When discussing mobile elements in DNA, we usually talk about jumping genes, meaning open reading frames, or even better - fully functional proteins whose encoding genes can either transfer themselves from one place to the other inside its genome or copy itself, thereby creating apparently redundant amount of genetic material.
However, one recent paper that caught my eye is describing a slightly different picture: it is not the gene that is jumping, it is the promoter:
The genome of Tolypothrix sp. PCC 7601 carries two copies of a novel insertion sequence, ISTosp1. One of the two copies is located upstream of the gene encoding glutamyl-tRNA synthetase, an enzyme playing a key role in protein and pigment synthesis. The tnpA gene of the IS element and gltX were co-transcribed and their expression was transiently upregulated upon retrieval of the ammonium source irrespective of whether nitrate or no nitrogen source were available. The second copy is also transcribed and shows a similar regulatory pattern. Structural elements of the promoter (-10 and -35 sequences) directing the expression of the tnpA-gltX operon have been localized within the IS. Regulatory sequences involving the NtcA transcription factor in the control of tnpA-gltX expression were found both within and in sequences upstream of the insertion element. The expression of gltX in a closely related cyanobacterium, Nostoc sp. PCC 7120, which lacks the insertion upstream of gltX, decreased upon ammonium retrieval, a regulatory pattern that markedly differs from that observed in Tolypothrix sp. PCC 7601. ISTosp1 constitutes a good example of how cells can make use of a transposable element to evolve an original regulatory mechanism.
It is quite an interesting concept. Just imagine a nitrogen response element being able to jump from place to place in cyanobacterial genome and regulate a different sets of genes every time, in response to a nitrogen stress. One can only assume that the number of different regulatory possibilities would be huge and that different variants produced by the regulon shifting would have different fitness levels at different nutrient stress conditions. For example, a variant whose nitrogen response element has inserted itself before the genes for photosynthetic complexes would be able to decrease the level of photosynthetic activity in response to the lack of nitrogen, in addition to its previous abilities to regulate the photosynthesis in response to light conditions. Such variant would have much greater fitness than the non-shuffled "wildtype". This removes a whole dimension of randomality from the shoulders of the evolution of microorganisms...