Eye color, hair color, height and weight -- the list of traits passed down from one generation to the next is long. It's almost an old saw: Half our genes come from mom, half from dad. Before we even knew what genes were, we understood the basic principle of heredity, thanks to Gregor Mendel's peas. Now, 150 years later,  scientists at the Washington University School of Medicine in St. Louis say the old arithmetic no longer adds up.

Turns out our pot bellies, or propensity to fly off the handle may have nothing to do with our parents' genes and instead everything to do with maternal bacteria. Yep, germs. The DNA of microbes carried by the mother during pregnancy can be passed on to offspring, according to mouse studies published online Feb. 16  in the journal Nature.

"We have kept bacteria on one side of a line separating the factors that shape our development -- the environmental side of that line, not the genetic line," co-author Herbert W. Virgin IV told ScienceDaily. "But our results show bacteria stepping over the line. This suggests we may need to substantially expand our thinking about their contributions, and perhaps the contributions of other microorganisms, to genetics and heredity."

Bacteria, both on the skin and deep in the gut, is about as fundamental to human life as blood, with the majority of it being beneficial. Far smaller than human cells, bacteria is also abundant. Each of us has 10 times the amount of bacteria as human cells.

As is often the case in science, the discovery of the Washington University scientists was prompted by unrelated research. Virgin, the head of the department of pathology and Thaddeus Stappenbeck, a professor of pathology and immunology, were studying inflammatory bowel diseases, such as Crohn's and ulcerative colitis.

About half their mice had low levels of IgA, a gut antibody linked to these bowel conditions, and half had high levels. When they were bred, the mothers with low levels of the antibody produced offspring with low levels.

Investigating further, the researchers learned that a little understood bacterium, Sutterella, was likely responsible for spreading the low antibody levels. In other words, the mice weren't just passing on their own DNA, but the DNA of the germ Sutterella as well, resulting in low levels of the gut antibody IgA.

In the short term, Stappenbeck said, the implications for all experiments using genetically engineered mice are "profound," since this now explains a persistent problem in those experiments: mice offspring that appear with inexplicable new traits.

The implications for humans are potentially monumental, though as yet unknown.