Uropathogenic Escherichia coli Superinfection Enhances the Severity of Mouse Bladder Infection

Urinary tract infections (UTIs) afflict over 9 million women in America every year, often necessitating long-term prophylactic antibiotics. One risk factor for UTI is frequent sexual intercourse, which dramatically increases the risk of UTI. The mechanism behind this increased risk is unknown; however, bacteriuria increases immediately after sexual intercourse episodes, suggesting that physical manipulation introduces periurethral flora into the urinary tract. In this paper, we investigated whether superinfection (repeat introduction of bacteria) resulted in increased risk of severe UTI, manifesting as persistent bacteriuria, high titer bladder bacterial burdens and chronic inflammation, an outcome referred to as chronic cystitis. Chronic cystitis represents unchecked luminal bacterial replication and is defined histologically by urothelial hyperplasia and submucosal lymphoid aggregates, a histological pattern similar to that seen in humans suffering chronic UTI. C57BL/6J mice are resistant to chronic cystitis after a single infection; however, they developed persistent bacteriuria and chronic cystitis when superinfected 24 hours apart. Elevated levels of interleukin-6 (IL-6), keratinocyte cytokine (KC/CXCL1), and granulocyte colony-stimulating factor (G-CSF) in the serum of C57BL/6J mice prior to the second infection predicted the development of chronic cystitis. These same cytokines have been found to precede chronic cystitis in singly infected C3H/HeN mice. Furthermore, inoculating C3H/HeN mice twice within a six-hour period doubled the proportion of mice that developed chronic cystitis. Intracellular bacterial replication, regulated hemolysin (HlyA) expression, and caspase 1/11 activation were essential for this increase. Microarrays conducted at four weeks post inoculation in both mouse strains revealed upregulation of IL-1 and antimicrobial peptides during chronic cystitis. These data suggest a mechanism by which caspase-1/11 activation and IL-1 secretion could predispose certain women to recurrent UTI after frequent intercourse, a predisposition predictable by several serum biomarkers in two murine models.     

Superinfection with R Factors by Transduction in Escherichia coli and Salmonella typhimurium

Superinfection immunity is found in the conjugal transfer of R factors between two fi(+) R factors and between two fi(-) R factors (fi = fertility inhibition), as we reported previously. In contrast, no reduction in the frequencies of transduction of an fi(+) R factor 222 was caused by the presence of fi(+) R factors in the recipients in transduction systems with phage P1kc in Escherichia coli K-12 and with phage P22 in Salmonella typhimurium LT-2. The absence of superinfection immunity in transduction may be due to the difference in the route of entry of the R factor. The frequencies of transduction of an fi(+) R factor were reduced, although slightly, by the presence of fi(-) R factors in the recipients. This reduction is probably due to host-controlled restriction of the entering fi(+) R factor by the fi(-) R factors in the recipients, since transduction of an fi(+) R factor by the transducing phage propagated on the strain carrying both fi(+) and fi(-) R factors was not reduced by the presence of homologous fi(-) R factors in the recipients. The fi(+) R factor 222, when transduced to the recipient strains carrying other R factors, recombined genetically at high frequencies with these resident R factors, regardless of their fi type.     

Nucleoid release from Escherichia coli cells.

The time course of morphological changes during lysis of Escherichia coli cells was examined with respect to an undisturbed release of nucleoids. The addition of detergents to plasmolyzed, osmotic sensitive cells resulted in the immediate reversal of plasmolysis followed by the appearance of rod-shaped ghost cells without any detectable spheroplast formation. Electron microscopic examination of the rod-shaped ghost cells revealed a zonal gap in the cell envelope, allowing the free release of the nucleoid. Due to the high ionic strength, a suitable cell lysis was shown to require higher incubation temperatures. However, in the absence of an appropriate control this may result in the sphering and vesiculation of ghost cell envelopes and even the unfolding of released nucleoids. To avoid this unfavorable consequence of lysis at high temperatures, a microscopic examination on the course of rod-shaped ghost formation is suggested.     

Peptidoglycan biosynthesis in stationary-phase cells of Escherichia coli.

The ability of stationary-phase cells of Escherichia coli W7 to incorporate radioactive precursors into macromolecular murein has been studied. During the initial 6 h of the stationary phase, resting cells incorporated meso-[3H]diaminopimelic acid at a rate corresponding to the insertion of 1.3 X 10(4) disaccharide units min-1 cell-1. Afterwards, the rate of incorporation dropped drastically (90%) to a low but still detectable level. Incorporation during stationary phase did not result in an increased amount of total murein in the culture, suggesting that it was related to a turnover process. Analysis of the effects of a number of beta-lactam antibiotics indicated that incorporation of murein precursors in stationary-phase cells was mediated by penicillin-binding proteins, suggesting that the activity of penicillin-binding protein 2 was particularly relevant to this process.     

Chemotactic signaling in filamentous cells of Escherichia coli.

Video techniques were used to record chemotactic responses of filamentous cells of Escherichia coli stimulated iontophoretically with aspartate. Long, nonseptate cells were produced from polyhook strains either by introducing a cell division mutation or by growth in the presence of cephalexin. Markers indicating rotation of flagellar motors were attached with anti-hook antibodies. Aspartate was applied by iontophoretic ejection from a micropipette, and the effects on the direction of rotation of the markers were measured. Motors near the pipette responded, whereas those sufficiently far away did not, even when the pipette was near the cell surface. The response of a given motor decreased as the pipette was moved away, but it did so less steeply when the pipette remained near the cell surface than when it was moved out into the external medium. This shows that there is an internal signal, but its range is short, only a few micrometers. These experiments rule out signaling by changes in membrane potential, by simple release or binding of a small molecule, or by diffusion of the receptor-attractant complex. A likely candidate for the signal is a protein or ligand that is activated by the receptor and inactivated as it diffuses through the cytoplasm. The range of the signal was found to be substantially longer in a cheZ mutant, suggesting that the product of the cheZ gene contributes to this inactivation.     

Production of giant cells of Escherichia coli.

Giant cells, with volumes up to 500-fold those of normal cells, have been produced by both genetic and pharmacological means in Escherichia coli K-12. In the genetic approach, an envB or mon mutation (conferring rounded or irregular morphology) was combined with a lon mutation (block of septation after irradiation). UV irradiation and subsequent incubation for 2 to 5 h in a rich medium supplemented with 1% sodium chloride led t; production of polymorphic giant cells. In the pharmacological approach, incubation of several different strains of E. coli K-12 with the drug 6-amidinopenicillanic acid (FL1060) in the same rich medium gave rise to a homogeneous population of smoothly rounded giant cells.     

Enumeration of stressed cells of Escherichia coli.

Cells of Escherichia coli K-12 were stressed by heating at 48 degrees C or by acid treatment at pH 4.2 for periods up to 1h. The addition of catalase to the selective medium increased the count of heat-stressed cells by 2.3-fold and acid-stressed cells by 4.8-fold. However, these values represented only a small percentage (3% for heat-stressed and 6% for acid-stressed cells respectively) of the population of injured but still viable cells. The addition of mannitol to the selective medium used to count acid-stressed cells did not increase the count. Whilst the presence of H2O2 in media may cause significant errors in the estimation of E. coli in certain situations these errors are unlikely to be significant in physiological studies of populations of cells injured by stress.     

The proton electrochemical gradient in Escherichia coli cells.

The internal pH of Escherichia coli cells was estimated from the distribution of either 5,5-[14C]dimethyl-2,4-oxazolidinedione or [14C]methylamine. EDTA/valinomycin treatment of cells was employed to estimate delta psi from 86Rb+ distribution concomitant with the delta pH for calculation of delta muH. Respiring intact cells maintained an internal pH more alkaline by 0.63-0.75 unit than that of the milieu at extracellular pH 7, both in growth medium and KCl solutions. The delta pH decreased when respiration was inhibited by anaerobiosis or in the presence of KCN. The delta muH, established by EDTA/valinomycin-treated cells, was constant (122-129 mV) over extracellular potassium concentration of 0.01 mM-1 mM. At the lower potassium concentration delta psi (110-120 mV) was the predominant component, and at the higher concentration delta pH increased to 0.7 units (42 mV). At 150 mM potassium delta muH was reduced to 70 mV mostly due to a delta pH component of 0.89 (53 mV). The interchangeability of the delta muH components is consistent with an electronic proton pump and with potassium serving as a counter ion in the presence of valinomycin. Indeed both parameters of delta muH decreased in the presence of carbonylcyanide p-trifluoromethoxyphenylhydrazone. The highest delta pH of 2 units was observed in the intact cells at pH 6; increasing the extracellular pH decreased the delta pH to 0 at pH 7.65 and to -0.51 at pH 9. A similar pattern of dependence of delta pH on extracellular pH was observed in EDTA/valinomycin-treated cells but the delta psi was almost constant over the whole range of extracellular pH values (6-8) implying electroneutral proton movement. Potassium is specifically required for respiration of EDTA-treated E. coli K12 cells since other monovalent or divalent cations could not replace potassium and valinomycin was not required.     

Branched Escherichia coli cells

We report that the normally rod-shaped bacterium Escherichia coli can form branched cells. These were found in strains in which chromosome replication or nucleoid segregation was disturbed, e.g. in minB mutants, intR1 strains, and in strains exhibiting stable DNA replication. Often, chromosome DNA was found to be located in the branch point of the cells. The branching frequency was dependent upon the growth medium: in rich medium no branched cells were found, whereas in minimal medium containing acetate and casamino acids the frequency of branched cells was increased. The genetic background of the strains also affected the tendency to branch. Furthermore, electron microscopy of thin-sectioned branched cells revealed additional membrane-like structures, which were not observed in wild-type cells. Finally, the branched cells are compared with bacteria that normally branch, and probable causes for branching in E. coli are discussed.     

Antitoxic immunity against the so-called lung toxin produced by Escherichia coli.

Cross neutralization test with antisera to crude haemolysins produced by some Escherichia coli strains indicated that there were no antigenic differences among the haemolysins tested. These crude preparates showed definite cytotoxicity which could also be cross neutralized by "antihaemolysin" sera. Neutralization experiments were performed in mouse lung test with homologous and heterologous anti-haemolysin sera, and with O and OK sera to the wild type strain and its toxic R mutant. The results showed that the immunity in the mouse lung model is antitoxic and antibacterial.     

Small arrays of electron-dense cylinders in Escherichia coli cells.

Electron microscopy of unstained Escherichia coli cells from cultures kept near 0 degrees C after incubation at 37 degrees C revealed small areas of geometrically arranged electron-dense cylinders. Their morphology, organization, and occurrence are described.     

Studies on superinfection immunity among transmissible plasmids in Escherichia coli

Superinfection immunity was studied by a method which permits the specific labeling of plasmid DNA following its entry into a recipient cell during conjugation. By measuring the incorporation of [³H]thymine during matings between a donor strain of Escherichia coli K12 carrying the R factor, Rl, and various recipients, we found that the presence in the recipient of a plasmid closely related to R1 (F or R factor 222), or isogenic to it (resistance transfer factor, from Rl), resulted in a reduction of 80 to 90% in the rate of [³H]thymine incorporation, relative to a mating with a plasmid-negative (F ^?) recipient. The DNA present in these recipients after 60 minutes of mating was further examined by neutral sucrose gradient eentrifugation. The DNA in the F⁺ and F ^? recipients sedimented similarly, in two major peaks at 50 S (relaxed circles) and 75 S (supercoiled circles). However, the DNA in the RTF recipient sedimented at rates intermediate between 50 S and 75 S. Pulse-chase experiments revealed that the DNA species seen after 60 minutes of an R1 × RTF mating are normal replicative intermediates which have disappeared by 60 minutes in the R1 × F^? or R1 × F⁺ matings.These data support genetic evidence suggesting that superinfection immunity is due to two distinct effects—entry exclusion and plasmid incompatibility. Thus, F (related to R1 but genetically compatible with it), as well as the incompatible plasmids, 222 and the RTF of R1 itself, when present in the recipient, greatly reduce the total synthesis of newly introduced R1 DNA in the recipient. We interpret this effect as entry exclusion. Incompatibility, manifested by RTF, but not F, further reduces the efficiency of conjugation by slowing the rate at which a newly acquired plasmid is replicated.     

The immunity (imm) gene of Escherichia coli bacteriophage T4.

The immunity (imm) gene of the Escherichia coli bacteriophage T4 effects exclusion of phage superinfecting cells already infected with T4. A candidate for this gene was placed under the control of the lac regulatory elements in a pUC plasmid. DNA sequencing revealed the presence of an open reading frame encoding a very lipophilic 83-residue (or 73-residue, depending on the unknown site of translation initiation) polypeptide which most likely represents a plasma membrane protein. This gene could be identified as the imm gene because expression from the plasmid caused exclusion of T4 and because interruption of the gene in the phage genome resulted in a phage no longer effecting superinfection immunity. It was found that the fraction of phage which was excluded upon infection of cells possessing the plasmid-encoded Imm protein ejected only about one-half of their DNA. Therefore, the Imm protein inhibited, directly or indirectly, DNA ejection.     

Cloning and biological characterization of the immunity region of Escherichia coli phage Mu.

The construction of three hybrid plasmids containing different parts of the left or immunity and end of phage Mu DNA is described. The recombinant plasmids pKN05 and pKN54 carry the HindIII.C and PstI.C fragments of Mu DNA, respectively. Neither of these plasmids expresses the killing function. Moreover, they do not allow plating of superinfecting Mu phages. Plasmid pKN62 harbors the fragment located in between the left PstI and EcoRI cleavage sites on Mu DNA, allows plating of superinfecting Mu phages, but does not express the killing function. These data suggest that the gene coding for the killing function is either positively regulated by a product from the EcoRI.C fragment, or the killing function requires a second product not coded for by pKN62. Mu Vir A- or Mu Vir B- phages are able to grow on bacteria harboring the recombinant plasmid pKN001 which carries the left and EcoRI-C fragment of Mu DNA. This indicates that the superinfecting phages can induce the corresponding gene functions from pKN001. No such induction could be detected in cells harboring the hybrid plasmids pKN05, pKN54 or pKN62.     

The phosphorylation of Escherichia coli isocitrate dehydrogenase in intact cells.

The isocitrate dehydrogenase of Escherichia coli ML308 can be reversibly activated by addition of pyruvate to cells growing on acetate [Bennett & Holms (1975) J. Gen. Microbiol. 87, 37-51]. By using cells pulse-labelled with [32P]Pi we showed that the activation and inactivation of the enzyme in these conditions correlate with its dephosphorylation and rephosphorylation respectively. Incubation of cell extracts prepared during an activation/inactivation cycle with purified isocitrate dehydrogenase phosphatase confirmed that the pyruvate-induced activation of the dehydrogenase goes essentially to completion. The results show that the reversible changes in the activity of the dehydrogenase in cells grown on acetate are solely due to phosphorylation/dephosphorylation. Inactive 32P-labelled isocitrate dehydrogenase was isolated from cells incubated with [32P]Pi in the presence of acetate. Both this material and purified enzyme phosphorylated in vitro were digested with chymotrypsin, and the phosphopeptides were isolated and analysed. Only one phosphopeptide was observed in each case; the results show that the residue phosphorylated in vivo is identical with that phosphorylated by purified isocitrate dehydrogenase kinase in vitro.