One of the deepest divisions among living things is the split between prokaryote and eukaryote cells. In eukaryote cells, the chromosomes are enveloped in a layer of phospholipids - these cells have a 'true' nucleus surrounded by a nuclear membrane, something that's absent from prokaryotes. And there are other differences: eukaryotes either have, or have had, mitochondria (tiny organelles where aerobic respiration occurs) and plants & algae also have chloroplasts, & there's a lot of other cellular infrastructure besides. Now, we teach students about the origins of mitochondria & chloroplasts (most likely by the process of endosymbiosis, first put forward as an explanation by Lynn Margulis), but what about that nucleus?
There have been various hypotheses put forward to explain the origin of the eukaryote nucleus. One is that the chromosomes of the 'pre-eukaryote' were somehow surrounded by a section of the cell membrane. Certainly some prokaryotes have extensive infoldings of the cell membrane, but if this hypothesis was correct it should mean that the nuclear membrane is a single continuous sheet of phospholipid bilayer, & it's not.
As well as 'how', scientists have also tried to answer the 'why' question - just why do eukaryotes surround their chromosomes in a membrane. The usual answer - the one I was taught - is that this protects the nuclear contents; against what, we weren't told. If a nucleus is so useful, why did this structure never evolve in bacteria? Reading Nick Lane's Life Ascending pointed me in the direction of another possible answer, first proposed in 2006 by Koonin & Martin.
Introns are non-coding sequences. apparently the molecular corpses of jumping genes, found within eukaryote genes. 'Active' jumping genes are able to cut themselves out of a sequence of DNA using what could be described as molecular scissors, & reinsert a copy of themselves somewhere else. Introns have lost this ability to cut themselves out of a gene, but the cell must do so in order to remove these non-coding sequences prior to the process of translation & production of a functional protein. It seems that eukaryote cells do this using the same molecular scissors as jumping genes, modifying messenger RNA sequences by 'splicing' out the introns. However, this process takes time (Martin & Koonin, 2006).
In general, prokaryotes don't have introns. A piece of prokaryote mRNA will be translated into a protein almost as fast as the mRNA is itself being produced, because DNA, mRNA, and everything else needed to manufacture proteins are all in very close proximity within the cell. It's suggested that eukaryotes acquired introns very early on, picking them up from their new endosymbionts, the mitochondria. Supporting this hypothesis, the presumed bacterial ancestors of mitochondria do have a particular type of intron; and Martin & Koonin note that the evolutionary relationships of the proteins associated with the nuclear membrane also suggest that this membrane formed in cells that already contained mitochondria.
The presence of introns would have presented significant problems for the early eukaryote cells. Because RNA splicing is relatively slow, if mRNA was translated into proteins as soon as the mRNA was produced, many of those proteins would be faulty because they'd be produced by translating the 'wrong' intron information as well as the 'right' information encoded in the functional part of the gene (the exons). Any cell with structural features that provided at least some separation of transcription & translation would thus be at a selective advantage, because they wouldn't be wasting energy in producing faulty proteins . Martin & Koonin are suggesting that the nuclear membrane evolved in response to this selection pressure, providing a mechanism to separate the two processes - production of mRNA & production of proteins - to give sufficient time for the introns to be splieced out before the assembly of a protein could begin. It's a fascinating hypothesis, although only time (& lots of research) will tell us if it's a good model for the origin of the eukaryote nucleus.
(NB introns aren't always spliced out in the same way every time a gene is expressed. This underlies the fact that, while the human genome contains around 25,000 genes, our cells can contain 60,000+ diffierent proteins!)
W,Martin & E.V.Koonin (2006) Introns and the origin of nucleus-cytosol compartmentalisation. Nature 440: 41-45. doi:10.1038/nature04531
PS below is a good explanation of how & why introns are spliced out (there is a large range of excellent science videos at the site this came from):