Figure 6. Continuum model.

See also Figure 6—figure supplements 1 and 2. (A) Proposed continuum model. In addition to switches to and from a closed state on the timescale of around ten minutes, the initiation rate in the active state fluctuates on a shorter timescale. (B) Simulation of the continuum model, resulting in temporal variation in the initiation rate (upper, green spikes). The short integration time of MS2 measurements (the time for which RNA is retained at the transcription site) means fluctuations in the active state of the gene can be visualized (lower). (C) In simulated smFISH data (right), using the RNA production events from the continuum model (B) and a cytoplasmic RNA decay time of 40 min, the distribution is well described by a standard two state bursting (negative binomial, NB) model. The long lifetime of cytoplasmic RNA averages out the temporal fluctuations in the initiation rate.

DOI: http://dx.doi.org/10.7554/eLife.13051.019

Figure 6—figure supplement 1. Potential mechanisms by which a continuum of activity (ii) may arise: (i) a ladder containing a large number of discrete states, each with a distinct initiation rate, caused by specific binding of transcription factors or epigenetic marks.

The states are too closely spaced to distinguish and count individual states. Alternatively, a model of fast switching between a primed state and an active state (iii) on a timescale of seconds or tens of seconds (shorter than the observation timescale of the MS2 system) produces a continuum of transcriptional activity. The fraction of time spent in the active state is modulated by the rates of switching into and out of the state, which depend on the local and time-varying concentration of polymerase and transcription factors.

DOI: http://dx.doi.org/10.7554/eLife.13051.020

Figure 6—figure supplement 2. Cartoon illustrating the continuum model and the predicted changes caused by TATA sequence modification.

Mutation of the TATA box may cause reduced rate of binding or increased rate of unbinding of activator. This results in a lower fraction of time spent in the active state, reducing the upper limit of initiation rates which can be realized. The rate of switching to and from the closed and inaccessible state is unaffected.

DOI: http://dx.doi.org/10.7554/eLife.13051.021