Feedback motifs for robust signal amplification and critical behavior

Many biological sensory systems are very sensitive to small input signals. A tempting hypothesis is that these systems operate close to bifurcation or critical points; at these points in parameter space the system's response exhibits a diverging susceptibility to the control parameter and small signals are amplified into a large collective response. However, a common concern is that being close to these points requires fine-tuning of parameters, which seems impossible for noisy biological systems. Based on several examples ranging from E. coli chemosensing to snake thermosensing, we have investigated a feedback motif inspired by self-organized criticality that robustly maintains these systems close to their respective critical point. The key ingredient is that the collective response feeds back onto the control parameter. Such a feedback scheme works well if just a single control parameter needs tuning. But robust critical behavior has also been found in high-dimensional systems like plasma membranes, made up from thousands of components. How do these systems find the critical manifold in the high-dimensional parameter space? Here we present an information theoretic perspective on this question and argue how critical behavior might occur naturally in high-dimensional systems.