Joined: July 2005
For any thread participants who are not familiar with Margulis's endosymbiont model: In this model, one prokaryotic (bacterial) cell engulfs another, but the engulfed cell is not "digested" at all. Rather, it persists, the association benefits one or both cells, and the ability to maintain an endosymbiont gives a selective advantage to the host. Margulis's hypothesis was not taken seriously when she proposed it in the 1960s, but in the ensuing years, genetic tools were developed to test it. It turns out that mitochondria and plastids contain their own DNA, and have considerable structural and biochemical similarities to bacterial cells.
Incidentally, I work with a different kind of endosymbiont -- Wolbachia, a group of bacteria adapted to persist and reproduce within the cells of arthropod and nematode reproductive tissues, and to be transmitted from mother to offspring. Wolbachia is very good at "manipulating" host reproduction to make more copies of itself, usually by biasing a female's reproductive output towards making more daughters. (Males either don't transmit it, or else transmit it much less efficiently than females.) In some cases, Wolbachia infection comes with considerable cost to the host, but in others, its presence has become important to host survival or reproduction.