Flexible response: DNA supercoiling, transcription and bacterial adaptation to environmental stress.

T he impressive ability of bacteria to adapt to environmental change is a cause of both wonder and concern. These 'simple' organisms can cope with a wide range of variation in external physical and chemical parameters, such as temperature, pH, barometric pressure, osmolarity, oxygen availability, nutrient supply and the presence of toxic substances. Much of this adaptation involves resetting the transcriptional machinery of the cell, a topic of current intensive study. This research is influenced in some c a s e s by the realization that bacteria of industrial importance produce useful products when undergoing adaptation to stressful growth conditions, and in other cases by the observation that harmful bacteria frequently express virulenceenhancing factors in response to environmental cues. Many mechanisms determine the response of the transcriptional machinery to environmental signals. None of these mechanisms is entirely straightforward, even when an apparently simple one-to-one relationship seems to have been established between a given gene and a regulator. For example, the role of the RpoS {5 factor in promoting the transcription of a subset of eubacterial g e n e s in the stationary phase of growth has become complicated by issues concerning the precise mechanism through which RpoS expression is itself controlled 1. O n e of the most complicated issues in the field of transcriptional responses to environmental signals concerns the role of DNA topology and, in particular, DNA supercoiling. Supercoiling has the potential to influence strand separation in DNA, and hence DNA transactions such as transcription. It has been recognized since the mid-1980s that the degree of supercoiling of bacterial DNA varies with the growth conditions 2-4. If supercoiling varies in response to environmental influences, logic suggests that this could provide a (partial) mechanism through which transcription could be influenced by the environment. This is a nice hypothesis which deserves to be tested rigorously. Historically, the proU operon of Escherichia coli and Salmonella typhimurium has played an important role as a research tool in this field. This operon encodes a transport system for the osmoprotectant glycine-betaine, and its transcription is induced by about a hundredfold when the bacterial cell experiences an increase in osmotic pressure s. These increases in osmolarity also correlate with a reduction in the linking number of bacterial DNA 4. In other words, the DNA experiences an increase in negative supercoiling. Moreover, th i s change in linking number is accompanied by an increase in the free energy of the DNA, resulting in an enhanced ability to drive processes (such as cruciform extrusion), which depend on DNA strand separation 6. It s e e m s reasonable to propose that these same changes in supercoiling could drive processes (such as the formation of an open complex at the promoter) that would facilitate the transcription of operons such as proU. These discoveries, and the hypotheses that quickly developed from them, raise two separate yet