Positioning of negative feedback loops within immune signaling pathways influences gene expression noise

Abstract

Signaling pathways depend on negative and positive feedback loops (NFLs and PFLs) to regulate internal noise. Across diverse organisms, signaling is regulated by NFLs that function at different cellular locations. These range from NFLs functioning upstream near signal-receiving receptors to those downstream within the nucleus. While previous studies have examined the relationship between NFLs, internal noise in signaling pathways, and network topology, none have directly addressed how the cellular location of NFLs impacts noise regulation. This is significant given the almost ubiquitous presence of multi-level regulation systems within signaling pathways. Here, we use stochastic models inspired by Imd and Toll signaling to address this gap within the context of immune signaling. We use both mechanistic and evolutionary models to demonstrate how noise is regulated by NFLs and how this, in turn, affects the hosts ability to fight off infection while minimizing immunopathologic effects of excessive immune gene expression. We found that downstream NFLs reduce noise in antimicrobial peptides (AMP) expression for some parameter values. On the other hand, upstream NFLs amplify the noise, but the presence of a strong PFL can reduce this noise. Our evolutionary simulations suggest that the mechanisms through which the downstream NFL operates within the cell can affect the evolution of the upstream NFLs. The results of our study provide insight into why distinct signaling pathways are regulated by varying numbers of NFLs, which operate in different cellular locations and employ diverse mechanisms to control gene expression. Author SummarySignaling pathways are noisy biological circuits. This noise is caused by random fluctuations in the number of proteins that function within these pathways. To properly respond to external stimuli, the ratio of noise to information needs to be minimized. To regulate noise, biological pathways produce proteins that either reduce (negative feedback) or amplify (positive feedback) signaling following stimulation. Negative feedback loops can shut down signaling by interfering with the first steps of signaling, which entail the detection of stimuli. Conversely, signaling might be left intact, and instead, negative feedback loops might interfere with the last step, which is the production of output. These regulatory differences can affect noise within signaling pathways. Here, we examined this using stochastic (inherently random) models to simulate immune signaling in response to pathogens. We found that negative feedback loops that function at later stages of signaling can decrease noise in the output, while negative feedback loops that act at earlier stages amplify the noise, but this noise can be reduced by the presence of a strong positive feedback loop.