Impairment of cell division in tolA mutants of Escherichia coli at low and high medium osmolarities.

Mutations in the tolA gene of Escherichia coli cause the cell to become sensitive to detergents and to some antibiotics, to release periplasmic enzymes and to be resistant to group A colicins; tolA mutations also lead to mucoid phenotype. TolA is a three-domain protein anchored in the inner membrane by its N-terminal domain. The second domain is proposed to span the periplasmic space and to interact with trimeric porins of the outer membrane. TolA proteins are considered to be located in the adhesion zones between inner and outer membranes. Our observations by confocal and electron microscopy have revealed that tolA mutants show modified morphology and produce DNA-free cells. Increasing or decreasing medium osmolarity amplifies these defects; mutants become essentially unable to locate the division site properly so that cells of highly unequal lengths are produced. Moreover, septation is impaired with asymmetric constrictions and oblique septa. These results suggest that TolA could play a role in positioning the division sites via the organisation of either the outer membrane or the possible adhesion zones.

Increased expression of a hemimethylated oriC binding protein, SeqA, in an aphA mutant.

In Escherichia coli, the origin of DNA replication, oriC, becomes transiently hemimethylated at the GATC sequences immediately after initiation of replication and this hemimethylated state is prolonged because of its sequestration by a fraction of outer membrane. This sequestration is dependent on a hemimethylated oriC binding protein such as SeqA. We previously isolated a clone of phage lambda gt11 called hobH, producing a LacZ fusion protein which recognizes hemimethylated oriC DNA. Very recently, Thaller et al. (FEMS Microbiol. Lett. 146 (1997) 191-198) found that the same DNA segment encodes a non-specific acid phosphatase, and named the gene aphA. We show here that the interruption of the aphA reading frame by kanamycin resistance gene insertion, abolishes acid phosphatase (NAP) activity. Interestingly, in the membrane of the null mutant, the amount of SeqA protein is about six times higher than that in the parental strain, suggesting the existence of a regulatory mechanism between SeqA and NAP expression.

Co-ordination between membrane oriC sequestration factors and a chromosome partitioning protein, TolC (MukA).

oriC DNA in the hemimethylated (but not in the fully methylated) state reacts with an Escherichia coli K-12 outer membrane preparation. This reaction is drastically reduced when the membrane preparation of a seqA null mutant is used. An in vitro reconstitution of the activity was undertaken by adding a partially purified SeqA protein to a seqA mutant membrane without success. A possible reason for this failure might be a profound modification of the outer membrane of the seqA mutant (as revealed by the fact that membrane from the mutant sediments more slowly than that from the wild type during ultracentrifugation). There is also a reduction in the content of OmpF protein. Moreover, one of the minor outer membrane proteins involved in partitioning of newly synthesized chromosomes, the ToiC (MukA) protein, was also found to be downregulated in the seqA mutant. This is also true of the hobH mutant grown in a high-osmolarity medium. Mutants of both seqA and hobH stop dividing after hyperosmotic shock, forming filaments (as observed in dam mutants).

Importance of the replication origin sequestration in cell division of Escherichia coli.

The DNA adenine methyltransferase of Escherichia coli methylates adenines at GATC sequences. The mutant deficient in this methylase has no apparent deficiency in the cell division process in spite of the absence of both synchrony in initiations of chromosomal DNA replication and sequestration of replication origin (oriC) at hemimethylated state. However, the dam mutant cannot resume cell division after hyperosmotic shock differing from the wild-type strain. This inhibition is not provoked by induction of the cell division inhibitor, SfiA protein. Although the FtsZ protein is present in the dam mutant in a reduced amount compared to wild-type, the quantitative difference of this protein is not the main reason of division arrest provoked by hyperosmotic shock. This observation supports the idea of oriC-membrane interaction playing a role both in chromosome partitioning and cell division as predicted by replicon theory.

Immediate and transient inhibition of the respiration of Escherichia coli under hyperosmotic shock.

The respiration of Escherichia coli is severely inhibited, during hyperosmotic stress period, as a consequence of plasmolysis; deplasmolysis allows the cell to recover respiration. A mutant lacking all K+ transport systems can neither deplasmolyze nor recover respiration unless betaine is present in the medium. Betaine, in these conditions, increases both cytoplasmic volume and respiration; this suggests a control of respiration by cytoplasmic volume.

A possible osmodependent protease in Escherichia coli.

When a strain of Escherichia coli, expressing a hybrid protein GalK-beta-Gal, is shifted to high osmolarity, the beta-galactosidase activity strongly decreases within 20 min of shock. The loss of beta-galactosidase activity results from degradation of the hybrid protein under osmotic stress. The results raise the possibility that osmotic stress induces a specific osmodependent protease.

Impairment of nucleoid segregation and cell division at high osmolarity in a strain of Escherichia coli overproducing the chaperone DnaK.

Escherichia coli strain WM1390, which overproduces 20-fold the chaperone protein DnaK, was able to grow exponentially without apparent abnormality in 300 mosM medium. In contrast, it showed both aberrant nucleoid segregation and strong inhibition of septation when shifted to high osmolarity. These impairments could not be accounted for by a bacteriocidal effect of the high DnaK content. Rather, the DnaK content appeared to promote faster growth than that of the parent C600, at least at high osmolarity in the presence of the osmoprotectant glycine betaine.

Potassium ions and changes in bacterial DNA supercoiling under osmotic stress.

Escherichia coli transiently increases both the [ATP]/[ADP] ratio and the negative supercoiling of plasmid DNA when it is shifted to high osmolarity. Here we report that a mutant lacking all saturable K+ transport systems increases the negative supercoiling of the plasmid DNA under upshock but cannot further relax DNA. The mutant dnaK756 behaves like the K+ transport mutant.

Role of heat shock protein DnaK in osmotic adaptation of Escherichia coli.

Escherichia coli can adapt and recover growth at high osmolarity. Adaptation requires the deplasmolysis of cells previously plasmolyzed by the fast efflux of water promoted by osmotic upshift. Deplasmolysis is essentially ensured by a net osmo-dependent influx of K+. The cellular content of the heat shock protein DnaK is increased in response to osmotic upshift and does not decrease as long as osmolarity is high. The dnaK756(Ts) mutant, which fails to deplasmolyze and recover growth, does not take up K+ at high osmolarity; DnaK protein is required directly or indirectly for the maintenance of K+ transport at high osmolarity. The temperature-sensitive mutations dnaJ259 and grpE280 do not affect the osmoadaptation of E. coli at 30 degrees C.

Glutathione-gated K+ channels of Escherichia coli carry out K+ efflux controlled by the redox state of the cell.

The kinetics of K+ efflux across the membranes of i) wild-type Escherichia coli poisoned by the thiol reagent N-ethylmaleimide, ii) K+ retention mutants and iii) glutathione-deficient mutants, have revealed a common "K+ leaky phenotype"; it is characterized by a very high rate of K+ efflux. The results suggest that the products of kefB and kefC genes could encode two K+ channels, both gated by glutathione. The possible function of these K+ channels seems to be a K+ exit controlled by the redox state of the cell; indeed, it can be inferred from the effects of several oxidants and reductants that turning on and off of the K+ efflux mediated by the channels can be correlated with the redox state of glutathione.

Glycine betaine reverses the effects of osmotic stress on DNA replication and cellular division in Escherichia coli.

The accumulation of glycine betaine to a high internal concentration by Escherichia coli cells in high osmolarity medium restores, within 1 h, a subnormal growth rate. The experimental results support the view that cell adaptation to high osmolarity involves a decrease in the initiation frequency of DNA replication via a stringent response; in contrast, glycine betaine transport and accumulation could suppress the stringent response within 1-2 min and restore a higher initiation frequency. High osmolarity also triggers the cells to lengthen, perhaps via an inhibition of cellular division; glycine betaine also reverses this process. It is inferred that turgor could control DNA replication and cell division in two separate ways. Glycine betaine action is not mediated by K+ ions as the internal level of K+ ions is not modified significantly following glycine betaine accumulation.

Turgor-controlled K+ fluxes and their pathways in Escherichia coli.

Escherichia coli like most gram-negative bacteria with walls maintains a cytoplasmic osmolarity exceeding that of the medium; the resulting hydrostatic pressure (turgor pressure) pushes the cytoplasmic membrane against the peptidoglycan and creates a tension in the two envelopes. Potassium is the only cation which takes part in the regulation of cellular osmolarity. The adaptation of intracellular K+ concentration to external osmolarity involves K+ turgor-controlled fluxes. When the medium osmolarity is raised an osmodependent influx of K+ can be observed; this is carried out by the K+ transport system TrkA which can also taken up rubidium. A specific and unidirectional pathway allows K+ ions to flow out of the cell when the medium osmolarity is decreased; this pathway reveals two characteristics: it has no affinity for rubidium and it can be blocked by the blockers of eukaryotic K+ channels. Osmodependent fluxes are turned on immediately after the medium osmolarity is disturbed; in contrast, they are turned off gradually as the rate of K+ fluxes approach zero. The rate of K+ influx seems to depend on the level of internal osmolarity and not on the extent of the increase in medium osmolarity. The rate of the efflux is directly proportional to the decrease in medium osmolarity and is independent on the level of internal osmolarity.

A high-affinity site for glutathione in the cytoplasm of Escherichia coli and its possible role in potassium retention.

Glutathione-deficient mutants of Escherichia coli, unlike the wild type, exhibit a fast leak and a fast turnover of K+ at steady state. In a medium with low K+ the growth rate is very slow when unsupplemented with glutathione; when supplemented with glutathione at a concentration as low as 0.1 microM the growth rate is similar to that of the wild type. The ability of glutathione to restore a wild-type growth rate can be accounted for by an immediate reduction of the K+ leak. Two analogs of glutathione also reduce the K+ leak: gamma-glutamylaminobutyrylglycine (ophthalmic acid) and alpha gamma-glutamylcystinylbisglycine. Glutathione binds with high affinity (Kd 50 nM) to a cytoplasmic protein; ophthalmic acid and alpha gamma-glutamylcystinylbisglycine compete with glutathione for the binding site (Ki 0.1 microM and 1 microM), thereby ruling out the possibility that the thiol is involved both in reducing the K+ leak and in binding to the high affinity binding site. For each of the three peptides the Kd for binding is similar to the minimal concentration that achieves the maximal reduction of the K+ leak. The results suggest that the binding of glutathione should play a role in the retention of K+ in E. coli.

Membrane potential stimulated binding of the maltose-binding protein to membrane vesicles of Escherichia coli.

We show in this study that radioactively labelled purified maltose-binding protein of Escherichia coli binds specifically to membrane vesicles, in the presence of maltose. When a potential is imposed across the membrane, the specific binding is increased, dependent on maltose, abolished in a mutant defective in the tar-gene product, one of the methyl-accepting chemotaxis proteins.

Glutathione and the gated potassium channels of Escherichia coli.

Glutathione-deficient mutants of Escherichia coli were found to require high potassium concentrations for growth, unless supplemented with glutathione. The unsupplemented mutants exhibited a rapid leak of potassium when transferred to a K+-free medium and a fast K+ turnover at the steady state of K+ accumulation, contrasting with the slow rate of the same processes in the wild-type. The steady-state level of K+ accumulation in low potassium medium increased immediately upon addition of glutathione, even in the absence of protein synthesis. K+-independent revertants were found to possess restored glutathione synthesis. Many properties of the glutathione-deficient mutants were identical with those of the potassium leaky K-B- and K-C- mutants, which, however, have a normal glutathione content. Both types of mutants differ from the wild-type in their response to thiol reagents in that no rapid loss of K+ is observed: they have, however, clear-cut differences under these circumstances. These results suggest that the products of trkB and trkC genes are essential for the formation of the potassium channel and glutathione plays an important role in the gating process.

The regulation of potassium fluxes for the adjustment and maintenance of potassium levels in Escherichia coli.

The regulation of K+ fluxes has been studied in Escherichia coli after depletion of K+ by an osmotic shock or at steady state of potassium accumulation. In the absence of a carbon source, bacteria accumulate K+ to an intracellular level near 0.1 M about half of actively metabolizing cells. During uptake the rate of net unidirectional influx decreased with time down to zero as a plateau was reached and no efflux was observed. Accumulated K+ was not exchanged with external K+ under these conditions. In the presence of a carbon source bacteria took up potassium to an intracellular concentration near 0.2 M. Efflux was delayed: it started only when 50% of final cellular K+ has been taken up. At steady state cellular K+ was exchangeable with external K+. The removal of the carbon source or the addition of respiratory inhibitors immediately stopped the K+ influx but did not affect efflux until the cellular K+ concentration has dropped to a level near 0.1 M; residual potassium was no longer exchangeable with external potassium. Maintaining such an impermeability to K+ ion does not require energy (delta psi or ATP), even though residual K+ is concentrated 100-fold compared to K+ in the medium. These results suggest that K+ efflux is dependent on active metabolism and on the concentration of intracellular K+ above a threshold. Unidirectional influx is also regulated by intracellular K+ according to a different concentration dependence. It appears that the trkA system, which is the only functional system under these experimental conditions is not a 'pump and leak' type of transport system.

Opening of potassium channels in Escherichia coli membranes by thiol reagents and recovery of potassium tightness.

The retention of high potassium levels in Escherichia coli is not dependent on intact energy metabolism, since without the presence of a carbon source or in the presence of energy inhibitors significant K+ gradients can be maintained. In contrast, with 0.5 mM N-ethylmaleimide, K+ depletion is immediate and complete. As a final result, intracellular K+ is approximately three times more concentrated than the K+ in the medium. Increase of K+ in the medium is immediately followed by K+ uptake whereas in the unpoisoned state only an increase in the osmotic pressure of the medium would result in an increase of the K+ pool. The intracellular K+ undergoes continuous turnover in the poisoned cells whereas in intact cells turnover is strictly dependent on the presence of a metabolizable carbon source. After removal of the thiol reagent the cell recovers its capacity to concentrate potassium. The recovery process is inhibited by energy inhibitors or by incubation at low temperature but not by chloramphenicol. It is only slightly slowed down by carbon or sulfur starvation. The leak provoked by N-ethylmaleimide is similar in wild-type E. coli cells when a derepressed kdp uptake system working in the micromolar range of the K+ concentration is responsible for the intracellular pool of K+ and when, in a medium millimolar K+ concentration range, the trkA and trkD systems are predominant.