SEB is cytotoxic and alters EC barrier function through protein tyrosine phosphorylation in vitro.

We studied whether Staphylococcal enterotoxin B (SEB) has direct effects on endothelial cells (EC) in the absence of effector cells or their products. Bovine or human pulmonary artery EC were grown to confluence on filters mounted in chemotaxis chambers. Barrier function was assessed by placing [14C]bovine serum albumin in the chamber and sampling the lower well for 14C activity. SEB exposures induced a significant (P < 0.001) dose- and time-dependent increase in albumin flux across both bovine and human EC monolayers. Albumin flux was temperature dependent, and cycloheximide pretreatment of the monolayers did not block the SEB-induced increase in permeability. Preincubation of SEB with trypsin or anti-SEB antibody significantly (P < 0.0001) reduced the effect, whereas pretreatment with polymyxin B did not. SEB at > or = 10 micrograms/ml significantly (P < 0.03) increased EC injury as measured by 51Cr release in a dose- and time-dependent manner. Herbimycin and genistein, inhibitors of protein tyrosine kinases, each protected against SEB-induced cytotoxicity, barrier dysfunction, and intercellular gap formation. We conclude that SEB perturbs endothelial barrier function and viability in the absence of effector cells or their mediators.

Mitogenic activities of amino acid substitution mutants of staphylococcal enterotoxin B in human and mouse lymphocyte cultures.

Site-directed mutagenesis has been used to introduce amino acid substitutions at specific residues of the staphylococcal enterotoxin B (SEB) gene cloned from Staphylococcus aureus 10-275. The mitogenic activities of these derivatives were determined in two assay systems: (i) mouse spleen cells and (ii) a mixture of human peripheral blood mononuclear cells and lymphocytes. Substitution of either His-12, His-32, His-121, His-166, Lys-152, or Gly-205 did not significantly alter the mitogenic activity from that of the wild-type toxin in either proliferation assay. Substitution of either residue Asn-23, Phe-44, or Cys-93 reduced the mitogenicity of SEB by a degree that depended upon the assay system used. Similar to the results reported by others measuring toxin activation of mouse lymphoid cells, we found that substitutions of these three residues of SEB caused at least 800-fold reductions of mitogenic activity from that of the wild-type toxin. When tested for toxicity in vivo in D-galactosamine-treated mice, the reduced activities of these mutant toxins, however, were not as pronounced. In contrast, when tested in the human cell mitogenicity assay, these mutant toxins were active. Small alterations in activity (two- to fivefold reduction) were observable only at low concentrations. Our findings reveal the importance of using human lymphocytes in addition to the traditional mouse spleen cell assay when assessing biological activities of staphylococcal enterotoxins.

The importance of a lipopolysaccharide-initiated, cytokine-mediated host defense mechanism in mice against extraintestinally invasive Escherichia coli.

Extraintestinally invasive Escherichia coli (EC) that possess both a complete LPS and K1 capsule evade both complement-mediated bacteriolysis and neutrophil-mediated killing. Since C3H/HeJ mice that are hyporesponsive to LPS were uniquely susceptible to lethal infection with EC of this phenotype, we speculated there was an LPS-initiated host defense mechanism against this pathogenic phenotype. The LPS-normoresponsive C3H/HeN as well as the C3H/HeJ mice cleared these EC from the circulation within 4 h of intravenous administration. Whereas electron micrographs of the liver demonstrated these EC undergoing degeneration within the phagolysosomes of of both macrophages and Kupffer cells of C3H/HeN mice, these EC replicated within these cells of the C3H/HeJ mice. Restoration of anti-EC activity of C3H/HeJ mice occurred with activation of Kupffer cells and peritoneal macrophages in vivo with BCG and in vitro with IFN-gamma, but not with LPS. Pretreatment of C3H/HeJ mice with a combination of recombinant murine IL-1 and TNF-alpha also restored the killing of K1(+)-EC but did not enhance the killing of a K1(-)-EC mutant. These data are consistent with the hypothesis that (a) there is no intrinsic inability of C3H/HeJ phagocytes to kill EC, but (b) an LPS-initiated, cytokine-mediated host defense mechanism is required for such killing. These studies emphasize the importance of bacterial surface characteristics in the interaction with specific host defenses.

Identification of staphylococcal enterotoxin B sequences important for induction of lymphocyte proliferation by using synthetic peptide fragments of the toxin.

A series of 13 synthetic peptides, approximately 30 amino acids each, which spanned the entire sequence of staphylococcal enterotoxin B (SEB) were tested to evaluate their effects on T-cell proliferation in a culture system containing elutriated human peripheral blood lymphocytes incubated with a specific ratio of mononuclear cells. Four peptide regions were found to inhibit SEB-induced proliferation; they included sequences 1 to 30 (previously thought to be involved in major histocompatibility complex class II binding), 61 to 92 (sequences which relate to the T-cell receptor site), 93 to 112 (a linear sequence corresponding to the cysteine loop), and 130 to 160 (containing a highly conserved sequence, KKKVTAQEL). Antisera raised to this last peptide were capable of neutralizing SEB-induced proliferation. Antisera raised against the peptides which overlapped this sequence also were somewhat inhibitory. Neutralizing antisera were not produced from any other peptide sequence tested. To determine if any of these effects were nonspecific with regard to SEB-induced proliferation, the peptides were tested for inhibition of phorbol dibutyryl ester-induced proliferation, and only the sequence 93 to 112 (corresponding to the cysteinyl loop region) was consistently inhibitory (40%). Of the regions which displayed inhibition of SEB-induced proliferation, the peptide 130 to 160 inhibited binding of 125I-SEB to lymphocytes. These data suggest that the residues containing and surrounding the sequence KKKVTAQEL may be critical in the SEB-induced proliferation and may be useful for developing neutralizing antisera to SEB.

Role of Shiga-like toxin I in bacterial enteritis: comparison between isogenic Escherichia coli strains induced in rabbits.

BACKGROUND/AIMS: Enteroadherent Escherichia coli that produce Shiga-like toxins are important causes of human disease, including enterohemorrhagic E. coli-induced colitis (EHEC). The role of Shiga-like toxins in these illnesses is unclear. The aim of this study was to establish an animal model for human EHEC and to determine the role of Shiga-like toxin I (SLT-I) in this model. METHODS: E. coli strain RDEC-1 is an enteroadherent rabbit diarrheal pathogen. An isogenic variant of RDEC-1 (termed RDEC-H19A) producing high levels of SLT-I was obtained by infecting RDEC-1 with an SLT-I-converting bacteriophage. The effects of in vivo enteric infection produced in rabbits by RDEC-H19A were compared with those in uninfected and RDEC-1-infected animals. RESULTS: SLT-I-producing RDEC-H19A induced a severe, noninvasive, enteroadherent infection in rabbits. Clinically, infection with RDEC-H19A was more severe than infection with RDEC-1 and caused more serious histological lesions including vascular changes, edema, and more severe inflammation. Interleukin 1 and platelet-activating factor appear to be important inflammatory mediators to this infection. CONCLUSIONS: The illness induced by RDEC-H19A in rabbits resembled enterohemorrhagic E. coli-induced colitis of humans. SLT-I is an important virulence factor in the pathogenesis of EHEC.

Method for simultaneous isolation and quantitation of platelet activating factor and multiple arachidonate metabolites from small samples: analysis of effects of Staphylococcus aureus enterotoxin B in mice.

A novel method has been developed for the extraction and simultaneous separation and quantitation of key arachidonate metabolites and platelet activating factor (PAF) from plasma samples of limited size. Aqueous solutions of these metabolites were added onto a solid phase C-18 cartridge and arachidonate metabolites and PAF were eluted, successively, with acetonitrile-methanol (85:15, v/v), followed by 100% methanol. Arachidonate metabolites (first eluate) were fractionated by C-18 reverse-phase high-performance liquid chromatography using a program designed for the resolution of 31 arachidonate metabolites (3-min separation). The fractions were collected and assayed by radioimmunoassay or radiography, if radioactively labeled. Two internal standards were added to each sample, 15-hydroxyeicosadienoic acid (detected at 235 nm) to determine gradient shifts and [1-14C]eicosatrienoic acid to estimate the recoveries of arachidonate metabolites at any stage of the process. This method was developed to answer specific questions concerning the mode of action of Staphylococcus aureus enterotoxin B (SEB). Plasma samples from mice challenged with SEB were analyzed and major differences were seen in 5-lipoxygenase metabolites. Low doses of SEB vs saline stimulated 5-hydroxyeicosatetraenoic acid production, while high doses of SEB stimulated leukotriene D4 production.

Investigation of a staphylococcal food poisoning outbreak in a centralized school lunch program.

The trend in many communities toward centralized school lunch preparation potentially increases the risk of foodborne illness. Foods often are prepared long before serving and may be distributed to satellite schools by persons with little formal training in safe techniques of food preparation or food service. In May 1990, an outbreak of staphylococcal food poisoning occurred in elementary schools in a Rhode Island community participating in such a program. In the investigation of the outbreak, students in schools that reported cases were interviewed. Food preparation, handling, and distribution were reviewed. At School E, 662 lunches were prepared and distributed to 4 additional schools (schools A-D). Schools A and B accounted for nearly all cases of the food poisoning, with rates of 47 percent and 18 percent. Eating ham increased the risk of illness (62 percent of those consuming ham and 3 percent of those who did not, relative risk = 18.0, 95 percent confidence interval = 4.0, 313.4). Large amounts of Staphylococcus aureus were cultured, and preformed enterotoxin A was identified in leftover ham. A food handler, who tested positive for the implicated enterotoxic strain S. aureus, reported having removed the casings from two of nine warm ham rolls 48 hours prior to service. Because of improper refrigeration, prolonged handling, and inadequate reheating, the ham was held at temperatures estimated at 10-49 degrees Celsius (50-120 degrees Fahrenheit) for a minimum of 15 hours. The potential for larger outbreaks prompted a statewide training program in safe food preparation for school lunch personnel, which may have applications for other communities.

The K1 capsule is the critical determinant in the development of Escherichia coli meningitis in the rat.

Although Escherichia coli strains possessing the K1 capsule are predominant among isolates from neonatal E. coli meningitis and most of these K1 isolates are associated with a limited number of 0 lipopolysaccharide (LPS) types, the basis of this association of K1 and certain 0 antigens with neonatal E. coli meningitis is not clear. The present study examined in experimental E. coli bacteremia and meningitis in newborn and adult rats whether or not the K1 capsule and/or O-LPS antigen are critical determinants in the development of meningitis. Rats received subcutaneously at K1 E. coli strain (018+K1+) or mutants lacking either the K1 capsule (018+K1-) or 0 side-chain (018-K1+). 12-24 h later, blood and cerebrospinal fluid (CSF) specimens were obtained for quantitative cultures. The isolation of E. coli from CSF was observed in both newborn and adult rats infected with K1+ strains regardless of LPS phenotype (018+ or 18-) who also developed a high degree of bacteremia (e.g., greater than 10(4) CFU/ml of blood). In contrast, none of the newborn and adult rats infected with 018+K1- and developing bacteremia of greater than 10(4) were found to have positive CSF cultures. These findings indicate that the presence of the K1 capsule and a high degree of bacteremia are key determinants in the development of E. coli meningitis, suggesting that there may be specific binding sites present in the brain which have an affinity for the K1 capsule and thus may be responsible for the entry of K1-encapsulated E. coli into the meninges.

Staphylococcus aureus enterotoxin B challenge of monkeys: correlation of plasma levels of arachidonic acid cascade products with occurrence of illness.

Arachidonic acid cascade products have been shown to be increased in vitro in Staphylococcus aureus enterotoxin B (SEB)-treated epithelial cell cultures in our laboratory. In order to confirm that these products were clinically related to SEB intoxication, monkeys were administered SEB by nasogastric intubation. It caused emesis in five of six monkeys (less than 4 h), and the sixth monkey showed signs of mild illness. The monkeys which vomited continued to display signs of gastrointestinal illness beyond 8 h but were without any apparent signs of illness by 24 h. Blood samples were collected prior to SEB administration, upon first indication of illness, and at twice that time interval. One week prior to SEB treatment, the same monkeys were administered saline by nasogastric intubation and in every way handled similarly in order to serve as their own controls. Blood samples were taken from the control animals at 0, 4, and 8 h. The plasma concentrations of prostaglandin E2 (PGE2), leukotriene B4 (LTB4), and 5-hydroxyeicosatetraenoic acid (5-HETE) did not vary significantly throughout the 8-h experiment for saline-treated controls, nor did they differ from the concentrations found in the plasma of monkeys just before administration of SEB. When the SEB-treated monkeys showed the first indication of illness (less than 4 h), the mean of the concentration in plasma of PGE2 increased 1.44-fold, that of LTB4 increased 2.23-fold, and that of 5-HETE was essentially unchanged. At twice the time interval of the first display of illness (less than 8 h), PGE2 was still elevated (1.48-fold), LTB4 had decreased slightly to 1.66-fold, and 5-HETE had soared (3,45-fold), suggesting a divergence in the enzymatic utilization of the parent compound of the latter two metabolites, 5-hydroperoxyeicosatetraenoic acid. These studies suggest that arachidonic acid cascade metabolites were a consequence of SEB intoxication and may provide a logical site for metabolic interference in SEB-induced toxicity.

Oligonucleotide probes for detection and differentiation of Staphylococcus aureus strains containing genes for enterotoxins A, B, and C and toxic shock syndrome toxin 1.

Different synthetic DNA nucleotide sequences were evaluated as gene probes for the specific detection and differentiation of Staphylococcus aureus strains encoding enterotoxins A (SEA), B (SEB), and C (SEC) and toxic shock syndrome toxin 1 (TSST-1). Identification of sequences unique to each toxin, based on knowledge of their nucleotide sequences, led to preparation of the specific 18-base oligonucleotide probes EA1 (encoding amino acids 177 to 182 of SEA), EB2 (encoding amino acids 105 to 110 of SEB), EC5 (encoding amino acids 125 to 131 of SEC1), and TS1 (encoding amino acids 160 to 166 of TSST-1). In colony blot hybridization analyses, these probes hybridized specifically with DNA from strains that produced the respective toxin serotypes. An excellent (greater than or equal to 93%) correlation between hybridization results (genotype) and toxin protein detection by an enzyme-linked immunosorbent assay (phenotype) was observed in the characterization of both reference and clinical strains of S. aureus for SEA, SEB, and TSST-1. A lower correlation (64%) for SEC reflected a lack of sensitivity in detecting toxin production. Our findings demonstrate that molecular DNA hybridization with synthetic oligonucleotide probes provides another approach for establishing the toxigenicity of S. aureus.

Aerobactin and alpha-hemolysin as virulence determinants in Escherichia coli isolated from human blood, urine, and stool.

Because iron acquisition is essential to the survival of invasive strains of Escherichia coli, the frequency of two potential iron acquisition systems, aerobactin and hemolysin production, were compared in E. coli isolated from human blood (n = 95), urine (n = 100), and stool (n = 50). By phenotypic and genotypic methods, the prevalence of hemolysin production was 22% in bacteremic, 38% in urinary, and 22% in fecal isolates of E. coli. Aerobactin production was detected in 76% of blood and in 73% of urinary isolates but in only 52% of fecal isolates (P less than .01). A reciprocal relationship was found in blood isolates between aerobactin and hemolysin; the majority of blood isolates (55%) that lacked aerobactin were hemolytic, whereas only 14% of blood isolates that expressed aerobactin were hemolytic (P less than .0001). Aerobactin may be the principal mechanism of iron acquisition in extraintestinal isolates of E. coli, and hemolysin may serve as an alternative mechanism in the absence of aerobactin genes.

Pretreatment with recombinant murine tumor necrosis factor alpha/cachectin and murine interleukin 1 alpha protects mice from lethal bacterial infection.

Tumor necrosis factor/cachectin (TNF/C) is the principal mediator of bacterial endotoxin-induced shock and death. We found that the C3H/HeJ mouse, which is less able to produce TNF/C in response to endotoxin, has a 1,000-fold greater susceptibility to lethal infection with Escherichia coli than the TNF-responsive congenic mouse, C3H/HeN. This surprising finding suggested that this lethal peptide may also be involved in host protection. To test this hypothesis we pretreated the C3H/HeJ mouse with a combination of recombinant murine TNF/C-alpha and IL-1 alpha. This combination protected these mice against an intraperitoneal bacterial challenge of greater than 20 LD50S (nearly 2 x 10(2) CFU) that grew to a level of greater than 10(7) CFU/ml of blood and per gram of liver in untreated mice. This suggests a significant role for these cytokines in host defenses against invasive infections that require bacterial replication within the host. These protective mechanisms may not be important for less virulent organisms. These findings may have important implications for the proposed use of anti-TNF/C agents in the treatment of septic shock.

Rapid purification of staphylococcal enterotoxin B by high-pressure liquid chromatography.

The Staphylococcus aureus enterotoxins represent a group of proteins that cause emesis and diarrhea in humans and other primates. We have developed a rapid two-step high-pressure liquid chromatography (HPLC) procedure for purification of staphylococcal enterotoxin B (SEB). Sterile filtrates (2.5 liters) of strain 10-275 were adsorbed directly onto a reversed-phase column (50 mm by 30 cm Delta Pak; 300 A [30 nm], 15 microns, C18). SEB was obtained by using a unique sequential gradient system. First, an aqueous ammonium acetate to acetonitrile gradient followed by an aqueous trifluoroacetic acid (TFA) wash was used to remove contaminants. A subsequent TFA to acetonitrile-TFA gradient eluted the bound SEB. Further purification was obtained by rechromatography on a cation-exchange column. From 35 to 45% of the SEB in starting filtrates was recovered. Analysis by immunoblotting of samples separated on sodium dodecyl sulfate-polyacrylamide gels indicated that HPLC-purified SEB exhibited immunological and biochemical properties similar to those of the SEB standard. Induction of an emetic response in rhesus monkeys showed that the HPLC-purified toxin also retained biological activity.

Deletion of the Shiga toxin gene in a chlorate-resistant derivative of Shigella dysenteriae type 1 that retains virulence.

We used a probe specific for detecting the structural-gene sequences of Shiga toxin to analyze the genetic nature of toxin synthesis in mutant derivatives of Shigella dysenteriae type 1. A chlorate-resistant (chl) mutant (725-78) of S. dysenteriae type 1 strain 3818T, which had retained virulence but had lost production of high levels of cytotoxic activity associated with Shiga toxin synthesis, contained a complete deletion of the Shiga toxin structural-gene sequences. These structural-gene sequences were also absent in a derivative of S. dysenteriae type 1 that contained a substitution of Escherichia coli DNA in the trp region of the chromosome. Isolates of Shigella flexneri and Shigella sonnei also did not react with the probe. The low-level cytotoxic activities associated with the mutant S. dysenteriae type 1 strains or with the virulent S. flexneri and S. sonnei strains are neutralizable with antiserum to Shiga toxin; however, these cytotoxic activities are not determined by the genes encoding classic Shiga toxin.

Survey of purported virulence factors of Escherichia coli isolated from blood, urine and stool.

One hundred randomly selected urinary and blood isolates and 50 stool isolates of Escherichia coli were analyzed for phenotypic characteristics which may contribute to their virulence potential. Bacteremic isolates were more likely to have K1 capsules and express mannose-sensitive hemagglutination compared to stool isolates. Blood-stream isolates more frequently contained complete 0 side-chains in their lipopolysaccharide layer and less frequently exhibited mannose-resistant hemagglutination when compared to urinary isolates. Total plasmid content, hemolysin, total colicin and colicin V production were not significantly increased in Escherichia coli from blood or urine when compared to those recovered from stool.

Role of lipopolysaccharide and capsule in the serum resistance of bacteremic strains of Escherichia coli.

To define the relative roles of capsule and lipopolysaccharide in the virulence of Escherichia coli obtained from blood, we compared the behavior of K1- and K5-encapsulated strains in serum bactericidal and rat virulence assays. Unencapsulated isogenic mutants selected from five parent strains of E. coli O12:K1, but not of O18:K1 or O7:K1 (all rough-specific phage insensitive), were lysed by normal human sera. In contrast, isogenic mutants from strains of serotypes O6:K5 and O18:K5 retained the serum resistance of the parent strains. There was a greater than 10(5) difference in LD50 in newborn rats between K1-positive and K1-negative pairs of E. coli serotypes O18 and O7 and a greater than 1 log difference between isogenic pairs of serotype O12; however, the K5 isogenic pairs had a similar LD50. Some non-O6 O serotypes, however, required the K5 capsule for serum resistance. We conclude that some O serotypes require encapsulation for optimal virulence but that other O serotypes may not.

Genetic studies of kanamycin resistance in Campylobacter jejuni.

Campylobacter jejuni 3H40 and 4B20 harbored 59-kilobase (kb) self-transmissible plasmids encoding resistance to kanamycin and tetracycline. Although the two antibiotic resistances were more frequently inherited together, some transconjugants and ethidium bromide segregants which were resistant to only one of these antibiotics were recovered. The kanamycin-susceptible, tetracycline-resistant segregants carried plasmids 4 kb smaller than the 59-kb plasmids of their parents, whereas the kanamycin-resistant, tetracycline-susceptible segregants contained no detectable plasmid DNA. Restriction endonuclease maps of deleted forms of the 59-kb plasmids revealed that deletions and rearrangements of 4-kb lengths of DNA were associated with loss of kanamycin resistance. Translocation of the kanamycin resistance determinant between plasmid and chromosomal DNA was demonstrated. Such phenomena have not been previously described in C. jejuni spp. and are consistent with the interpretation that the kanamycin resistance determinant is encoded by a translocatable element.

High frequency transduction by phage hybrids between coliphage phi 80 and Salmonella phage P22.

phi 80immP22dis, a hybrid between phi 80 and P22, carries all the late genes of phi 80 and most of the P22 early region including the immC and immI bipartite immunity loci. The presence of the immI region allows this hybrid to grow on lysogens of phi 80immP22 hybrids which have the immC locus, but not the immI locus. In addition to these P22 immunity regions, phi 80immP22dis contains the P22 att marker so that the prophage can be inserted into the chromosomal P22 attachment site adjacent to the proA-proB region of the host. Unlike its phi 80 parent which performs specialized transduction of the trp region, phi 80immP22dis transduces markers located adjacent to its attachment site to Escherichia coli K12 recipients at high frequencies (0.3% for argF and 0.18% for proA). Induction of phi 80immP22dis lysogens yields new hybrid phage clones which have incorporated E. coli K12 chromosomal segments in place of the P22 immI to att segment. Having lost the immI region, the new hybrids no longer grow in phi 80immP22 lysogens. These new hybrids, termed phi 80immP22dis-, possess specialized transducing properties, transferring the argF and proA markers at higher frequencies (21% for argF and 12% for proA) than previously obtained with the phi 80immP22dis phage.