Analysis of the regulation of transcriptional noise in NF-κB-induced transcription

Transcriptional noise can regulate biological processes including the immune response and reactivation of latent HIV, both of which depend on the transcription factor NF-κB. To better understand the molecular mechanisms that can alter this noise, we quantified transcripts in single cells and assayed events at gene promoters to inform a mathematical model of transcriptional bursting. Initially, we compared transcriptional noise from different HIV integrations to test the effect of basal chromatin environment on noise in response to stimulation of NF-κB by the inflammatory cytokine TNF. We then characterized basal and TNF-induced transcription of endogenous NF-κB targets to test whether the same model could describe their transcription and whether the same molecular events are implicated in regulation of gene expression noise.Chapter 2. Differences in chromatin environment have previously been show to correlate with transcriptional bursting behavior for different HIV integration sites with some basal transcriptional activity. We quantified transcription before and after reactivation of basally silent integrations and found different transcriptional noise profiles with similar mean transcript levels. Higher basal histone H3 acetylation at the integration site correlated with more skewed single-cell transcript distributions. Based on fits to a two-state promoter model, promoters in acetylated environments increased transcriptional burst size in response to stimulation, while those in deacetylated environments increased burst frequency. We found that increases in burst size or burst frequency correlated with differential regulation of RNAPII pause release. Using small molecule inhibitors, we remodeled basal chromatin and demonstrated that this was sufficient change transcriptional noise, RNAPII regulation, and bursting behavior in response to activation of NF-κB by TNF. Chapter 3. Transcriptional noise of endogenous NF-κB targets remains under explored, and has not been analyzed in terms of transcriptional bursting. We quantified transcription of six endogenous targets found in a range of chromatin environments before and after stimulation. We found similar effects of basal environment on transcriptional noise, with higher histone H3 acetylation leading to more skewed transcript distributions and higher induced burst size. We found that the same mathematical model of promoter activity used to describe HIV transcription fit five of our targets. As with the HIV integrations, increases in burst size correlated with increases in RNAPII promoter-proximal pausing. Using a histone acetyl transferase inhibitor to decrease basal acetylation, we found we could move genes along a spectrum of induced bursting behaviors. Additionally, we showed that inhibiting RNAPII pause release decreased induced burst sizes while increasing burst frequency. Taken together, our single-cell endogenous and HIV transcript measurements support existing theories of global constraints on mammalian gene expression noise