Bile acids stimulate invasion and haptotaxis in human colorectal cancer cells through activation of multiple oncogenic signaling pathways.

Bile acids are implicated in colorectal carcinogenesis as evidenced by epidemiological and experimental studies. We examined whether bile acids stimulate cellular invasion of human colorectal and dog kidney epithelial cells at different stages of tumor progression. Colon PC/AA/C1, PCmsrc, and HCT-8/E11 cells and kidney MDCKT23 cells were seeded on top of collagen type I gels and invasive cells were counted after 24 h incubation. Activation of the Rac1 and RhoA small GTPases was investigated by pull-down assays. Haptotaxis was analysed with modified Boyden chambers. Lithocholic acid, chenodeoxycholic acid, cholic acid and deoxycholic acid stimulated cellular invasion of SRC- and RhoA-transformed PCmsrc and MDCKT23-RhoAV14 cells, and of HCT-8/E11 cells originating from a sporadic tumor, but were ineffective in premalignant PC/AA/C1 and MDCKT23 cells. Bile acid-stimulated invasion occurred through stimulation of haptotaxis and was dependent on the RhoA/Rho-kinase pathway and signaling cascades using protein kinase C, mitogen-activated protein kinase, and cyclooxygenase-2. Accordingly, BA-induced invasion was associated with activation of the Rac1 and RhoA GTPases and expression of the farnesoid X receptor. We conclude that bile acids stimulate invasion and haptotaxis in colorectal cancer cells via several cancer invasion signaling pathways.

Activation of cellular invasion by trefoil peptides and src is mediated by cyclooxygenase- and thromboxane A2 receptor-dependent signaling pathways.

We have investigated the possible functional relationships between cellular invasion pathways induced by trefoil factors (TFFs), src, and the cyclooxygenases COX-1 and COX-2. Pharmacological inhibitors of the Rho small GTPase (C3 exoenzyme), phospholipase C (U-73122), cyclooxygenases (SC-560, NS-398), and the thromboxane A2 receptor (TXA2-R) antagonist SQ-295 completely abolished invasion induced by intestinal trefoil factor, pS2, and src in kidney and colonic epithelial cells MDCKts.src and PCmsrc. In contrast, invasion was induced by the TXA2-R mimetic U-46619, constitutively activated forms of the heterotrimeric G-proteins Galphaq (AGalphaq), Galpha12, Galpha13 (AGalpha12/13), which are signaling elements downstream of TXA2-R. Ectopic overexpression of pS2 cDNA and protein in MDCKts.src-pS2 cells and human colorectal cancer cells HCT8/S11-pS2 initiate distinct invasion signals that are Rho independent and COX and TXA2-R dependent. We detected a marked induction of COX-2 protein and accumulation of the stable PGH2/TXA2 metabolite TXB2 in the conditioned medium from cells transformed by src. This led to activation of the TXA2-R-dependent invasion pathway, which is monitored via a Rho- and Galpha12/Galpha13-independent mechanism using the Galphaq/PKC signaling cascade. These findings identify a new intracrine/paracrine loop that can be monitored by TFFs and src in inflammatory diseases and progression of colorectal cancers.

Structure of the Rho-activating domain of Escherichia coli cytotoxic necrotizing factor 1.

Certain uropathogenic and neonatal meningitis-causing strains of Escherichia coli express a 114 kDa protein toxin called cytotoxic necrotizing factor 1 (CNF1). The toxin causes alteration of the host cell actin cytoskeleton and promotes bacterial invasion of blood-brain barrier endothelial cells. CNF1 belongs to a unique group of large cytotoxins that cause constitutive activation of Rho guanosine triphosphatases (GTPases), which are key regulators of the actin cytoskeleton. This group also includes E. coli cytotoxic necrotizing factor 2 (CNF2, 114 kDa) and dermonecrotic toxins (DNT, 159 kDa) of Bordetella spp. with related sequences occurring in Yersinia spp. Here we show that the catalytic region of CNF1 exhibits a novel protein fold as determined by its 1.83 A resolution crystal structure. The structure reveals that CNF1 has a Cys-His-main chain oxygen catalytic triad reminiscent of enzymes belonging to the catalytic triad superfamily. The position of the catalytic Cys residue at the base of a deep pocket restricts access to potential substrates and helps explain the high specificity of this and related toxins.

Escherichia coli cytotoxic necrotizing factor-1 (CNF-1) increases the adherence to epithelia and the oxidative burst of human polymorphonuclear leukocytes but decreases bacteria phagocytosis.

Recruitment of polymorphonuclear leukocytes (PMNL) is a hallmark of both urinary and digestive infections caused by Escherichia coli. Cytotoxic necrotizing factor 1 (CNF-1) is a toxin produced by uropathogenic E. coli strains that mediates its effects via the activation of small GTP-binding proteins. However, the role and the consequences of CNF-1 on PMNL physiology remain largely unknown. In this study, we provide evidence that CNF-1 dramatically affects the PMNL cytoskeleton architecture by inducing an increased content of F-actin. Furthermore, we demonstrate that CNF-1 increases functional features of PMNL, such as superoxide generation and adherence on epithelial T84 monolayers, but significantly decreases their phagocytic function. Our results suggest that CNF-1 may behave as a virulence factor in urinary or digestive infection by stimulating PMNL cytotoxicity as a result of its enhancing effect on their adherence to epithelial cells as well as the production of radical oxygen products. Moreover, the decreased phagocytosis of PMNL induced by CNF-1 likely facilitates growth of bacteria. In these conditions, CNF-1 would intervene in the initiation and in the perpetuation of the inflammatory process.

The p21 Rho-activating toxin cytotoxic necrotizing factor 1 is endocytosed by a clathrin-independent mechanism and enters the cytosol by an acidic-dependent membrane translocation step.

Cytotoxic necrotizing factor 1 (CNF1), a protein produced by pathogenic strains of Escherichia coli, activates the p21 Rho-GTP-binding protein, inducing a profound reorganization of the actin cytoskeleton. CNF1 binds to its cell surface receptor on HEp-2 cells with high affinity (K(d) = 20 pM). In HEp-2 cells the action of CNF1 is not blocked in the presence of filipin, a drug described to reduce cholera toxin internalization by the caveolae-like mechanism. Moreover, HEp-2 cells, which express a dominant negative form of proteins that impair the formation of clathrin coated-vesicles and internalization of transferrin (Eps15, dynamin or intersectin-Src homology 3), are still sensitive to CNF1. In this respect, the endocytosis of CNF1 is similar to the plant toxin ricin. However, unlike ricin toxin, CNF1 does not cross the Golgi apparatus and requires an acidic cell compartment to transfer its enzymatic activity into the cytosol in a manner similar to that required by diphtheria toxin. As shown for diphtheria toxin, the pH-dependent membrane translocation step of CNF1 could be mimicked at the level of the plasma membrane by a brief exposure to a pH of

Deamidation of RhoA glutamine 63 by the Escherichia coli CNF1 toxin requires a short sequence of the GTPase switch 2 domain.

CNF1, a toxin produced by pathogenic Escherichia coli strains, deamidates the RhoA GTP-binding protein glutamine 63 and impairs RhoGAP-mediated GTP hydrolysis resulting in RhoA permanent activation. Using peptides derived from the RhoA sequence, we found that DTAGQEDYDRL (corresponding to RhoA 59-69 residues) was the minimum RhoA-derived peptide which could be deamidated in vitro by the CNF1 catalytic domain (CNF1-Cter). Site-directed mutagenesis outside the RhoA 59-69 sequence had no influence on glutamine 63 deamidation by CNF1-Cter. RhoA proteins with substitutions L57G, D65G, Y66G, or R70G were not affected in their ability to be deamidated by CNF1-Cter, whereas this was abolished by the R68G substitution. Arginine 68 is part of the DYDRL motif that is strictly conserved in Rho, Rac, and Cdc42 but not in other small GTP-binding proteins consistent with the observation that only Rho, Rac, and Cdc42 can be modified by CNF1.

Effects of cytotoxic necrotizing factor 1 and lethal toxin on actin cytoskeleton and VE-cadherin localization in human endothelial cell monolayers.

Integrity of the vascular endothelium is largely dependent on endothelial cell shape and establishment of intercellular junctions. Certain pathogenic bacterial toxins alter the cytoskeletal architecture of intoxicated cells by modulating the GTPase activity of p21 Rho family proteins. In the present study we have analyzed the effect of Rho-directed toxins on the actin cytoskeleton and monolayer integrity of endothelial cells. We report here that Escherichia coli cytotoxic necrotizing factor 1 (CNF1) activates Rho in human umbilical vein endothelial cells (HUVEC). In confluent monolayers, CNF1 treatment induces prominent stress fiber formation without significantly modifying peripheral localization of VE-cadherin, a specific marker of vascular endothelial cell adherens junctions. Further, Rho activation with CNF1 blocks thrombin-induced redistribution of VE-cadherin staining and gap formation in HUVEC monolayers. Inhibition of Rho by prolonged treatment of cells with C3 exoenzyme (Clostridium botulinum) eliminates actin stress fibers without disrupting the continuity of VE-cadherin staining, indicating that Rho-dependent stress fibers are not required for maintaining this adhesion receptor at sites of intercellular contact. Lethal toxin (Clostridium sordellii), an inhibitor of Rac as well as Ras and Rap, potently disrupts the actin microfilament system and monolayer integrity in HUVEC cultures.

Effect of cytotoxic necrotizing factor-1 on actin cytoskeleton in human monocytes: role in the regulation of integrin-dependent phagocytosis.

Cytotoxic necrotizing factor-1 (CNF1) is isolated from pathogenic strains of Escherichia coli and catalyzes the activation of Rho GTPases by the deamidation of a glutamine residue. This toxin induces stress fiber formation, cell spreading, and membrane folding and promotes phagocytosis competence in epithelial cells. We show that CNF1 induces morphologic changes in monocytic cells: polarized-like shape in THP-1 cells, lamellipodia, and cell spreading in adherent monocytes. CNF1 also increased filamentous actin (F-actin) content in a time- and dose-dependent manner. In addition, the toxin profoundly reorganized the actin cytoskeleton: redistribution of F-actin in polarized deformations of THP-1 cells and disorganization of microfilament network in monocytes. We also studied the effects of CNF1 on phagocytosis. It markedly impaired the ingestion of unopsonized zymosan involving CR type 3. However, CNF1 had no effect on the uptake of iC3b-coated zymosan or IgG-mediated phagocytosis of SRBC. In addition, CNF1 induced clustering of CR3 and Fc gammaRII (CD32) but selectively impaired the colocalization of CR3 with F-actin. It is likely that CNF1-induced reorganization of actin cytoskeleton down-modulates integrin activation-dependent phagocytosis by preventing the codistribution of CR3 with F-actin. CNF1 may control some features of integrin-dependent phagocytosis in myeloid cells through its action on Rho GTP binding proteins and cytoskeletal organization.

Escherichia coli cytotoxic necrotizing factor 1 effaces microvilli and decreases transmigration of polymorphonuclear leukocytes in intestinal T84 epithelial cell monolayers.

Cytotoxic necrotizing factor type 1 (CNF1), a 110-kDa toxin-like protein from pathogenic Escherichia coli strains, induces an actin cytoskeleton reorganization consisting of the formation of prominent stress fibers by permanent activation of the small GTP-binding protein Rho. Since p21Rho regulates tight-junction permeability and perijunctional actin reorganization in epithelial intestinal cells (A. Nusrat, M. Giry, J. R. Turner, S. P. Colgan, C. A. Parkos, E. Lemichez, P. Boquet, and J. L. Madara, Proc. Natl. Acad. Sci. USA 92:10629-10633, 1995), we used polarized T84 epithelial intestinal cell monolayers to examine whether CNF1 could affect microvillus structure, transepithelial resistance, and polymorphonuclear leukocyte (PMN) transmigration. Incubation of T84 cells with CNF1 did not influence transepithelial resistance, suggesting that barrier function and surface polarity were not affected by the toxin. However, CNF1 effaced intestinal cell microvilli and induced a strong decrease of PMN transepithelial migration in either the luminal-to-basolateral or the basolateral-to-luminal direction. CNF1 could thus be a virulence factor exhibiting a new type of combined activity consisting of effacing of microvilli and occlusion of the epithelial barrier to PMNs. Attenuated transepithelial migration of PMNs could result in the enhanced growth and protection of luminal bacteria.

Escherichia coli cytotoxic necrotizing factor 1 (CNF1), a toxin that activates the Rho GTPase.

Cytotoxic necrotizing factor 1 (CNF1), a 110-kDa protein toxin from pathogenic Escherichia coli induces actin reorganization into stress fibers and retraction fibers in human epithelial cultured cells allowing them to spread. CNF1 is acting in the cytosol since microinjection of the toxin into HEp-2 cells mimics the effects of the externally applied CNF1. Incubation in vitro of CNF1 with recombinant small GTPases induces a modification of Rho (but not of Rac, Cdc42, Ras, or Rab6) as demonstrated by a discrete increase in the apparent molecular weight of the molecule. Preincubation of cells with CNF1 impairs the cytotoxic effects of Clostridium difficile toxin B, which inactivates Rho but not those of Clostridium sordellii LT toxin, which inhibits Ras and Rac. As shown for Rho-GTP, CNF1 activates, in a time- and dose-dependent manner, a cytoskeleton-associated phosphatidylinositol 4-phosphate 5-kinase. However, neither the phosphatidylinositol 4,5-bisphosphate (PIP2) nor the phosphatidylinositol 3,4-bisphosphate (PI 3,4-P2) or 3,4,5-trisphosphate (PIP3) cellular content were found increased in CNF1 treated HEp-2 cells. Cellular effects of CNF1 were not blocked by LY294002, a stable inhibitor of the phosphoinositide 3-kinase. Incubation of HEp-2 cells with CNF1 induces relocalization of myosin 2 in stress fibers but not in retraction fibers. Altogether, our data indicate that CNF1 is a toxin that selectively activates the Rho GTP-binding protein, thus inducing contractility and cell spreading.

Toxin-induced activation of the G protein p21 Rho by deamidation of glutamine.

Pathogenic Escherichia coli are responsible for a variety of diseases, including diarrhoea, haemolytic uraemic syndrome, kidney infection, septicaemia, pneumonia and meningitis. Toxins called cytotoxic necrotizing factors (CNFs) are among the virulence factors produced by uropathogenic (CNF1) or enteropathogenic (CNF2) E. coli strains that cause diseases in humans and animals, respectively. CNFs induce an increase in the content of actin stress fibres and focal contacts in cultured cells. Effects of CNFs on the actin cytoskeleton correlated with a decrease in the electrophoretic mobility of the GTP-binding protein Rho and indirect evidence indicates that CNF1 might constitutively activate Rho. Here we show that CNF1 catalyses the deamidation of a glutamine residue at position 63 of Rho, turning it into glutamic acid, which inhibits both intrinsic GTP hydrolysis and that stimulated by its GTPase-activating protein (GAP). Thus, this deamidation of glutamine 63 by CNF1 leads to the constitutive activation of Rho, and induces the reorganization of actin stress fibres. To our knowledge, CNF1 is the first example of a bacterial toxin acting by deamidation of a specific target protein.

Molecular localization of the Escherichia coli cytotoxic necrotizing factor CNF1 cell-binding and catalytic domains.

Cytotoxic necrotizing factor type 1 (CNF1) induces, in epithelial cells, the development of stress fibres via the GTPase Rho pathway. We showed that CNF1 is able to modify Rho both in vitro and in vivo. Recombinant N-terminal 33kDa (CNF1Nter) and C-terminal 14.8-31.5 kDa (CNF1Cter) regions of the CNF1 protein allowed us to demonstrate that the N-terminal region contains the cell-binding domain of the toxin and that the C-terminal region is responsible for its catalytic activity. CNF1Nter lowered the activity of CNF1 when provided to cells before the toxin whereas CNF1Cter had no effect on CNF1 cell toxicity. CNF1Cter was sufficient to induce a typical CNF1 phenotype when microinjected into African green monkey kidney cells (Vero cells), and was able to modify Rho as previously reported for CNF1. The C-terminal domain lost its catalytic activity when deleted of various subdomains, suggesting a scattered distribution of catalytic-site amino acids. Elucidation of the CNF1 functional organization and analysis of amino acid homologies between CNFs (CNF1, CNF2), Pasteurella multocida toxin (PMT) and dermonecrotic toxin of Bordetella pertussis (DNT) allowed us to postulate that CNFs and DNT act on Rho via the same enzymatic activity located in their C-terminus, and that CNFs and PMT probably bind to analogous cell receptors.

Ras, Rap, and Rac small GTP-binding proteins are targets for Clostridium sordellii lethal toxin glucosylation.

Lethal toxin (LT) from Clostridium sordellii is one of the high molecular mass clostridial cytotoxins. On cultured cells, it causes a rounding of cell bodies and a disruption of actin stress fibers. We demonstrate that LT is a glucosyltransferase that uses UDP-Glc as a cofactor to covalently modify 21-kDa proteins both in vitro and in vivo. LT glucosylates Ras, Rap, and Rac. In Ras, threonine at position 35 was identified as the target amino acid glucosylated by LT. Other related members of the Ras GTPase superfamily, including RhoA, Cdc42, and Rab6, were not modified by LT. Incubation of serum-starved Swiss 3T3 cells with LT prevents the epidermal growth factor-induced phosphorylation of mitogen-activated protein kinases ERK1 and ERK2, indicating that the toxin blocks Ras function in vivo. We also demonstrate that LT acts inside the cell and that the glucosylation reaction is required to observe its dramatic effect on cell morphology. LT is thus a powerful tool to inhibit Ras function in vivo.

Effects of some saturated n-alkanes on the survival of Escherichia coli resting cells in sea water.

The influence of some linear alkanes on the survival of Escherichia coli natural sea water was investigated. Alkanes with fifteen or more carbon atoms induced a large decrease in viability of E. coli cells in natural sea water. In this case and for concentrations higher than 100 ppm, the loss of viability followed an exponential relationship with the carbon chain length. In the presence of 500 mg l-1 heptane, the survival was 1.6 times higher than that of controls. The progressive disappearance of heptane from the survival medium with a low and temporary accumulation by cells, suggests that this alkane may have been responsible for the increase of cell viability.

Influence of the RpoS (KatF) sigma factor on maintenance of viability and culturability of Escherichia coli and Salmonella typhimurium in seawater.

The sigma factor RpoS is essential for stationary-phase-specific, multiple-stress resistance. We compared the viabilities (direct viable counts) and culturabilities (colony counts) in seawater of Escherichia coli and Salmonella typhimurium strains and those in which rpoS was deleted or which were deficient in guanosine 3',5'-bispyrophosphate (ppGpp) synthesis (relA spoT). RpoS, possibly via ppGpp regulation, positively influenced the culturability of these bacteria in oligotrophic seawater. This influence closely depended, however, upon the growth state of the cells and the conditions under which they were grown prior to their transfer to seawater. The protective effect of RpoS was observed only in stationary-phase cells grown at low osmolarity. A previous exposure of cells to high osmolarity (0.5 M NaCl) also had a strong influence on the effect of RpoS on cell culturability in seawater. Both E. coli and S. typhimurium RpoS mutants lost the ability to acquire a high resistance to seawater, as observed in both logarithmic-phase and stationary-phase RpoS+ cells grown at high osmolarity. A previous growth of S. typhimurium cells under anoxic conditions also modulated the incidence of RpoS on their culturability. When grown anaerobically at high osmolarity, logarithmic-phase S. typhimurium RpoS+ cells partly lost their resistance to seawater through preadaptation to high osmolarity. When grown anaerobically at high osmolarity until stationary phase, both RpoS+ and RpoS- cells retained very high levels of both viability and culturability and then did not enter the viable but nonculturable state for over 8 days in seawater because of an RpoS-independent, unknown mechanism.

Effect of thermal, oxidative, acidic, osmotic, or nutritional stresses on subsequent culturability of Escherichia coli in seawater.

Survival of stressed Escherichia coli with or without the rpoS gene was assessed after 2 and 6 days in sterile seawater. Cells were submitted to thermal (48°C), acidic (pH 5.1), oxidative (H2O2 1mM), nutritional (C, N, P starvation), or osmotic (NaCl 0.5M) stresses for periods ranging from 0 to 4 h. We found a stress-mediated cross protection against seawater relative to controls. Viability was higher when cells were acid, oxidatively, nutritionally or osmotically stressed. Survival increased in cells stressed at 37°C as compared with 20°C. With the exception of osmotic stress, we found that this stress-induced cross protection was rpoS dependent.

The loss of culturability by Escherichia coli cells in seawater depends on availability of phosphate ions and phosphate transport systems.

Using strains with or without the PhoE porin or different components of the phosphate regulon, we determined that maintenance of the culturability of Escherichia coli in seawater depended significantly on the presence of structures allowing access of phosphate ions to the periplasm, then to the cytoplasm of cells. Cells totally deprived of the two main phosphate transport systems (Pit, Pst) exhibited the highest loss of culturability. Most of this effect resulted from the loss of the high-affinity Pst system, and more specifically that of the periplasmic phosphate-binding protein PhoS. Survival was enhanced in seawater supplemented with phosphate (0.5 mM), whether or not these structures were present. From an ecological point of view, it is assumed that the presence of phosphate ions, even at low concentrations, can influence the behavior of E. coli cells in seawater.

Sensitivity of Escherichia coli cells to seawater closely depends on their growth stage.

Sensitivity of Escherichia coli cells in seawater, considered in terms of culturability loss, was examined after different growth periods in a mineral medium supplemented with glucose (M9) at 37 degrees C under aerobic or anaerobic conditions. Their sensitivity varied considerably during the different growth phases and differed when cells were grown under aerobic or anaerobic conditions. Sensitivity of aerobic cells rapidly increased during the lag phase, then decreased during the exponential phase and became minimal during the stationary phase. Coliforms isolated from human faeces showed a similar sensitivity after incubation in wastewater at 37 degrees C for 3 h. The sensitivity phase was completely eliminated when cells were incubated with chloramphenicol. Variation of sensitivity in anaerobic cells according to their growth phase was comparable with that found for aerobic cells which had been left in seawater for a long period (6 d). However, for shorter periods in this medium (1-2 d), cells grown until the mid-exponential phase remained resistant to seawater. During the second half of the growth phase, they were as sensitive as aerobic cells at lag phase. Escherichia coli cells grown under anaerobic conditions, such as found in the intestine, progressively adapt to aerobic conditions after their transfer into aerated seawater and their sensitivity to seawater increases. On a practical level, these observations show that it is necessary to control accurately the age of cells before inoculation in seawater microcosms to conserve a comparative value in results. The importance of this factor is vital as all variations in sensitivity of cells to seawater according to their prior growth phase proved to be logarithmic functions of time.

OmpC and OmpF porins influence viability and culturability of Escherichia coli cells incubated in seawater.

The contribution of the major outer membrane porins OmpF and OmpC to the maintenance of viability and culturability of Escherichia coli cells in seawater was analyzed using isogenic mutant strains lacking one or both porins. Cells that possessed OmpF and OmpC survived better than those lacking one or both of them. However, the results differed, depending on whether the cells were adapted to high osmolarity or not before transfer to seawater. When cells were grown at low osmolarity, survival was largely influenced by porins, the OmpF+ strains surviving better than those lacking this porin. Addition of an OmpF plasmid to OmpF- OmpC- cells also improved their viability. When grown at high osmolarity, the role of porins was less critical since both the viability and culturability of the cells increased. However, cells that expressed only OmpC showed the most dramatic loss of viability. Cells lacking both OmpF and OmpC exhibited a higher loss of viability and culturability in seawater. Regarding the influence of porins on survival, these results show that the conditions that prevail during the growth of cells before their transfer to seawater are highly influential: cells that express the porin corresponding to the growth conditions they are in at the time of transfer survive better.

Intracellular accumulation of potassium and glutamate specifically enhances survival of Escherichia coli in seawater.

The high resistance of Escherichia coli grown in saline media to seawater was suppressed by an osmotic down-shock. The shock released several molecules into the medium, including potassium, glutamate, and glycine betaine when cells were previously grown in the presence of this osmolyte. Incubation of such sensitized cells in a solution containing K+ (80 mM) and glutamate (50 mM) at pH 7.4 restored their resistance to seawater up to a level close to that observed initially. The protective effect was partly due to the rapid accumulation of K+; a significant exponential relationship between intracellular concentration of K+ and resistance to seawater was observed. Glutamate was accumulated more slowly and progressively completed the action of K+. These data emphasize the specific influence of potassium glutamate on osmotically stressed E. coli cells. They confirm that regulation of osmotic pressure and, probably, of intracellular pH strongly enhances survival of E. coli in seawater. Osmotic fluctuations in waters carrying enteric bacteria from intestines to seawater, together with variations in their K+ and amino acid contents, could modify the ability of cells to survive in marine environments. These results demonstrate the need to strictly control conditions (K+ content, temperature) used to wash cells before their transfer to seawater microcosms. They suggest that the K+ and glutamate contents of media in which E. coli cells are transported to the sea can influence their subsequent survival in marine environments.