High efficiency? Receptors on the surface of a cell may have an unusual way of sensing chemicals.

Adapted from an illustration by Nicolle Rager, National Science Foundation

Cells Mind the Gaps

To be successful, a cell must have a nose for chemicals. A cell needs to know where nutrients are, for example, or--in the case of yeast--when a potential mating partner is close. Now, researchers report that cells may have developed a surprising strategy to optimize this chemical detector.

Scientists know that cells measure chemical concentrations in their environment via receptor proteins on their surface. When a chemical binds to the receptor, it generates a signal within the cell that alerts it to a nearby toxin, for example. In the 1970s, biophysicist Howard Berg and physicist Edward Purcell proposed that cells estimate chemical concentrations based on the fraction of time that receptors have molecules bound to them. (The more time the receptor is bound, the higher the concentration of the chemical.)

Until now, scientists have considered this to be the best possible strategy for cells. However, Robert Endres of Imperial College London and Ned Wingreen of Princeton University wanted to test this assumption. "The Berg and Purcell statement was such a straightforward thing to say that it didn't raise any eyebrows," says Wingreen. "But we asked: For the purposes of argument, is that really the best you can do?" Their answer turned out to be no, at least in theory.

To reach this conclusion, the researchers assumed unlimited processing power on the part of a cell. They then used a statistical method called the maximum likelihood technique, which helped them work out what processing of inputs at a cell's receptors would provide the best estimate of the chemical concentration. The duo found that, when cells want to detect chemicals in their environment, the best strategy is to discard information on how long those chemicals bind to receptors and instead record just the duration of the unbound intervals.

The basic reason why the alternative method improves sensitivity is fairly simple, researchers report in an article in press at Physical Review Letters. The fraction of time a receptor is bound depends on both how often a molecule hits it and the amount of time the molecule sticks to the receptor. In contrast, the length of an unbound interval depends only on how often the receptor gets hit and is thus more purely a measure of the concentration of molecules. In all, Endres and Wingreen's strategy halves the uncertainty in the concentration measurements and therefore allows cells to make measurements of a given certainty more quickly.

Of course, the scheme requires the cells to collect and process more information--specifically, the length of individual time intervals--and researchers aren't sure exactly how a cell would do that. But the team suggests that if the binding events can be made very short compared with the time between bindings, then simply recording a single impulse at each binding event and counting the number of impulses within a given period of time would be good enough. Wingreen thinks that eukaryotic cells, that is cells that contain a distinct nucleus, might take advantage of that approach because they withdraw each receptor as soon as a binding takes place. That suggests, he says, that the cell is paying attention to the binding event rather than the duration that the molecule stays bound.

As for practical applications, Wingreen says that the general sensing strategy might build into the artificial cells researchers are trying to create in synthetic biology. Or it may find applications in medical sensors and detectors.

Others had suggested that cells might use the rate of binding events to calculate concentrations as early as the 1970s. But cell biologist Peter van Haastert of the University of Groningen in the Netherlands says that the new research is important because it proves statistically that such an approach would be an improvement over that of Berg and Purcell's. He also believes it should be possible to test whether cells use such a strategy. The measurement of the rate of binding events changes significantly with temperature, van Haastert notes, so a cell measures that rate, and its sensitivity to chemical concentrations should change dramatically with temperature. If instead the cell follows the Berg and Purcell scheme, its sensitivity should remain more or less unchanged. The test could be made, he suggests, by exposing cultured muscle cells to adrenaline at temperatures between 20°C and 40°C.