Bacterial chemotaxis is based on modulation of the probability to switch the direction of flagellar rotation. Responses to many stimuli are transduced by a two-component system via reversible phosphorylation of CheY, a small cytoplasmic protein that directly interacts with the switch complex at the flagellar motor. We found that the chemorepellents indole and benzoate induce motor switching in Escherichia coli cells with a disabled phosphorylation cascade. This phosphorylation-independent chemoresponse is explained by reversible inhibition of fumarase by indole or benzoate which leads to an increased level of cellular fumarate, a compound involved in motor switching for bacteria and archaea. Genetic deletion of fumarase increased the intracellular concentration of fumarate and enhanced the switching frequency of the flagellar motors irrespective of the presence or absence of the phosphorylation cascade. These correlations provide evidence for fumaratedependent metabolic signal transduction in bacterial chemosensing.Motile procaryotes actively search for optimal living conditions. This phenomenon, called taxis, is based on the ability to sense environmental stimuli like chemicals, oxygen, and light and to respond by modulating the probability of switching the rotational sense of the motility organelle, the flagellar motor. The interplay of excitation and adaptation allows the cell to detect time-dependent changes in the strengths of different stimulus types; these changes are integrated to give a graded motor response (for reviews, see references 4, 12, and 22). Components of the bacterial sensory machinery have been identified and characterized by genetic, biochemical, and cell physiological methods. Bacterial chemotaxis provides a fascinating example of how computational information processing is performed by a network of proteins (8).

No abstract

Halobacterium salinarium responds to blue light by reversing its swimming direction. Fumarate has been proposed as one of the molecular components of this sensory system and is involved in the switching process of the flagellar motor. In order to obtain chemical proof for this role of fumarate, cells were stimulated with a pulse of blue light and lysed by rapid mixing with distilled water. The lysate contained fumarate in free and bound form, which were separated by ultrafiltration. The fumarate concentration in the low-molecular-mass fraction (< 5 kDa) of the lysate was assayed enzymatically and a light-induced increase was observed. Additionally, the total cellular fumarate content decreased in response to light, indicating that fumarate was released from a cellular pool rather than being formed by de novo synthesis. The light-induced release was not detected in a mutant defective in sensory rhodopsin-I and -II. Therefore it is concluded that photoreceptor activation rather than a direct effect of light on the activity of metabolic enzymes causes fumarate release. For each photoactivated sensory rhodopsin-II molecule at least 350 molecules of fumarate were liberated demonstrating efficient amplification. The rate of light-induced fumarate release is at least 10-times faster than the fumarate turnover number of the citric acid cycle which was estimated as approximately 4300 per cell and second. Therefore this metabolic process is not expected to be part of the signal transduction chain in the halobacterial cell.

The effect of CheY and fumarate on switching frequency and rotational bias of the bacterial flagellar motor was analyzed by computer-aided tracking of tethered Escherichia coli. Plots of cells overexpressing CheY in a gutted background showed a bell-shaped correlation curve of switching frequency and bias centering at about 50% clockwise rotation. Gutted cells (i.e., withcheA to cheZ deleted) with a low CheY level but a high cytoplasmic fumarate concentration displayed the same correlation of switching frequency and bias as cells overexpressing CheY at the wild-type fumarate level. Hence, a high fumarate level can phenotypically mimic CheY overexpression by simultaneously changing the switching frequency and the bias. A linear correlation of cytoplasmic fumarate concentration and clockwise rotation bias was found and predicts exclusively counterclockwise rotation without switching when fumarate is absent. This suggests that (i) fumarate is essential for clockwise rotation in vivo and (ii) any metabolically induced fluctuation of its cytoplasmic concentration will result in a transient change in bias and switching probability. A high fumarate level resulted in a dose-response curve linking bias and cytoplasmic CheY concentration that was offset but with a slope similar to that for a low fumarate level. It is concluded that fumarate and CheY act additively presumably at different reaction steps in the conformational transition of the switch complex from counterclockwise to clockwise motor rotation.