The kinetics and the temperature dependence of potassium loss from Escherichia coli cells treated with colicin K have been examined. At 37 C, after a single lethal hit, essentially all of the intracellular potassium is lost within the first few minutes of treatment. The initial rate of loss is linearly related to colicin concentration up to a multiplicity of 30. As the temperature is decreased over the range from 37 to 1 C, an increasing delay is seen in the initiation of potassium loss after colicin adsorption. This delay can be overcome by increasing colicin multiplicity and probably reflects an alteration of the cell membrane at these temperatures. A comparison of this effect with an apparently related effect of temperature on the action of irehdiamine A indicates that the delay may represent the inhibition of a transmission process occurring in the membrane.Colicin K is one of three colicins known to inhibit many energy-dependent processes in Escherichia coil (2,4,12,16) and has been previously shown to inhibit active transport of potassium and to cause the loss of intracellular potassium (17). These effects apparently occur by a direct action on the cell membrane (2) and have been assumed to result from changes which are transmitted over or through the membrane from a local site of action (9,15,16). A physical model which could acc'ount for the possible spreading of the local action of a single colicin molecule has been proposed (3,4). This model supposes that conformational changes may be propagated between membrane subunits in much the same manner as linked subunits of allosteric enzymes respond to the binding of a ligand molecule to one of the units (cooperativity within the membrane).The major evidence for a transmission of the local action of a single colicin molecule seems to be the single-hit killing shown by colicins. The conclusion has been drawn from single-hit killing that a cell attacked by a single colicin molecule has a definite probability of being killed (9,15,16), and this in turn implies that the effect of the single molecule can be transmitted to the whole cell. On this basis it might be expected that the effect of a single lethal unit on a cell should be demonstrable at the level of the affected process as well as at the level of cell killing. That is, at low multiplicities the fraction of the cell population which received single lethal hits should be correlated with the fraction of which the metabolism was blocked. In experimental systems for which data on this point are available, the relationship is not clear. The effects of colicins E2 and E3 are very slow to appear and require high multiplicities before deoxyribonucleic acid or protein synthesis are markedly affected (7,15,17). At low multiplicities (which are still sufficient to kill 95% of the cells), the effects are only slight. The effects of colicins K or El are quite rapid, but the fraction of cells killed is generally greater than the extent of inhibition (5, 6).Another conclusion from the single-hit killing has been that of ...