Speaking as a biologist... duh. Something that's been pretty obvious for a long time, really, when you look back at the data. Take the DES story, for example (diethylstilbesterol, a hormone-analog used an awful lot in the late '60's), which is just as wierd.

Mom takes DES to fight off chances of miscarriage and other pregnancy problems (doesn't work in humans). Mom suffers no significant ill effects (slight increase in breast cancer odds). Daughters have insanely higher risk for an extremely rare form of extremely aggressive cervical cancer (something like a change from 1/(1 x 10^8) down to 1/1000). Sons have no apparent problems (though they're now showing up with some consistent, non-lethal conditions). Grandsons have significant increases in specific defects. All without germline mutation.

It's something that I'm constantly coming up against--if there's a way for biology to do something in a wholly novel way, it will. Not only are biological organisms evolving, biological mechanisms are evolving in their own ways. We humans just think we're so smart and know everything. There will be surprises like this for the next century, is my guess.