wilms’ tumor, the second most common childhood solid tumor, often occurs in infancy. Feinberg’s team found loss of imprinting in another type of childhood solid tumor as well. “Our data suggests that relaxation of imprinting may be a first step in cancer. It causes over-expression of this growth factor so tumors grow,” says Feinberg, who also is a member of the U-M Comprehensive Cancer Center.

“A large number of childhood tumors show increased IGF2 expression, and it’s thought to be important in breast, colon and lung tumors in adults. We know that if you block IGF2 expression, some tumors don’t grow. My guess is that relaxation of imprinting will turn out to be a widespread alteration involved in many tumors,” he says. Loss of imprinting eventually could prove to be easier to fix than a classic mutation in which DNA is rearranged or missing, he explains.

“It will be a long time before we know how to correct gene mutations. This kind of alteration offers hope for cancer treatment because it is by definition reversible. Turning off one of the genes is something that normally happens.

“Right now, we don’t know how that works. The first thing we have to figure out is if the information that’s supposed to shut down the gene is irretrievably lost, or just suppressed,” he notes. However, imprinting—an entirely new pattern of inheritance—complicates genetic counseling because it bends the rules of dominant and recessive inheritance based on Austrian monk Gregor Mendel’s cross-breeding experiments 130 years ago.

“It means that some traits are inherited exclusively from the father and some are inherited only if they come from the mother,” Feinberg explains. An imprinted disease gene that is silent if carried in eggs might not affect a woman or her offspring, but her son could transmit an active copy in sperm. This would drastically change how counselors calculate the chance of passing on disease to offspring and future generations.

Feinberg’s research was supported by HHMI and the National Institutes of Health.