We report that the water permeability of wild-type Escherichia coli during exponential growth is comparable to that of an aqpZ disruption mutant. In contrast, an increase in permeability is observed for the wild type at the onset of the stationary stage with no significant corresponding change in the permeability of the mutant.Membrane channels are thought to be required for the osmoregulation of plant, animal, and bacterial cells (8). Aquaporin Z (AqpZ), which was identified a decade ago in wild-type Escherichia coli on the basis of sequence homology to other aquaporins, was the first water channel to be recognized in prokaryotes (5). In contrast to earlier reports that expression of E. coli aqpZ peaks during mid-exponential phase (6), a recent study by Soupene et al. challenged the hypothesis that expression of the gene is an adaptive response by rapidly growing cells and instead found that high levels of transcription of aqpZ occurred during transition into the stationary growth phase. In enriched medium, increased transcription was under the control of the RpoS sigma factor (9). Although it has been previously demonstrated that purified E. coli AqpZ increases the permeability of water in liposomes (2, 3), there is a notable absence of direct evidence that AqpZ mediates water permeability in vivo. We report here the water permeability of wildtype E. coli and an aqpZ disruption mutant as a function of the stages of cell growth.Relationship between growth rate and cell permeability of the wild type. Starter cultures of E. coli MG1655 (CGSC 7740) for our study were grown to full yield at 37°C in Luria-Bertani (LB) medium, and after 15 h they were diluted 1:100 into the same medium to initiate the experiments. At 1-h intervals, 30 ml of the culture was centrifuged for 5 min at 10,000 ϫ g at 5°C, the cell pellet was washed twice with 10 ml 100 mM (pH 7.4) phosphate-buffered solution (PBS), and the final cell pellet was resuspended in 30 ml PBS. The entire manipulation was complete, and permeability measurements were made within 10 min of sampling the culture. We employed a HI-TECH SF61 DX2 and stopped-flow methods to determine the permeabilities (1, 7). E. coli cells were subjected to an outwardly oriented osmotic gradient (see below), and the change in light scattering at em ϭ 600 nm was recorded as a function of time. Two features were observed in the transient spectra that followed mixing: a rapid exponential increase in scattering and a subsequent slower exponential decrease in scattering (data not shown). These events are attributed to egress of water from the cytoplasm and a subsequent slower movement of the osmolyte into the cytoplasm, as indicated by the direction of the change in scattering, the similarities of the observed rate of the first event to previous measurements of the egress of water from E. coli (1), and the insensitivity of the rate of the first event (and a sensitivity of the second event) to the chemical nature of the osmolyte (see below). The magnitude of change in scattering was proportio...