Analytic theory of stochastic oscillations in single-cell gene expression

Single-cell stochastic gene expression, with gene state switching, transcription, translation, and negative feedback, can exhibit oscillatory kinetics that is statistically characterized in terms of a non-monotonic power spectrum. Using a solvable model, we illustrate the oscillation as a stochastic circulation along a hysteresis loop. A triphasic bifurcation upon the increasing strength of negative feedback is observed, which reveals how random bursts evolve into stochastic oscillations. Translational bursting is found to enhance the efficiency and the regime of stochastic oscillations. Time-lapse data of the p53 protein from MCF7 single cells validate our theory; the general conclusions are further supported by numerical computations for more realistic models. These results provide a resolution to R. Thomas' two conjectures for the single-cell stochastic gene expression kinetics.