INTRODUCTION

Studies of spontaneous mutation rates in bacteria have tended to utilise exponentially growing populations, despite the fact that in their natural environments cells probably spend only a fraction of their lives in log-phase growth (85). Of equal relevance, especially clinically, are questions pertaining to the physiological state of bacteria during latent disease, and the effects of an altered physiology on mutagenesis. A common misconception is that spontaneous mutations arise as a consequence of errors made exclusively during DNA replication in actively growing cells. As early as the 1950s, Ryan (116) showed that mutations could arise in apparently static (or non-dividing) bacterial populations subjected to non-lethal selective conditions. Subsequent work extending this observation has redefined our understanding of the fundamental environmental and intracellular conditions promoting mutagenesis (18). While actively growing cells have the advantage of elevated rates of mutation per unit time, the increased periods spent in stationary phase by most cellular populations mean that even very low mutation rates can lead to the accumulation of high mutation frequencies per cell (13). Compounding the effects of an extended stationary phase are inducible systems, such as the bacterial SOS response (139) and its associated mutator polymerases (37), as well as hypermutable genes (29), all of which have potentially important implications for the development of subpopulations of bacteria with genetically encoded drug and stress resistance.