The kinetics of transcription elongation

Transcription is the first step in gene expression and as such it is the most highly regulated. In particular, regulation of the elongation phase of transcription has a wide range of functions. Understanding transcription cannot be complete without models of the system which predict and agree with the large amount of experimental data available. This dissertation describes the construction of a kinetic model for transcription elongation.;The model is experimentally and physically motivated. It is highly simplified in order to avoid overparameterisation and improvements are added incrementally. RNA folding is found to be of paramount importance in maintaining an elongation competent RNA polymerase. A phenomenological model of the effect of cotranscriptional RNA folding is described and used in the prediction of sequence dependent pausing during elongation. The pause positions predicted by the transcription model agree extremely well with experiment.;Further refinement produces a model which displays agreement with the results of many single molecule publications. The addition of an intermediate state is found to be necessary for agreement with observations. The model reproduces and predicts RNA polymerase velocity distributions, pause duration distributions, pause densities, pause frequencies and arrest densities. The force dependence of these quantities predicted by the model illustrates the difficulty in understanding some experiments which find force independence for many of the statistical properties of the system. The effect of temperature and the intermediate state are detailed.;The elongation mechanism is discussed as an isothermal ratchet which rectifies fluctuations be they equilibrium or not. A physical model for the translocation of RNA polymerase along double stranded DNA is presented. Future developmental directions are indicated by producing analytical solutions for the mean elongation velocity without making the usual pseudo-steady state hypothesis. The resulting expressions are fit to published force-velocity data. The simple model constructed in this manuscript fits the data and implies slow translocation at 0 applied force. However, the interpretation of the fits is model dependent. This is explicitly shown for two models which are more complex. Several other issues are discussed as they arise throughout this dissertation.