The molecular structure and transferability of Tn1546 in 143 vancomycin-resistant Enterococcus faecium (VREF) isolates obtained from patients (n = 49), surface water (n = 28), and urban and hospital sewage (n = 66) in Tehran, Iran, were investigated. Molecular characterization of Tn1546 elements in vanA VREF was performed using a combination of restriction fragment length polymorphism analysis and DNA sequencing of the internal PCR fragments of vanA transposons. Long-PCR amplification showed that the molecular size of Tn1546 elements varied from 10.8 to 12.8 kb. The molecular analysis of Tn1546 showed that 45 isolates (31.5%) harbored a deletion/mutation upstream from nucleotide 170. No horizontal transfer of Tn1546 was observed following filter-mating conjugation with these isolates. Nevertheless, the rates of transferability for other isolates were 10−5 to 10−6 per donor. Insertion sequences IS1216V and IS1542 were present in 103 (72%) and 138 (96.5%) of the isolates, respectively. The molecular analysis of Tn1546 elements resulted in three genomic organizations. The genomic organization lineage 1 was dominated by the isolates from clinical samples (3.4%), lineage 2 was dominated mostly by sewage isolates (24.5%), and lineage 3 contained isolates obtained from all sources (72.1%). The genetic diversity determined using pulsed-field gel electrophoresis (PFGE) revealed a single E. faecium clone, designated 44, which was common to the samples obtained from clinical specimens and hospital and municipal sewage. Furthermore, the results suggest that lineage 3 Tn1546 was highly disseminated among our enterococcal isolates in different PFGE patterns.

The prevalence of diarrhoea in calves was investigated in 8 dairy farms in Mozambique at 4 occasions during 2 consecutive years. A total of 1241 calves up to 6 months of age were reared in the farms, and 63 (5%) of them had signs of diarrhoea. Two farms had an overall higher prevalence (13% and 21%) of diarrhoea. Faecal samples were collected from all diarrhoeal calves (n = 63) and from 330 healthy calves and analysed for Salmonella species, Campylobacter jejuni and enterotoxigenic Escherichia coli (ETEC). Salmonella spp. was isolated in only 2% of all calves. Campylobacter was isolated in 11% of all calves, irrespective of health condition, and was more frequent (25%) in one of the 2 diarrhoeal farms (p = 0.001). 80% of the isolates were identified as C. jejuni. No ETEC strains were detected among the 55 tested strains from diarrhoeal calves, but 22/55 (40%) strains from diarrhoeal calves and 14/88 (16%) strains from healthy calves carried the K99 adhesin (p = 0.001). 6,757 E. coli isolates were typed with a biochemical fingerprinting method (the PhenePlate™) giving the same E. coli diversity in healthy and diarrhoeal calves. Thus it was concluded: i) the overall prevalence of diarrhoea was low, but 2 farms had a higher prevalence that could be due to an outbreak situation, ii) Salmonella did not seem to be associated with diarrhoea, iii) Campylobacter jejuni was common at one of the 2 diarrhoeal farms and iv) ETEC strains were not found, but K99 antigen was more prevalent in E. coli strains from diarrhoeal calves than from healthy, as well as more prevalent in one diarrhoeal farm.

Clinical observations suggest that immune mechanisms affect etiology and course of recurrent cystitis. A primate infection model was used to show that primary bladder infection with a uropathogenic P-fimbriated strain (binding to globoside present in the bladder wall) protects against rechallenge with homologous as well as heterologous Escherichia coli strains for up to 5-6 mo. In contrast, mutant derivatives producing P-fimbriae either lacking the tip adhesin protein or carrying an adhesin for which no bladder receptor was present, were unable to induce protection, even though they generated bladder infections of similar duration as the wild type. Therefore, the protective effect mediated by the adhesin seemed to depend upon the presence of its cognate receptor. Since the wild strain also mediated protection against mutants that lacked the adhesin, our data suggest that the globoside-binding PapG adhesin acts as an adjuvant during infection to enhance a specific response against other bacterial antigens. In fact, the globoside-binding strain DS17, but not the mutant DS17-1, unable to bind to membrane-bound globoside, elicited a secretory IgA response to LPS in urine. These in vivo findings suggest that bacterial adhesin-ligand interactions may have signaling functions of importance for the immune response.

Aeromonas isolates from patients with diarrhea in Bangladesh (n = 69), from healthy controls (n = 11), and from surface water (n = 40) were analyzed with respect to their hybridization groups (HGs) by the aid of fatty acid methyl ester (FAME) characterization and DNA fingerprinting by AFLP, biochemical phenotypes (Phe-nePlate [PhP] types), and the production of hemolysin and cytotoxin. The aim of the investigation was to find out whether certain strains carrying virulence factors predominated among patient isolates. According to FAME and/or AFLP analysis, most human isolates were allocated to DNA HGs 4 (Aeromonas caviae) and 1 (A. hydrophila). Most environmental strains were allocated to HG8 (A. veronii biogroup sobria) and HG4 (A. caviae), and only one was of HG1. According to PhP typing, the diversity among patient isolates was lower than that among other strains, and two dominating PhP types (types BD-1 and BD-2) were identified in 29 and 30% of the patient isolates, respectively. PhP type BD-1 was also common among the environmental isolates, whereas PhP type BD-2 was only identified in two of the other isolates. Twenty-five of 26 isolates belonging to HG1 were of the same PhP type (BD-2), whereas isolates of other common HGs were more diverse according to their PhP types. Hemolytic and cytotoxin-producing strains occurred more frequently among the environmental isolates than among patient isolates. However, the hemolytic and cytotoxic activities among human isolates was strongly correlated to the HG1/BD-2 type, which, in addition, showed high cytotoxin titers (median values, 1/512 compared to 1/128 for cytotoxin-positive isolates belonging to other types). Thus, the HG1/BD-2 type may represent a pathogenic A. hydrophila type that is able to produce diarrhea in humans.

A nested PCR, using a 239-bp sequence in the pertussis toxin promoter region, was developed and evaluated. The assay differentiates Bordetella pertussis, Bordetella parapertussis, and Bordetella bronchiseptica by restriction enzyme analysis of the amplified fragments. The diagnostic performance of the PCR was evaluated in a Swedish pertussis vaccine efficacy trial which took place from 1992 to 1995, including study children and household members and using culture and serology for laboratory confirmation of suspected cases. In total 2,421 nasopharyngeal aspirates were analyzed. The total diagnostic sensitivity for B. pertussis was 90.2% (194 of 215). During the study period samples were processed with and without the cation-exchange resin Chelex. The PCR diagnostic sensitivity for B. pertussis among the Chelex-treated aspirates was 94.9% (75 of 79), and that for B. pertussis among 124 aspirates in a consecutive non-Chelex-treated material was 89.5% (111 of 124). After Chelex treatment of the 13 PCR-negative samples, an additional six became PCR positive, giving a final sensitivity of 94.3%. In addition, PCR was positive for B. pertussis with 57 of 1,744 samples negative by culture but with available serological data. The specificity of PCR with these samples was supported by a significant increase in antibody levels between acute and convalescent sera in 45 cases and by epidemiological or clinical data in all but two of the remaining cases. PCR was also positive for B. pertussis with 26 of 415 aspirates from episodes lacking serology. The diagnostic sensitivity of PCR for B. parapertussis was 74.0% (37 of 50). There were an additional seven culture-negative B. parapertussis PCR findings, six from cases with significant antibody increases against filamentous hemagglutinin only and one from a case lacking serology. There were no samples positive for B. bronchiseptica. In conclusion, PCR detection of B. pertussis and/or B. parapertussis enabled us to identify 90 positive nasopharyngeal aspirates, in addition to the 262 culture-positive samples (an increase of 34%). Relating these cases to serology and clinical data indicated a PCR specificity approaching 100%.

A highly discriminatory and standardized biochemical fingerprinting method was used to monitor the persistence and colonization of intestinal Escherichia coli isolated from the feces of four sows and their litters (four piglets from each) during the suckling, postweaning, and fattening periods. Altogether, 195 fecal samples were collected and 1,827 E. coli strains were tested (mean number of isolates tested per fecal sample per pig, 9.5). Strains were divided into similarity groups on the basis of their biochemical phenotypes (BPTs). The diversity of E. coli strains in each sample was measured with Simpson's index of diversity, and similarity between E. coli floras of piglets was calculated with a population similarity index. Each fecal sample contained several BPTs of E. coli, some of which dominated that population. The intestinal colonization of piglets consisted of successive waves of different E. coli BPTs, the tenure of which varied from a few days to 2 weeks. Most of these BPTs disappeared in the succeeding samples and were not recovered again from the same piglets. On the other hand, some E. coli strains which colonized piglets early during the suckling period persisted for a long period and were referred to as resident BPTs. Each piglet carried more than one resident BPT (mean of 2.4 BPTs per pig), some of which were also found in other piglets.(ABSTRACT TRUNCATED AT 250 WORDS)

Coliform bacteria are the most frequently reported bacteria to translocate after hemorrhage. We investigated the correlation between composition and diversity of the cecal coliform flora and the degree of translocation in a rat model of hemorrhagic stress. Two groups of nine rats each were bled to 60 and 50 mm Hg mean arterial blood pressure, respectively. A sham-operated group without bleeding (n = 9) and a noninstrumented group (n = 6) served as controls. From each rat, 40 coliform isolates from the cecum and up to 16 from positive mesenteric lymph node (MLN) cultures were tested with an automated biochemical fingerprinting method. The phenotypic diversity of coliforms in each cecal sample was calculated as Simpson's diversity index (DI), and similarities between bacterial types in different samples were calculated as population similarity coefficients. Three rats in the sham-operated group and seven in each of the bled groups showed bacterial translocation. Of the different biochemical phenotypes (BPTs) found in the cecum of bled rats (mean, 6.5 BPTs), only a few were detected in MLNs (mean, 1.9 BPTs per MLN), with Escherichia coli being the dominant species. The translocating E. coli strains were mainly of two BPTs. Rats showing no translocation either did not carry these strains or had a high diversity of coliforms in the cecum. Furthermore, translocation of these coliform types was independent of their proportion in the cecum. In bled rats, the diversity of coliforms (mean DI, 0.53) was significantly higher than that in control groups (mean DI, 0.30; P = 0.004), suggesting that hemorrhage stimulates an increase in diversity of cecal coliforms. Rats with similar coliform flora and subjected to the same treatment showed similar patterns of translocation. Our results suggest that the composition of the coliform flora is an important factor in translocation and that certain coliform strains have the ability to translocate and survive in MLNs more easily than others.

The Phene Plate system for typing Salmonella serotypes (PhP-S) is a simple automated typing method based on biochemical fingerprinting. It gives a quantitative value of the metabolism of various substrates by measuring the speed and intensity of each reaction. The 'biochemical fingerprint' of each isolate is used to calculate similarities among the tested strains with a personal computer program. We used this system to examine a collection of 86 strains of Salmonella enteritidis isolated from human sporadic cases in Germany between 1980 and 1992. Twenty-three biochemical phenotypes (BPTs) consisting of 9 common (C) and 14 single (S) BPTs were identified. BPTs C2 and C4 containing 20 and 36 strains respectively accounted for 65% of the isolates. Strains of BPT C2 were found over a wide period of time whereas strains of BPT C4 were isolated during the period between 1988 and 1992. With phage typing, 11 discrete phage types (PTs) and 18 strains designated as non-specific type (NST) were identified. PTs 4 and 8 with 39 and 17 strains respectively were the dominant PTs. Strains of PT 8 were isolated over a wide period of time whereas all (except one) strains of PT 4 were isolated between 1988 and 1992. Combination of biochemical fingerprinting and phage typing divided the strains into 25 phenotypes (BPT:PTs). Whilst phenotype C2:8 was found over a number of different years, phenotype C4:4 was isolated only between 1988 and 1992. These findings indicate the presence of one persistent and one recently emerged phenotype among S. enteritidis strains in Germany.(ABSTRACT TRUNCATED AT 250 WORDS)

The gene encoding alpha-toxin from Staphylococcus aureus was cloned into a Bacillus subtilis expression vector (pEF 231/alpha-Tox). The protease-deficient B. subtilis strain DB 104 transformed with pEF 231/alpha-Tox expressed and secreted 5 mg of alpha-toxin per liter into the growth medium. The alpha-toxin-containing supernatant was diluted 200-fold and used as coating antigen in an enzyme-linked immunosorbent assay (ELISA) for serodiagnosis of septicemia and endocarditis caused by S. aureus. Paired sera from patients in acute and convalescent stages of S. aureus and non-S. aureus infections were used to evaluate this ELISA. To evaluate the effectiveness of the crude preparation, the results were compared with those of an ELISA based on a commercially available alpha-toxin. Similar rises in serum titers were obtained with either type of alpha-toxin preparation. This is the first time a crude supernatant without any further purification has been used as an ELISA coating antigen. We therefore conclude that B. subtilis is a suitable host organism for cheap and simple production of prokaryotic recombinant antigens to be used in serodiagnosis.

Images

The faecal Escherichia coli flora was studied in 89 infants. Each infant was followed with a mean of 12 faecal samples (range 5-21) between 0 and 18 months of age. All isolates were assayed for P fimbriae and biochemically phenotyped and the persistence of each strain (phenotype) in the infant's gut was determined. In a subset of strains the occurrence of type 1 fimbriae and adherence to HeLa cells was studied. Thirty-one per cent of isolates belonging to strains colonizing for longer than 6 months expressed P fimbriae compared to 19% of the isolates from strains colonizing 1-6 months or transient strains colonizing less than 1 month. Type 1 fimbriae and adherence to HeLa cells occurred similarly often in all groups of strains. We conclude that P fimbriae, but not type 1 fimbriae or HeLa cell adherence seemed to contribute to the ability of the E. coli strain to colonize the human intestine.

Reproducible and discriminating typing methods are required for epidemiological investigations. Numerical typing systems analyse patterns obtained in various ways by calculating similarity coefficients between isolates. In the present study, various measures of the efficiency of a numerical typing system are quantified. These include reproducibility, accuracy, and discrimination power. Three different numerical typing methods for Escherichia coli were compared using these measures: (a) Biotyping with API 50 CH system, (b) Biochemical fingerprinting with the API 50 CH system and (c) Biochemical fingerprinting with the PhP-EC system. Biotyping qualitatively measures the results of a set of biochemical reactions as + or -. Biochemical fingerprinting also uses biochemical reactions, but the tests are scored quantitatively by measuring the kinetics and intensity of each reaction. It was found that biotyping yielded poor reproducibility. When biochemical fingerprinting analysis was used with the API 50 CHE system the reproducibility and the discrimination was good. The PhP-EC system for biochemical fingerprinting showed equal reproducibility but was superior to the API 50 CH system with regard to discrimination power.

A simple, automated microplate system for biochemical characterization of water isolates can be used to obtain fingerprints of the bacterial flora from various water samples. Mathematical models for calculating the diversities and similarities between bacterial populations are described for such fingerprints. The diversity may give information on whether an indigenous or allochthonous flora is present, and the similarities between bacterial populations, as calculated by using a population similarity coefficient (Sp), may indicate contaminations between different water samples. The system was demonstrated on coliform bacterial populations from various water samples, with or without suspected intercontamination. For unrelated water samples, the Sps were close to 0, whereas repeated samples of the same source showed Sps of 0.64 to 0.74. The Sp values from several water samples were also clustered to form a dendrogram, thus indicating the relative similarities between the bacterial populations to confirm suspected common sources of pollution.

A persistent vaginal colonization with a pyelonephritogenic strain of Escherichia coli, induced by administration of amoxicillin, was established in four adult cynomolgus monkeys. This colonization mimicked the one seen in urinary tract infection-prone human females. Attempts to eliminate the E. coli colonization and restore normal conditions were made. Either suspensions of lactobacilli or vaginal fluid from a healthy unmanipulated monkey was administered as repeated vaginal flushes for 5 to 9 days. A total elimination of vaginal E. coli was observed in two of six experiments with lactobacilli, and a decrease was observed in the other four. A better result was obtained with flushes of vaginal fluid, which eliminated the E. coli colonization in eight of eight experiments. In two of these, a single flush was sufficient to obtain a decolonization. The ability of fresh vaginal fluid to eliminate E. coli from the vagina could be transferred from one monkey to another. This study demonstrates the role of the normal flora in the defense against genital colonization with potentially pathogenic adhering E. coli. The possible clinical relevance of these findings must be further examined.

A prospective study of faecal colonization with P-fimbriated Escherichia coli between 0 and 18 months of age was conducted in 751 healthy infants. The influence of breast-feeding and treatment with antibiotics on this colonization was studied. Colonization with P-fimbriated E. coli increased with age from 10% at 6 days to 30% at 18 months of age (P less than 0.01). Breast-feeding influenced colonization at 6 weeks of age when breast-fed children harboured fewer bacterial species (P less than 0.001) and fewer P-fimbriated E. coli (P = 0.06) than bottle-fed infants. Treatment with antibiotics increased the colonization rate with P-fimbriated E. coli at the age of 11 months (P less than 0.05). However, this was not true for treatment with ampicillin, which increased colonization rate with Gram-negative species other than E. coli (P less than 0.05). Fifty per cent (378) of all children were colonized and a quarter (183) had pure cultures of P-fimbriated E. coli in at least one faecal sample. The clinical importance of this colonization remains to be shown.

Parenteral vaccination of sows against Escherichia coli diarrhea in their newborn piglets has become more common during the last decade in Sweden, and the vaccination has generally had positive effects. For more than 20 years we have investigated E. coli strains isolated from piglets and weaned pigs with enteric disorders, noting the presence of O groups, enterotoxins, and adhesins. There has been a continuous change in the frequency of these virulence factors. The present study was performed during 1983 and 1984 to follow this change, since such information is essential for the proper choice of vaccines. A total of 856 E. coli strains were obtained from 683 herds divided into three age groups: 1 to 6 days old, 1 to 6 weeks old, and weaned pigs. O group 149 still dominated in the last two age groups, while O group 101 was, for the first time, the most frequent O group in neonatal piglets. All but four O149 strains carried the K88 antigen, which was found in only one other strain (O group 8). K99 antigen was most often found in O groups 101 and 64, and among all the K99 strains ST mouse was the most common (44 of 57), followed by ST mouse-ST pig strains (12 of 57). The 987P antigen was demonstrated in 26 strains belonging to O groups 141 and OX46 and nontypable strains. Two strains belonging to O group 101 were positive for F41 antigen; one of them also carried the K99 antigen. Among all non-O149 strains, ST mouse was the most common type of enterotoxigenic E. coli ( n = 88), followed in decreasing order by ST mouse-ST pig strains ( n = 69) and ST pig strains ( n = 33). In 114 strains producing enterotoxins no adhesive factor was found. Thus, vaccination of the Swedish sow population for more than 5 years with vaccines containing O149 and K88 antigens has apparently changed the pattern of enterotoxigenic E. coli in neonatal diarrhea. The frequency of O149:K88 strains has been reduced, and O101:K99:ST mouse strains now dominate. However, O149 strains remain the dominant O group in piglets older than 1 week. In spite of all our diagnostic efforts, no virulence factors were detected in about one third of the piglets and weaned pigs with enteric disorders.

Serum antibodies to purified pneumolysin were determined by enzyme-linked immunosorbent assay (ELISA) in paired samples from 406 adult patients with community-acquired pneumonia and in samples from 184 healthy controls. A high sensitivity (83%) was obtained in patients with blood culture-confirmed pneumococcal pneumonia. In patients with a tentative pneumococcal diagnosis based on culture of samples from the sputum or the nasopharynx, 45% were positive by ELISA. The difference likely reflected the different relevance of cultural findings for the diagnosis of pneumococcal pneumonia. A significant rise in ELISA titer was found in 17% of the patients. When the diagnosis was also based on high titers, 25% were positive. Pneumococcal pneumonia diagnosed by the pneumolysin ELISA was significantly more common in the patients with a more severe disease and who required hospitalization (21 versus 5% for outpatients). Younger patients were more often positive for pneumococci as determined by high titers, while older patients showed titer rises. Mixed infections with other infectious agents were not uncommon. The finding of low titers in acute-phase samples from positive patients and in the youngest and oldest age groups of healthy controls were unexpected, indicating that further studies on the role of pneumolysin in pneumococcal disease are warranted.

An enzyme-linked immunosorbent assay (ELISA) with a highly purified pneumolysin as the antigen was evaluated for serological diagnosis of pneumococcal pneumonia. One hundred four healthy controls were tested, and the specificity of the test was set to 95%. In samples from patients with bacteremic pneumococcal pneumonia, 82% (18 of 22) were positive, i.e., at least one serum sample had a titer above the upper normal limit or at least a twofold rise in antibody titers was noted. In nonbacteremic pneumococcal pneumonia, 45% (21 of 47) of samples were positive. All sera were negative for patients with pneumonia caused by Haemophilus influenzae, Legionella pneumophila, Chlamydia psittaci, and influenza A virus. However, in patients with a diagnosis of Mycoplasma pneumoniae infection, 8 of 25 (32%) samples were positive for antibodies to pneumolysin. All sera, including those from patients with mycoplasma infection, were negative to a protein control antigen by ELISA. Serum immunoglobulin G response to pneumolysin as measured by ELISA might thus be an aid in the laboratory diagnosis of pneumococcal pneumonia. This assay may also help to further elucidate the occurrence of dual infections with pneumococci.

Pneumolysin was found to be produced by 112 of 113 clinical isolates of Streptococcus pneumoniae and to be an intracellular hemolysin. A 10-liter-scale fermentor production and purification procedure was developed for this hemolysin. The culture was concentrated by filtration 10 times before centrifugation. The cellular content was purified by ion-exchange chromatography, covalent thiopropyl gel chromatography, and gel filtration. One batch operation resulted in 6 mg of highly purified pneumolysin, with a yield of 66% and a specific activity of 1,400,000 hemolytic units per mg. The pneumolysin had a molecular weight of 53,000 and an isoelectric point of 5.2. The purification method developed will be of value in future studies on this hemolysin.

Images

A highly purified teichoic acid preparation was used in an enzyme-linked immunosorbent assay to measure the specific immunoglobulin G (IgG) and IgM response in staphylococcal disease. Antibody determination in a normal population, showing a difference of up to 20-fold in the mean IgG titers between the youngest children and adults, was used to establish age-correlated upper normal values. IgM antibodies were found to be of little diagnostic value since their response was often low or absent. Increased IgG titers were found in 24 of 27 (89%) patients with endocarditis, in 11 of 14 (79%) with complicated septicemia, and in 10 of 20 (50%) with uncomplicated septicemia with serum samples drawn between days 7 and 30 of disease. With paired samples, the numbers of patients with increased IgG titers were 17 of 17, 3 of 4, and 6 of 7, respectively, in the same patient groups. Increased IgG titers were less often demonstrated in patients with chronic osteomyelitis (7 of 22). The enzyme-linked immunosorbent assay for teichoic acid antibodies was found to be a sensitive and specific method for diagnosing staphylococcal endocarditis and septicemia. For optimal results, both the substantial age-correlated variation in normal titers and the importance of adequately spaced samples should be considered.

To evaluate the effect of widely used parenteral vaccination of dams against neonatal colibacillosis, the virulence factors of the intestinal Escherichia coli flora, namely, O serogroup, enterotoxin(s) produced (heat labile, porcine heat stable, and murine heat stable) and adhesins (K88, K99, and 987P antigens) of 149 piglets from different herds in Sweden were investigated. Three categories were investigated: healthy piglets, diarrheal piglets born to unvaccinated dams, and diarrheal piglets born to dams vaccinated with a polyvalent Formalin-killed whole-cell vaccine containing K88 antigen (Porcovac; Hoechst Pharmaceuticals, Hounslow, England). Piglets less than 1 week old and those 1 to 8 weeks old were evaluated separately. Diarrheal piglets less than 1 week old from vaccinated dams yielded a higher incidence of K99 antigen-positive E. coli of the murine heat-stable enterotoxigenicity type compared with piglets of the same age group from unvaccinated dams. The percentage of diarrheal cases from which E. coli lacking recognized virulence attributes were isolated was also higher in the former compared with the latter group. In the 1- to 8-week-old diarrheal piglets of vaccinated dams, the overall incidence, enterotoxigenicity type, and serotype of the E. coli isolates resembled those of diarrheal piglets less than 1 week of age from unvaccinated herds. Enterotoxigenic E. coli bearing 987P antigen detectable in vitro was rare. Most of the enterotoxigenic isolates lacking K88, K99, and 987P antigens produced only ST. The investigation pinpoints some of the inadequacies of vaccines of the type studied under field conditions.

Human and porcine enterotoxigenic strains of Escherichia coli were cultivated in tryptone-yeast extract medium or brain heart infusion broth and tested for production of heat-labile enterotoxin by the GM1 ganglioside enzyme-linked immunosorbent assay (GM1-ELISA) and the Y1 adrenal cell assay. When testing for enterotoxigenicity by the GM1-ELISA technique, homologous antisera for human and porcine heat-labile enterotoxins had to be used to detect enterotoxigenic strains of human and porcine origin, respectively. This observation indicates a serological difference between the heat-labile enterotoxins produced by human and porcine strains. Furthermore, brain heart infusion broth was found to have an inhibitory effect on detection of enterotoxin both in the GM1-ELISA and in a toxin-binding modification of the Y1 adrenal cell test, but not in the conventional adrenal cell assay.

We investigated membrane damage to human diploid, embryonic lung fibroblasts caused by highly purified alveolysin by measuring leakage of cytoplasmic markers and uptake of various metabolites, and we observed a leakage pattern typical of SH-activated cytolysins. However, the mode of membrane interaction resembled the mode of membrane interaction of theta-toxin from Clostridium perfringens rather than that of streptolysin O in the following respects: the activity on fibroblast membranes was high compared with the activity on sheep erythrocytes; the toxin did not bind irreversibly to fibroblast cytoplasmic membranes; considerable membrane damage was caused at 0 degrees C; and inhibition of amino acid uptake occurred in the absence of significant structural membrane damage. These findings imply that data on membrane effects caused by streptolysin O cannot be related indiscriminately to other SH-activated cytolysins. With regard to the mode of membrane interaction, two apparently different groups of SH-activated cytolysins exist.

Purified staphylococcal alpha-, gamma-, and delta-toxin were shown to cause activation of lymphoid cells from adult human donors and of cord cells in vitro as measured by [14C]thymidine incorporation after 7 days of incubation. T cell-enriched and T cell-depleted lymphocyte suspensions were activated in a similar fashion. Beta-toxin, on the other hand, exerted no valid stimulation of the various lymphocyte preparations. The lymphocyte-activating properties of alpha- and gamma-toxin were shown to be independent of their hemolytic capacity. The results probably reflect unspecific mitogen effects, but a component of specific reactivity cannot be excluded. We suggest that the unspecific triggering of lymphocytes in vitro is caused by surface-active properties of the toxins.

The membrane-damaging properties on human diploid embryonic lung fibroblasts of streptolysin O (from Streptococcus pyogenes) and theta-toxin (from Clostridium perfringens) were compared. The results are consistent with the suggested mechanism for hemolysis by streptolysin O involving one fixation site and one lytic site of this cytolysin. However, the membrane-damaging activity of the two toxins differed with respect to (i) relative cytolytic activity on human diploid lung fibroblasts compared with that on sheep erythrocytes, (ii) binding to the fibroblast membrane, (iii) activity at 0 degrees C, (iv) membrane repair after more than 30 min, and (v) effect on influx of amino acids. It is concluded that the mechanism of membrane damage caused by theta-toxin differs from that of cytoplasmic membrane. These results question the current concept that all thiol-activated, cholesterol-inactivated bacterial toxins are similar both structurally and functionally.