Using Oscillations from Negative Feedback to Control Timing of Operations

Engineers have long recognized that negative feedback often does not restore a system to its target value, but results in an oscillation around it. This is observed when a thermostat controls a furnace or air conditioner—first the temperature exceeds the target, then it drops below the target, etc. Rather than stabilizing at the target temperature, it oscillates around it. In some cases oscillations generated by feedback mechanisms do dampen, but in other cases they sustain themselves. Negative feedback systems in biology also generate oscillations. Rather than just being a nuisance, as they often are in human-designed machines, oscillations are often employed as control systems in living organisms. Oscillations generate a repeating pattern of activity through time. The different activity states at different phases in the oscillation can be used to orchestrate operations of other mechanisms in time.

The glycolytic mechanism described above offers an example of feedback that generates oscillation. When Ghosh and Chance (1964) measured the concentration of NADH in their experimental preparation of yeast, they discovered it oscillated with a period of approximately one minute. Subsequently, Hess, Boiteux and Kruger (1969) demonstrated periodic oscillations in the concentrations of other reactants, with those generated in adjacent reactions generally being in phase with each other, but with phase shifts occurring at the phosphorylation of fructose-6-phosphate to fructose- 1,6-diphosphate and the dephosphorylation of phosphoenolpyruvate to pyruvate (left side of Figure 12.2). They also observed a small phase delay between glyceraldehye-3-phosphate and 1,3-diphosphoglycerate, which is the step at which the oxidation reaction occurs. This phenomenon, known as glycolytic oscillation, is explained by the feedback loop involving the allosteric enzyme phosphofructokinase-1 discussed above. When AMP or ADP activates it, more 1,3-diphosphoglycerate is produced, which provides the input to subsequent reactions. Eventually NADH and ATP levels increase. The increased concentration of ATP serves to inhibit the reaction (and the declining concentration of ADP as it is phosphorylated to ATP also reduces its activating effect). As NADH is reduced in the formation of lactate and as ATP is consumed in performing different cell activities, the concentrations of NADH and ATP decline again.

Although glycolytic oscillation is readily demonstrated in laboratory conditions, it is uncertain whether oscillations occur under physiological conditions and whether it has any physiological functions (Richard et al. 1994; Richard et al. 1996). But there are many other negative feedback systems that produce oscillations in biological systems that have been demonstrated to perform regulatory roles. Among the best known are circadian oscillations, which are exhibited in a host of our own activities from sleep to athletic performance and in physiological processes such as metabolism and immune responses. Although in the following section I will identify an important role for neurons in circadian rhythms in animals, these rhythms are in fact generated within nearly all cells of our bodies. The core mechanism involves a transcription-translation feedback loop whereby the proteins PERIOD (PER) and CRYPTOCHROME (CRY) feed back to inhibit their own transcription. The steps in the process (accumulation of PER and

CRY in the cytoplasm, transport to the nucleus, binding to the proteins that activate transcription and removing them from the promoter, and then degrading) together take about 24 hours. The result is that concentrations of these (and several other proteins that are centrally involved in the mechanism) oscillate with a period of approximately 24 hours. Some of these oscillating proteins in turn serve as activators or inhibitors to other genes, causing them to be synthesized at appropriate times of day (e.g., proteins required for immune responses are synthesized at those times of day when we are most likely to encounter other people).

Although the core of the circadian clock mechanism involves negative feedback, it is a much more elaborate mechanism than the simple feedback loop in glycolysis. It involves a set of proteins (including many more than those indicated above) dedicated to the task of generating an oscillation with a period of about 24 hours (Reppert and Weaver 2002; Zhang and Kay 2010). Moreover, it is paradigmatically a control mechanism. It regulates a host of other mechanisms by sending signals that alter constraints (enzymes) within them. The various mechanisms that the circadian clock regulates can continue to function without it. Under such circumstances, these mechanisms cease to be coordinated with the light-dark cycle of our planet. This can have untoward effects on the health of the organism. The circadian system is at a higher level than these individual mechanisms and, when functioning properly, imposes top-down control that enables these mechanisms to generate their respective phenomena when appropriate for the organism.