Cytokines are messengers of the immune system: How far does the message travel?
Cytokines are small, (often) secreted proteins that mediate cell communication in the immune system. They recruit and organize immune cells, and also scale and specify the immune response. We know a great deal about the identities of various different cytokines, and their effects on different cell types. However, the spatial extent of spread away from their source of production, and the factors that govern that spread, were unclear.
Two key barriers prevented our understanding. First, although in vitro experiments offer the tractability to make perturbations and precise measurements, they lack the high cell density and 3-D geometry that exists in vivo, making them effectively well-mixed. Second, we lacked a general, mathematical framework to describe their spread quantitatively. Such a framework would allow us to compare spread of the same molecule in different tissues or contexts, where variation in the parameters controlling spread might lead to different outcomes.
To overcome these barriers, we engineered ex vivo culture setups that mimic the tissue geometry and cell density that exists in vivo. We hypothesized that cytokine spread could be explained quantitatively by the biophysics of diffusion and receptor-mediated consumption. We used these experimental tools to rigorously test predictions from our model, which validated our hypothesis (Oyler-Yaniv et al., Immunity 2017). We also confirmed that this framework held true in vivo, and - most exciting - that natural immune responses tune some of the parameters in our model , which predictably extends or confines spread. In parallel, we discovered a novel mechanism by which certain plasma membrane lipids can modulate a cytokines’ effective diffusion rate (Oyler-Yaniv et al., Molecular Cell 2017). These lipids were aberrantly localized in cancer cells, leading to dysregulated inflammation in tumors.
In 2023, we confirmed that the same quantitative framework accurately predicts the spread of a different cytokine in human melanomas, and discovered that spatial variation in cytokine spread leads to non-genetic variation in protein expression and tumor cell evasion of T cell surveillance (Centofanti, Wang, & Iyer et al., PNAS 2023).
Our ongoing work addresses several related questions:
Can our quantitative framework describe the spatial spread of small, immuno-transmitters, such as 2’3’ cGAMP, and other signaling metabolites through cell channels and junctions?
What are the natural length-scales (average distance traveled by a cytokine molecule) for common T cell-derived cytokines in the lymphoid organs?
How do cytokine length-scales evolve and change over time as a result of feedbacks between cells communicating within the niche?
How do cellular interactions within cytokine niches impact immune function?
Viral infection begins within tissues at a single isolated site or a few scattered foci. Infection triggers a signal relay, which drives the formation of a zone of inflammation surrounding the infected site. This inflammatory microenvironment serves to constrain the early spread of the pathogen, and recruit and coordinate the activity of infiltrating innate and adaptive immune cells.
Activation of ubiquitously-expressed pattern recognition receptors by PAMPs (conserved molecular signatures of infection), within virus-infected cells is thought to be the crucial signal triggering the signal relay. However, our recent work challenges this paradigm, by demonstrating that the vast majority of virus-infected cells shut down translation of host proteins, preventing activation of the immune response (Sonnett et al., BioRxiv 2024). Nevertheless, a viral inflammatory microenvironment still forms, begging the question as to what triggers it’s formation, and how do cells within it collaborate to contain infection?
While transcriptomics measurements are widely-accessible and technically straightforward, mRNA expression often doesn’t accurately represent protein expression. With this in mind, we employ high throughput mass spectrometry-based proteomics to measure the cells’ proteome during viral infection. We demonstrated that cells infected with most viruses (excepting extremely attenuated ones), exhibit profound shut down in host translation, which completely hamstrings the antiviral response (Sonnett et al., BioRxiv 2024). This translational shutdown leads to the rapid decay of short-lived pro-survival proteins, whose loss dramatically sensitizes cells to cytotoxic cytokines.
We showed that a key cytotoxic cytokine (TNF-alpha) originates from activated, inflammatory macrophages present in tissues, and sensitization to death short-circuits the viral life cycle, restricting spread throughout the tissue (Oyler-Yanivs et al., Nat Comm 2021). In addition to investigating the timing in cell death during infection, we also established deep learning-based image analysis tools to study the different mechanisms by which infected cells die (Centofanti et al., MBoC 2025). These studies demonstrated that within an isogenic population of cells infected with the same virus, the mode of cell death can vary dramatically, impacting the resulting inflammatory response.
Our ongoing work asks:
By what mechanisms do virus-infected cells communicate with their neighbors to warn of infection?
How is generic cellular stress differentiated from pathogen-induced stress in infected cells and in their neighbors?
Can spatial inflammatory relays enable epithelial tissues to outpace the rapid rate of viral replication and confine spread?
Timelapse movie of Herpes Simplex Virus-1 infected cells (pink)
dying (green) after exposure to the cytotoxic cytokine TNF-alpha.