The C-terminal domain promotes the hemorrhagic damage caused by Vibrio vulnificus metalloprotease.

Vibrio vulnificus, an opportunistic human pathogen, produces a 45-kDa zinc metalloprotease (V. vulnificus protease; VVP) as an important virulence determinant. VVP injected intradermally into the dorsal skin causes the hemorrhagic damage through specific degradation of type IV collage in the vascular basement membrane. The N-terminal 35-kDa polypeptide (VVP-N), the catalytic domain, also evoked the hemorrhagic skin reaction within minutes. However, the hemorrhagic activity of VVP-N was one-third of that of VVP. Besides, the proteolytic activity of VVP-N toward the reconstituted basement membrane or type IV collagen was found to be about 50 % of VVP. VVP-N, like VVP, was quickly inactivated by an equimolar amount of alpha(2)-macroglobulin, a broad-spectrum plasma protease inhibitor. These findings indicate that the C-terminal 10-kDa polypeptide, the substrate-binding domain mediating the effective binding to protein substrates, functions to augment the hemorrhagic reaction of VVP.

Role of the sulphate assimilation pathway in utilization of glutathione as a sulphur source by Saccharomyces cerevisiae.

Mutants unable to grow on medium containing glutathione as a sole source of sulphur (GSH medium) were isolated from Saccharomyces cerevisiae strains carrying met17(deficiency of O-acetylserine and O-acetylhomoserine sulphydrylase). They were defective in the high-affinity glutathione transport system, GSH-P1. Newly acquired mutations belonged to the same complementation group, gsh11. However, it became apparent that gsh11 conferred the mutant phenotype not by itself but in collaboration with met17. Moreover, mutations conferring the defect in sulphate assimilation made the cell unable to grow on GSH medium in collaboration with gsh11. From this finding, we propose that the sulphate assimilation pathway acts as a sulphur-recycling system and that this function is especially vital to the cell when the supply of glutathione is limited.

The hemagglutinating action of Vibrio vulnificus metalloprotease.

Vibrio vulnificus protease (VVP), a 45-kDa zinc metalloprotease, consists of two functional domains: an N-terminal 35-kDa polypeptide having endoproteinase activity, and a C-terminal 10-kDa polypeptide that mediates the binding of VVP to the erythrocyte membrane. Therefore, VVP, but not its N-terminal endoproteinase domain alone, has agglutinating activity to rabbit erythrocytes. When a single zinc atom in the catalytic center was substituted by treatment with CuCl2 or NiCl2, proteolytic and hemagglutinating activities were reduced by Ni substitution but not by Cu substitution. Cu-treated 35-kDa polypeptide showed sufficient affinity of the catalytic center and weak binding ability to the erythrocyte membrane, but the Ni-treated polypeptide did not. These results suggest that the binding of endoproteinase domain to membrane is also necessary for hemagglutination.

The ability of Vibrio vulnificus to use a synthetic hydrophilic heme compound, Fe-TPPS, as a single iron source.

Vibrio vulnificus, an opportunistic human pathogen, can obtain iron from a variety of heme proteins. This process involves the digestion of heme proteins by an exoprotease to liberate protoheme (iron-protoporphyrin IX). In the present study, we tested whether this pathogen also uses a synthetic heme compound, Fe-alpha,beta,gamma,delta-tetraphenylporphine tetrasulfonic acid (Fe-TPPS), as an iron source. When inoculated into a medium containing Fe-TPPS, V. vulnificus L-180 multiplication was seen to be dependent on the concentration of the synthetic heme compound; a mutant lacking the ability to utilize protoheme did not multiply. Cells of the strain grown under the iron-restricted condition showed time-dependent uptake of Fe-TPPS. The ability to use either protoheme or Fe-TPPS was significantly reduced by the addition of an excess amount of free TPPS or Cu-TPPS. The data suggest that, V. vulnificus may assimilate Fe-TPPS, at least partially, through the same system as that for protoheme.

Glutathione transport systems of the budding yeast Saccharomyces cerevisiae.

The budding yeast Saccharomyces cerevisiae was shown to have two kinetically distinguishable glutathione transport systems. While one with high affinity (GSH-P1; KT = 0.045 mM) was regulated, the other with low affinity (GSH-P2; KT > 2 mM) was not. GSH-P1 was highly specific to glutathione, and its activity was quickly lost by suspending the cells in buffer solutions. This activity loss was not observed if glucose-containing buffer was used. In addition, rho-isolates had only about one half of the glutathione transport activity of the original (rho+) strain. Therefore, it is concluded that GSH-P1 is an ATP-driven transport system. Strong and moderate inhibition of GSH-P1 by protonophores and ionophores, respectively, are attributed to competition for ATP between GSH-P1 and proton- and cation-pumps, respectively.

Preparation and characterization of everted membrane vesicles from cells of Staphylococcus aureus.

We developed a method for the preparation of everted membrane vesicles from cells of Staphylococcus aureus. The cells were first treated with ampicillin to weaken the peptidoglycan layer, then the cells were passed through a French press cell. The resulting vesicles were roughly 0.1 microm in diameter, judging from electron microscopic observations. We detected fairly high membrane-bound ATPase activity in the membrane vesicles. We observed respiratory-driven quenching of quinacrine fluorescence, which indicates that inward H+ transport took place. These results indicate that the vesicles are everted. We characterized the membrane-bound ATPase. We also detected Na+/H+ antiport, erythromycin/H+ antiport and chloramphenicol/H+ antiport activities in the membranes of S. aureus.

Functional domains of a zinc metalloprotease from Vibrio vulnificus.

Vibrio vulnificus, an opportunistic human pathogen causing wound infection and septicemia, secretes a 45-kDa metalloprotease (V. vulnificus protease; VVP). A plasmid which carries the entire vvp gene subcloned into pBluescriptIIKS+ was transformed into Escherichia coli DH5alpha for overproduction of the protease. The 45-kDa recombinant protease (rVVP) was isolated from the periplasmic fraction of the transformant by ammonium sulfate precipitation followed by column chromatography on phenyl Sepharose. Biochemical characterization of the isolated rVVP showed that the recombinant protease was identical to that produced by V. vulnificus. When rVVP was incubated at 37 degrees C, a 35-kDa fragment was generated through autoproteolytic removal of the C-terminal peptide. This 35-kDa fragment (rVVP-N) was found to have sufficient proteolytic activity toward oligopeptides and soluble proteins but had markedly reduced activity toward insoluble proteins. Lineweaver-Burk plot analysis indicated increased Km values of rVVP-N for all of the protein substrates. rVVP, but not rVVP-N, was shown to agglutinate rabbit erythrocytes, bind to the erythrocyte ghosts, and digest the ghost membrane proteins. These results strongly suggest that rVVP (and VVP) consists of at least two functional domains: an N-terminal 35-kDa polypeptide mediating proteolysis and a C-terminal 10-kDa polypeptide which may be essential for efficient attachment to protein substrates and erythrocyte membranes.

Vibrio mimicus attaches to the intestinal mucosa by outer membrane hemagglutinins specific to polypeptide moieties of glycoproteins.

Vibrio mimicus is the closest organism to Vibrio cholerae. V. mimicus E-33, which is a highly adhesive and enteropathogenic strain, is known to produce three types of hemagglutinins (HAs), i.e., a 31-kDa exocellular metalloprotease (Vm-HA/protease), lipopolysaccharide (Vm-LPSHA), and a 39-kDa major outer membrane protein (Vm-OMPHA). Hemagglutination induced by Vm-LPSHA and Vm-OMPHA was inhibited by glycoproteins, including mucin, fetuin, and asialofetuin, but not by monosaccharides, disaccharides, or N-acetylated saccharides. The inhibitory potential of each glycoprotein for Vm-OMPHA was greatly augmented by treatment with a glycolytic enzyme such as beta-D-galactosidase or beta-D-glucosidase, while pronase treatment achieved complete abolition of the inhibitory potential. The inhibitory ability of the glycoproteins for Vm-LPSHA was also abolished by pronase treatment; however, glycolytic enzyme treatment showed no effect. Hence, the polypeptide portion of glycoproteins may directly associate with Vm-OMPHA and Vm-LPSHA, but the sugar moiety may act as a barrier to interaction with Vm-OMPHA. The glycoproteins as well as Fab antibodies against Vm-OMPHA and Vm-LPSHA eliminated the ability of E-33 cells to agglutinate rabbit erythrocytes and to attach to rabbit intestinal mucosa. Additionally, expression of the hemagglutinating ability by the bacterial cells was accompanied by efficient bacterial adherence to the intestinal mucosa. Finally, the hemagglutinating activity of Vm-OMPHA was markedly increased by incubation with Vm-HA/protease. These results indicate that all three HAs may have significant roles in the glycoprotein-mediated intestinal adherence of V. mimicus E-33.

Vacuolar-type H(+)-ATPase in mouse bladder epithelium is responsible for urinary acidification.

The urine in the mouse bladder was found to be acidic, ranging from pH 5.3 to 5.5 in the daytime and pH 6.0 to 6.3 at night. Administration of bafilomycin A1 or concanamycin A, specific inhibitors of vacuolar-type H(+)-ATPase, into bladder lumen caused neutralization of urinary pH at least for 36 h, whereas inhibitors of mitochondrial ATP synthase (F-type H(+)-ATPase) or P-type H(+)-ATPases did not. Bafilomycin A1-sensitive proton secretion from isolated inside-out bladder was also observed. Immuno-electron microscopy with antibodies against vacuolar H(+)-ATPase revealed that vacuolar-type H(+)-ATPase is rich in luminal plasma membrane and endosomes of superficial cells of the bladder epithelium. These results indicate that vacuolar-type H(+)-ATPases present in luminal plasma membrane of the superficial epithelial cells secrete protons so as to acidify the urine in mouse bladder.

Microvesicles isolated from bovine posterior pituitary accumulate norepinephrine.

Histochemical study indicated that the posterior pituitary possesses numerous microvesicles (MVs) containing synaptophysin, a marker protein specific for brain synaptic vesicles (Navone, F., Di Gioia, G., Jahn, R., Browning, M., Greengard, P., and De Camilli, P. (1989) J. Cell Biol. 109, 3425-2433). By monitoring cross-reactivity with anti-synaptophysin antibody, the MVs were highly purified from bovine posterior pituitaries by a combination of differential and sucrose density gradient centrifugations. The purified MVs had an average diameter of about 60 nm and were associated with synaptophysin as revealed by immunoelectron microscopy. The vesicles contained ATPase activity partially sensitive to bafilomycin A1 and to vanadate. The membrane fraction immunoisolated with anti-synaptophysin antibody also exhibited similar ATPase activity. The two ATPases could be purified separately; the vandate-sensitive enzyme was identified as a 115-kDa polypeptide immunochemically similar to chromaffin granule P-ATPase (forming phosphoenzyme intermediate), and the bafilomycin A1-sensitive ATPase showed essentially the same properties as those of vacuolar type H(+)-ATPases. Upon addition of ATP, the MVs formed an electrochemical gradient of protons and took up norepinephrine in a reserpine-sensitive manner, indicating the presence of secondary monoamine transporter coupled with vacuolar type H(+)-ATPase. No uptake of L-glutamate, gamma-aminobutyrate, glycine, or acetylcholine was observed. The identification of MVs as organelles responsible for storage of monoamines is important for understanding the physiological function of the posterior pituitary.

Utilization of hemin and hemoglobin as iron sources by Vibrio parahaemolyticus and identification of an iron-repressible hemin-binding protein.

Several clinical isolates of Vibrio parahaemolyticus were examined for their ability to utilize either hemin or hemoglobin as a sole source of iron. Both compounds appeared to be equally good iron sources. Maximum growth was obtained at 5 microM hemin or 1.25 microM hemoglobin under the conditions tested. Using a hemin-agarose batch affinity method, the hemin-binding protein was isolated from crude total membranes of a hemin-utilizing strain, WP1, grown under iron-deficient but not under iron-sufficient conditions. This protein was identical to the 83 kDa outer membrane protein which was expressed in response to iron limitation. The protein was susceptible to proteinase K cleavage in whole cells, indicating its exposure at the cell surface. Hemin and hemoglobin, but not protoporphyrin IX, inhibited binding of the protein to hemin-agarose.

A sandwich cup method for the penetration assay of antimicrobial agents through Pseudomonas exopolysaccharides.

We developed new sandwich cup method to assay the penetration of various antimicrobial agents through Pseudomonas exopolysaccharides. Using alginate extracted from mucoid-type Pseudomonas aeruginosa and gellan gum from Pseudomonas elodea, the role of exopolysaccharides as a barrier against drug penetration was examined. The penetration of positively charged hydrophilic drugs such as aminoglycosides and polypeptides was markedly inhibited by the gels tested, but that of beta-lactams, quinolones, and macrolides was not inhibited. The penetration of gentamicin was strongly influenced by the gel concentration, the solution to be used, and the presence of Ca2+. These results suggest that the microenvironment at the infection site could greatly influence drug penetration through biofilms in vivo.

Suppressive effect of liposomes containing DNA coding for diphtheria toxin A-chain on cells transformed with bovine leukemia virus.

A recombinant plasmid which contained a gene for diphtheria toxin A-chain (DT-A) under the control of the long terminal repeat (LTR) of bovine leukemia virus (BLV) (BLV-LTR) was constructed to test a novel application of liposomes as antiviral agents. The promoter activity of BLV-LTR was estimated by the chloramphenicol acetyltransferase (CAT) assay using a plasmid which contains the coding sequence of CAT under the control of BLV-LTR (pBLVCAT). When BLV-infected cells were transfected with pBLVCAT, CAT activity was detected. BLV-uninfected cell lines, however, showed no detectable CAT activity. The plasmid DNA entrapped in liposomes was added to BLV-infected cells in culture. Syncytium formation induced by BLV-infected cells was effectively suppressed by the liposomes containing the gene for DT-A under the control of BLV-LTR. Conversely, liposomes containing the gene for DT-A without a promoter showed no such effect. DT-A gene-containing liposomes with BLV-LTR did not affect formation of syncytium induced by bovine immunodeficiency virus. These observations indicate that BLV-infected cells were readily targeted on the level of gene expression. This strategy could be applied to the treatment of BLV-induced B-cell proliferation of cattle, and further to other viral/neoplastic diseases where specific gene expression is exerted.

Establishment of a mouse model of cystitis and roles of type 1 fimbriated Escherichia coli in its pathogenesis.

The role of type 1 fimbriae in promoting bladder colonization and the course of Escherichia coli cystitis were examined with type 1 fimbriated strains of clinically isolated E. coli. In the experiments of mice in vivo, intact bladder epithelium showed natural resistance to the adherence of type 1 fimbriated and non-fimbriated E. coli. However, the exfoliation of bladder superficial cells by trypsinization before the bacterial inoculation promoted the adhesion and colonization of type 1 fimbriated E. coli onto bladder epithelium. After colonization of E. coli, maximum numbers of E. coli and leukocytes were observed 3 days after inoculation. Nine days after inoculation, both of E. coli and leukocytes disappeared and the regeneration of superficial cells was observed. On the other hand, superficial cells in mice injected with phosphate-buffered saline or non-fimbriated E. coli regenerated 5 days after trypsinization. The present study demonstrated that the removal of superficial cells is essential for the adhesion and colonization of type 1 fimbriated E. coli onto bladder epithelium in vivo and a new model of E. coli cystitis in mice was established. The model which we established is valuable for histopathological, immunological, and therapeutic studies.

Overproduction of truncated subunit a of H+-ATPase causes growth inhibition of Escherichia coli.

Genes (uncB) for wild-type and mutant a subunits of Escherichia coli H+-ATPase (F0F1) were cloned into recombinant plasmids. The subunits were expressed under the control of a weak promoter of the unc operon at 30 degrees C and strong promoters of lambda phage at 42 degrees C. At 30 degrees C, the wild type and a truncated (Glu-269----end) a subunit complemented the defect of the a subunit mutant KF24A (Trp-111----end), whereas the other mutant subunits (Trp-111----end, Trp-231----end, Gln-252----end, and a subunit with a deletion of residues 21 to 227) did not. Three mutant subunits (Trp-231----end, Gln-252----end, and Glu-269----end) and the wild-type a subunit caused growth inhibition associated with cell elongation, an uneven distribution of membrane proteins, and an altered septum structure when they were expressed at 42 degrees C. These phenomena were not observed with the other mutant subunits, suggesting that overproduction of the middle region (between residues 111 and 230) of the a subunit causes growth inhibition.