Quantifying cell fate change under different stochastic gene activation frameworks

Gene transcription is a stochastic process characterized by fluctuations in mRNA levels of the same gene in isogenic cell populations. A central question in single-cell studies is how to map transcriptional variability to phenotypic differences between isogenic cells. We introduced a measurable and statistical transcription threshold I for critical genes that determine the entry level of Waddington's canal toward a specific cell fate. Subsequently, J _I , which is the probability that a cell has at least I mRNA molecules of a given gene, approximates the likelihood of a cell committing to the corresponding fate. In this study, we extended the previous results of J _I of the classical telegraph model by considering more complex models with different gene activation frameworks. We showed that (a) the upregulation of the critical gene may significantly suppress cell fate change and (b) increasing transcription noise performs a bidirectional role that can either enhance or suppress the cell fate change. These observations matched accurately with the data from bacterial, yeast, and mammalian cells. We estimated the threshold I from these data and predicted that (a) the traditional human immunodeficiency virus (HIV) activators that modulate gene activation frequency at high doses may largely suppress HIV reactivation and (b) the cells may favor noisier (or less noisy) regulation of stress genes under high (or low) environmental pressures to maintain cell viability.