In a startling exception to classical genetics, mice in a lab experiment have inherited an effect of an aberrant gene without inheriting the gene itself.

Experts say the result may someday help scientists understand aspects of diabetes, infertility and other problems. They also said that while such an inheritance pattern has been seen before, it's not clear how common or important it is.

DNA is the stuff of genes in mice and men. But the new study indicates DNA's chemical cousin, RNA, produced the odd-sounding result. In this case the result was mice with distinctive white tail tips.

For the study, reported in Thursday's issue of the journal Nature, scientists produced mice that carried one normal copy of a particular gene and one aberrant copy.

When the scientists bred them, each mouse passed along one of its gene copies to each offspring. The offspring, in turn, ended up with two copies, one from each parent.

Under the rules of classical genetics, one would expect that mice that ended up with two normal copies of the gene would show normal coloring. But surprisingly, 24 out of 27 mice with two normal copies still showed the telltale white patches associated with the aberrant copy.

When scientists bred their special mice to others that had only normal copies of the gene, most offspring with only normal gene copies still showed the distinctive coloring. Experiments showed the trait could be inherited through either the mother or father, and it went on for generations in absence of the abnormal gene.

How can this be? Researchers focused on sperm, which is simpler to analyze than an egg, and found evidence that RNA molecules there were carrying the hereditary signal.

For example, when RNA from mice bearing the aberrant gene was injected into early embryos, about half the resulting mice showed the distinctive white tail tip.

RNA normally delivers instructions from genes to a cell's protein-making machinery, so it makes sense that it might be involved in transmitting a gene's effect.

Just how it's operating in the mice is not clear, said lead study author Minoo Rassoulzadegan of the University of Nice Sophia Antipolis in Nice, France.

Paul Soloway of Cornell University, who was not involved in the study, said that when viewed in context with other research, the new evidence that RNA is involved is "compelling."

Scientists have shown that human sperm delivers RNA to eggs during fertilization, and the new work is important for establishing that sperm RNA actually has an effect, some experts said.

"I think it could open a whole new line of investigation" into what this sperm RNA does in people, said Chuck Ostermeier of The Jackson Laboratory in Bar Harbor, Maine.

For one thing, it might play a role in unexplained male infertility, noted Stephen Krawetz of the Wayne State University School of Medicine in Detroit.

Neither scientist was involved in the mouse study.

Soloway said it's not clear how common or important this kind of inheritance is, but he noted that a similar pattern of effects appeared in a 1997 analysis of a human gene that predisposes people to Type 1 diabetes.

The mouse study is not the first demonstration that a gene's effect can be inherited without the gene itself. That phenomenon, called paramutation, was shown 50 years ago in corn and later in other plants.

Similar findings have also been made in animals, but the new work is notable because it presents evidence for how it happens, Soloway said.