page 883, Schneider and Haugh show that fibroblasts need stricter instructions and apparently lack the sophisticated signaling loops that are found in neutrophils.
Neutrophils take small differences in chemoattractant levels at the front and back of the cell and amplify them, via GTPase-driven positive feedback loops, into large differences in 3′ phosphoinositide (PI) production that drive the polarization of cytoskeletal changes. Neutrophils also adapt quickly to uniform chemoattractant levels by returning 3′ PI to near basal levels (possibly through global inhibition of 3′ PI production, as seen in Dictyostelium).
PDGF-stimulated fibroblasts, by contrast, do not adapt, prompting the authors to wonder how this sensing mechanism works. The team formulated a mathematical model to describe the simplest possible mechanism of this gradient sensing: PDGF-bound receptors activate PI3K locally, which creates 3′ PI; receptors compete for limiting amounts of PI3K; but no global regulatory mechanisms or feedback loops are included.
This basic model correctly predicted the PDGF gradient-sensing behavior of fibroblasts under all tested conditions. It also highlighted some differences from neutrophils. For one, fibroblasts have a much narrower range of chemoattractant concentrations (∼20 fold) within which they will respond well. They also require steeper gradients than do neutrophils. No evidence for the involvement of positive feedback loops was found.
Each sensing mechanism is well-suited to its owner. Neutrophils are like heat-seeking missiles tracking elusive bacterial invaders over long distances; they must respond and adapt quickly to directional changes. PDGF gradients direct fibroblasts into wounds from the adjacent tissue on a much longer time scale, during which the gradient orientation does not change. Desensitization of PDGF signaling through PI3K would be detrimental, as these pathways also control survival and proliferation.PDGF is perceived by a receptor tyrosine kinase, whereas neutrophils sense chemoattractants via G-protein–coupled receptors. The two receptor types might elicit diverse responses via distinct PI3K isoforms, as suggested by recent reports.