The BBC reports on a study in the American Journal of Physiology claiming to have evidence for nanobacteria. If I rely on the BBC account, though, the researchers seem to have missed the boat.
In the lab, they stained the specimens and examined them under a high power electron microscope.
The team found tiny spheres ranging in size from 30-100 nanometres (sic), which is smaller even than many viruses.
When the tissue was broken up, filtered to remove anything more than 200nm and the filtrate added to a sterile medium, the optical density - or cloudiness - of the medium increased.
This, the researchers argue, means the nanoparticles were multiplying of their own accord.
They should have taken samples of the medium at different times and tried to count the number of "nanobacteria" they could centrifuge out as a function of time. I'm not sure I believe optical density changes by themselves.