Phase variation is a mechanism of ON-OFF switching that is widely utilized by bacterial pathogens. There is currently no standardization to how the rate of phase variation is determined experimentally, and traditional methods of mutation rate estimation may not be appropriate to this process. Here, the history of mutation rate estimation is reviewed, describing the existing methods available. A new mathematical model that can be applied to this problem is also presented. This model specifically includes the confounding factors of back-mutation and the influence of fitness differences between the alternate phenotypes. These are central features of phase variation but are rarely addressed, with the result that some previously estimated phase variation rates may have been significantly overestimated. It is shown that, conversely, the model can also be used to investigate fitness differences if mutation rates are approximately known. In addition, stochastic simulations of the model are used to explore the impact of 'jackpot cultures' on the mutation rate estimation. Using the model, the impact of realistic rates and selection on population structure is investigated. In the absence of fitness differences it is predicted that there will be phenotypic stability over many generations. The rate of phenotypic change within a population is likely, therefore, to be principally determined by selection. A greater insight into the population dynamics of mutation rate processes can be gained if populations are monitored over successive time points.
INTRODUCTIONPhase variation describes a process of reversible, highfrequency phenotypic switching that is mediated by DNA mutations, reorganization or modification. Phase variation is used by several bacterial species to generate population diversity that increases bacterial fitness and is important in niche adaptation including immune evasion (Saunders, 2003; Salaün et al., 2003). Being able to determine the rate at which these processes occur and the nature of any factors that influence them is integral to understanding the impact of these processes on the evolution and dynamics of the population as a whole and on the host-bacterium interaction. To do this, tools with which to reliably determine and compare phase variation rates within and between experiments and bacterial populations are needed. The estimation of mutation rates in bacteria has a long history. The methods in use, however, are not general across all systems and it is important that the assumptions behind the methods are recognized. Here, we present a new mathematical model that can be used to estimate phase variation rates in the presence of fitness differences, and explore the impact of proposed rates on population structure over time. It is also timely to compare and contrast some of the many different methods and terminology in this field. To put our approach in context, we begin with a brief review of the previous work in this field.