I had the privilege of listening to Dr. Bhattacharya speak about the evolution of the chloroplast a few days ago. It inspired this notion in me: what if the nucleus was originally a parasitic archaeon that infected bacterial colonies?
One of the main tricks that he mentioned was the acquisition of a sugar transporter by the chloroplast/cyanelle/cyanobacteria within the eukaryotic cell; i.e. they changed to release some of the sugar that their photosynthesis generated into the cell.
The other point was that gene transfer seems to always go from the organelle to the nucleus; in at least one species of algae there’s no single chromosome left in the chloroplasts, just 14 single-gene plasmids.
There is an ongoing controversy over the origin of the nuclear pore complex, a remarkably complex “gate-keeper” that sits around holes in the nuclear membrane and controls what molecules can and can’t enter and leave the nucleus. It seems reasonable that a parasitic archaea would find controlling such things to be a selective advantage.
Basically the picture I’m painting is this: there are a lot of specialized bacteria, that do all sorts of interesting chemistry, often times in colonies that may even be membrane bound. By entering the colonies and releasing membrane translation proteins that cause the release of useful material (ATP or Sugar), you ensure your dinner; by translocating interesting genetic tricks from the organelles to the nucleus, you feed your grandchildren.
The way to falsify this prediction would be to do a systematic cross comparison of the cladistics of whatever dna-bearing organelles we can find, and see if there are similar patterns of heredity. If we were to find evidence that other organelles have ancestral prokaryotic cousins, it would be additional circumstantial evidence for this notion.
This reminds me that I need to look into why chromosomes are perpetually condensed in some species of algae, particularly the ones that don’t have histones. It would be nice, for example, to know how the heck they get transcription done.