Monday 29 Apr 2019: Dynamics Seminar: Regulation of NF-kB dynamics and function through control of noisy feedback loops
Mike White - Manchester
LSI Seminar Room B 14:30-15:30
It is increasingly important to quantify the dynamic molecular processes that underlie cell-fate decisions in single cells. Early signalling events often occur within seconds of stimulation, whereas intracellular signalling and transcriptional changes may take minutes or hours. Cell-fate decisions can take many hours or days. Multi-parameter experimental and computational approaches to integrate quantitative measurements and mathematical simulations are required in order to understand the highly dynamic mechanisms that control cell fate (Spiller et al., (2010) Nature 465, 736).
Work on the NF-κB signalling system has indicated a key role for dynamic processes in signal transduction through this key pathway. We have used live cell imaging to show that NF-κB oscillates between the cytoplasm and nucleus in TNF-α stimulated cells (Nelson et al., (2004) Science 306: 705) and obtained preliminary support for the hypothesis that the frequency of these oscillations may control the pattern of downstream gene expression (Ashall et al., Science, (2009) 324: 242). Cell heterogeneity in oscillations may be regulated & advantageous (Paszek et al., (2010) PNAS 107: 11644).
One hallmark of the mammalian response to infection is the fever response. Temperature changes that occur in normal physiology (e.g. in response to fever or the circadian cycle) would be expected to change the rate of all biological reactions. We investigated whether temperature change within the physiological range (34 – 40C) affected NF-κB dynamics and function in human cell lines and mouse primary cells (Harper et al., PNAS (2018) 115: e5243). Changes of temperature caused a substantial change in the frequency of nuclear/cytoplasmic NF-κB oscillations, with higher temperatures increasing oscillation frequency. Analysis of a mathematical model of the NF-κB system suggested that it could not predict the observed temperature sensitivity. Our mathematical model and transcription data suggested an A20-dependent mechanism, involving temperature dependent delay in activation of A20 transcription. A20 knock-down of led to a substantially longer period of NF-κB oscillations. Remarkably, when A20 was knocked down there was no change in oscillation frequency between 37 and 40C. The change in temperature leads to tightly controlled temperature and TNFα-dependent changes in expression of an immunologically-relevant set of NF-κB target genes. A key set of immediate genes showed temperature control of the timing of their response, but many genes did not show significant control of their cytokine-induced expression at different temperatures (i.e. they were temperature compensated). We have analysed the functional networks of these altered genes and find that they participate in cross talk with MAP kinase, Wnt and p53-dependent pathways. These data suggest the hypothesis that temperature may subtly regulate the NF-κB system to modulate the cellular immune response to pathogens.
The different levels of A20 feedback in different cells appears to give rise to different cell specific oscillatory dynamics and gene expression responses. A real surprise was that levels of A20 mRNA were found to be very low, suggesting that this feedback will be very noisy and stochastic in single cells. We also previously described an A20-dependent refractory period in the response to repeated pulses of TNFα (Adamson et al., (2016) Nat. Comm. 7: 12057) . These and other data have led us to suggest that unknown epigenetic mechanisms may regulate A20. Recent data has implicated additional feedback loops through miRNAs in this process.