Identification of Staphylococcal Hemolysins by an Electrophoretic Localization Technique

A technique for identifying and characterizing staphylococcal hemolysins by first separating them electrophoretically in barbital-buffered agar gel (pH 8.4) at 5 ma/cm for 2 hr and then determining their hemolytic activities by exposing them to human, horse, rabbit, and sheep erythrocytes is described. The alpha-hemolysin produced by a White variant of the Wood 46 strain of Staphylococcus aureus migrated 18 mm towards the cathode, and it lysed horse, rabbit, and sheep erythrocytes, whereas a Clear variant of the Wood 46 strain of S. aureus produced a lysin which migrated similarly to the alpha-hemolysin but lysed only rabbit cells. This latter lysin was tentatively named alpha(1)-lysin. This strain of S. aureus also produced beta-hemolysin which migrated 36 mm towards the cathode and lysed sheep cells. beta-Hemolysin produced by some strains of S. aureus showed considerable tailing during electrophoresis, whereas beta-hemolysin produced by other strains of S. aureus migrated as a well-defined peak. A lysin migrating 11 mm towards the anode was probably delta-lysin. It was, however, not produced in sufficient concentration when the cultures were grown in semisolid medium.     

Isolation and characterization of hemolytic genes from Actinobacillus actinomycetemcomitans

Abstract Periodontopathic Actinobacillus actinomycetemcomitans produces hemolysin and other leukotoxins. In the present study, two distinct clones which lysed horse erythrocytes were isolated by screening genomic DNA libraries of A. actinomycetemcomitans ATCC 43718 on blood agar plates. DNA hybridization analysis indicates that there were two distinct hemolytic genes present. Sonicated extracts from both Escherichia coli clones possessed hemolytic activities on horse, sheep and human erythrocytes, but not those of rabbit. Rabbit antiserum to A. actinomycetemcomitans ATCC 43718 whole cells inhibited the hemolytic activities of these clones.     

Hemolytic activity of Serratia marcescens

A cell-bound hemolytic activity was found in several strains of Serratia marcescens. One Serratia cell per ten erythrocytes was sufficient to cause complete lysis of human erythrocytes within 2 h in the liquid assay. The hemolytic activity resided in the membrane fraction and could be inactivated by incubating cells with proteases. The hemolytic activity was greatly enhanced in actively metabolizing Serratia cells and was partially controlled by the iron supply. Hemolysis was accompanied by degradation of erythrocyte membrane proteins (band 3 and 6, glycophorin) and was independent of the blood group. The exoprotease secreted by S. marcescens in large amounts was not involved in hemolysis. Comparison with various hemolytic strains of Escherichia coli showed that hemolysis of erythrocytes was more pronounced with S. marcescens than with E. coli. In contrast to hemolysis by E. coli, lysis of erythrocytes by S. marcescens was not enhanced by Ca²⁺ ions.Dedicated to Professor Dr. Gerhart Drews on the occasion of his 60th birthday     

Haemolysins of Salmonella, their role in pathogenesis and subtyping of Salmonella serovars

Haemolysin patterns of 175 strains of different Salmonella enterica subspecies enterica serovars isolated from different animal sources and places were determined using 11 different blood agar media made with either non-washed horse/sheep erythrocytes or with washed erythrocytes of cattle, sheep, horse, goat, rabbit, guinea pig, and human A, O and B blood groups. Study on 47 strains belonging to 10 serovars of Salmonella from buffalo meat (buffen), 42 strains of 11 serovars from goat meat (chevon): 16 strains of Salmonella enterica serovar Paratyphi B and 25 of S. enterica serovar Paratyphi B var Java from fish, meat, meat products and clinical cases; 45 isolates of S. Abortusequi from aborted mares (18), fetal contents (21), aborted donkey mares (2) and 4 reference strains, revealed that all host restricted Salmonella namely, S. enterica serovar Gallinarum, S. enterica serovar Anatum, S. enterica serovar Abortusequi and S. enterica serovar Paratyphi B could be divided into different haemolysin types based on their inability to produce haemolysis on one or more types of blood agar, while strains of all zoonotic Salmonella serovars induced haemolysis on all the 9 types of blood agar made of washed erythrocytes. None of 175 Salmonella could produce hemolytic colonies on blood agar made of non-washed horse/ sheep erythrocytes. Haemolysin type I (lysing all types of washed erythrocytes) was the commonest one among all serovars except S. Abortusequi, none of which lysed horse erythrocytes. Salmonella enterica serovar Abortusequi having hemolytic activity against sheep erythrocytes were more invasive but had lesser ability to survive in sheep mononuclear cells than non-hemolytic strains. Multiplicity of haemolysins appeared significant epidemiological tool.     

Binding of Extracellular Matrix Proteins by Enterococci

Forty-four enterococcal strains isolated from human clinical specimens were investigated for binding of ¹²⁵I-labeled fibronectin, vitronectin, thrombospondin, lactoferrin, and collagen type I and IV, and for cell surface hydrophobicity. Most strains expressed low binding of iodine-labeled human fibronectin, collagen I and IV, and higher binding of human vitronectin, human lactoferrin, and human thrombospondin. Bacteria grown in Todd-Hewitt broth exhibited increased binding to vitronectin and thrombospondin. In particle agglutination assays (PAA), Enterococcus faecalis strains reacted strongly with coated latex beads in contrast to E. faecium strains, which generally did not react. The ability of enterococci to bind ECM proteins was affected by heating and proteolytic digestion, suggesting that some protein-binding components become surface exposed after treatment with proteases. The binding of ¹²⁵I-labeled proteins to E. faecalis strain E70 was inhibited when cells were preincubated with unlabeled proteins. Preincubating cells with sulfated polymers such as dextran sulfate (M _r 5000 and 8000), pentosan sulfate and heparin decreased binding of vitronectin, lactoferrin, and thrombospondin. The binding of lactoferrin and thrombospondin was also decreased when bacteria were preincubated with galactose, fucose, and mannosamine, but not with mannose. All of 30 E. faecalis strains expressed pronounced surface hydrophobicity, but 10 of 14 E. faecium strains showed hydrophilic cell surface. Received: 22 April 1996 / Accepted: 29 June 1996     

A mitogenic agglutinin from the red alga Carpopeltis flabellata

An agglutinin has been isolated from the marine red alga Carpopeltis flabellata by ammonium sulphate fractionation, affinity chromatography on a yeast mannan-Sepharose 4B column and gel permeation HPLC. This new protein, designated carnin, is a monomeric glycoprotein with a M_r of 25 000, and it contains large amounts of Gly and Asx. It strongly agglutinated untreated rabbit, mouse and horse erythrocytes, and very weakly untreated human erythrocytes, whereas it did not agglutinate untreated sheep and chicken erythrocytes. Treatment of the erythrocytes with trypsin affected their sensitivity to haemagglutination by the agglutinin. The haemagglutinating activity was inhibited only by glycoproteins with N-glycosidic sugar chains. The activity was not affected by divalent cations. Carnin also showed mitogenic activity for T lymphocytes from mouse spleen. It inhibited the normal embryonic development of marine invertebrates.     

Hemagglutinating activity inCorynebacterium pyogenes

Thirty-eight strains ofCorynebacterium pyogenes isolated from cases of heifer- and dry-cow mastitis and from other infections of sheep, cows, pigs, and man were screened for agglutination of sheep erythrocytes. Bacteria grown either in serum broth or on blood agar in the presence of CO₂ hemagglutinated. Performance of titrations at 4°C avoided the hemolytic effects ofC. pyogenes. Erythrocytes of cat, chicken, cow, dog, guinea pig, horse, man (Group A), pig, and rabbit were also agglutinated. Pretreatment of sheep erythrocytes with trypsin, pepsin, A1 proteinase or pronase had no effect on agglutinability. Pretreatment ofC. pyogenes with pronase, but not with trypsin, A1 proteinase, or pepsin, abolished hemagglutinating capacity. The hemagglutinin was inactivated by exposure to 60°C for 10 min. Agglutination of sheep erythrocytes was inhibited by five glycoproteins. None of 12 mono-, di-, or trisaccharides nor heparin, chondroitin sulfate, or dextrin inhibited hemagglutination. These data suggest that the receptor may possibly be an oligohexosyl group of a glycoconjugate of lipid nature. Although a few cells of three mastitic strains ofC. pyogenes possessed fimbriae-like surface structures, no correlation between fimbriation and hemagglutinating activity was apparent.     

Characterization of Actinobacillus actinomycetemcomitans Hemolysin

Twenty out of 33 Actinobacillus actinomycetemcomitans strains formed hemolytic colonies on horse blood agar plates under anaerobic conditions. The hemolytic activity found in A. actinomycetemcomitans strain 137HE was examined. This activity was detected in the late exponential to early stationary phases of growth. Human erythrocytes were the most susceptible, followed by rabbit, sheep, horse and swine red blood cells. The majority of activity was detected in the cell-associated vesicle fraction. Zwitterionic detergent 3-[(3-cholamidopropyl)-dimethyl-ammonio]-1-propanesulfonate (CHAPS) extract from whole cells was semipurified by ammonium sulfate precipitation, preparative isoelectric focusing (IEF) and gel-filtration chromatography to yield a major band on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) with a molecular mass of 12 kDa. Heating at 80 C for 30 min and treatment with proteinase K or trypsin resulted in complete disappearance of the hemolytic activity. Sulphydryl reagents enhanced activity and small amounts of cholesterol inhibited it. In summary, we demonstrated the presence of hemolysin in A. actinomycetemcomitans, and examined and characterized it.     

Studies on the hemolytic and hydrolytic actions of phospholipases against mammalian erythrocyte membranes.

The hemolytic actions of three kinds of phospholipase C on horse and sheep erythrocytes were studied in relation to their hydrolytic activities on the phospholipid components of these red cells. Clostridium novyi (oedematiens) type A phospholipase C hemolyzed horse red cells by hydrolyzing phosphatidylcholine. However, the enzyme did not lyse sheep cells nor did it hydrolyze any phospholipid under the same conditions, although this enzyme hydrolyzed both sphingomyelin and phosphatidylethanolamine in the phospholipid mixture extracted from sheep red cells. Clostridium perfringens phospholipase C hemolyzed not only horse red cells by hydrolyzing phosphatidylcholine but also sheep red cells by hydrolyzing sphingomyelin. Sphingomyelin on sheep red cell membrane was hydrolyzed 10 times faster by this enzyme than that on horse red cell membrane. Pseudomonas aureofaciens phospholipase C hemolyzed horse red cells by attacking phosphatidylcholine and phosphatidylethanolamine. The enzyme did not attack sheep red cells but it did hydrolyze phosphatidylethanolamine in the extracted phospholipid mixture from sheep cells. The hemolytic activity of phospholipase C depends not only on the enzyme and the asymmetric distribution of phospholipids in the erythrocyte membrane but also on the accessibility of the enzymes to the phospholipids in the surface of the membranes. Hemolysis by phospholipase C belongs to a hot-cold type of lysis.     

Existence of a Novel Hemagglutinin Having No Protease Activity in Vibrio mimicus

The protease elaborated by Vibrio mimicus is known to possess hemagglutinating ability to chicken erythrocytes, the well-known HA/protease. A non-protease hemagglutinin (HA) with strong agglutinating ability towards rabbit erythrocytes was obtained from 32 hr culture supernatant of a pathogenic environmental strain of V. mimicus. This HA (V. mimicus HA: VMHA) appeared stable at relatively higher temperature and agglutinated the erythrocytes from rabbit, guinea pig and mouse but not the erythrocytes from chicken, bovine, horse and sheep. Simple sugars, metal ions and chelating agents failed to inhibit the activity of VMHA. The activity of VMHA was found to be sensitive to digestion by proteolytic enzymes including HA/protease. These results provide evidence for the existence of novel HA other than HA/protease in V. mimicus.     

Hemagglutination and hemolysis by : lichen extracts

Twenty-two species of lichens from 10 different genera possessed a hemagglutinin for one or more of human, sheep, horse, cow, rabbit, guinea pig, and chicken erythrocytes. Hemolysins were also detected occasionally, but these were only active at low dilutions. In those species tested, the hemolytic principle was dialyzable; the hemagglutinating agent was not. Preliminary studies have indicated that the lichen hemagglutinins are nonspecific.     

Evidence for the production of a novel proteinaceous hemolytic exotoxin by dinoflagellate Alexandrium taylori

We found that a whole cell suspension of Alexandrium taylori, which is toxic to Artemia, causes species-specific hemolysis against mammalian erythrocytes. Among the erythrocytes tested, rabbit and guinea-pig erythrocytes were highly sensitive, but human, sheep, and cattle erythrocytes were insensitive. The cell-free culture supernatant also showed potent hemolytic activity toward rabbit erythrocytes as seen in whole cell suspension. The hemolytic activity in the culture medium gradually increased with increase in cell number during exponential growth phase, and relatively high activity was maintained even after reaching the death phase. These results suggest that the hemolytic substance is actively released into the medium from A. taylori cells rather than simple leakage from ruptured or dead cells, and a part of them are steadily accumulated in the medium during the algal growth. Chemical characterization with ultrafiltration and trypsin-treatment suggested that the hemolytic substance released into the medium is protein-like compound with molecular weight more than 10,000 Da. The ammonium sulfate precipitated fraction obtained from the cell-free supernatant of A. taylori showed cytotoxic effect on HeLa cells as well as the hemolytic activity in a similar concentration range on a protein content basis. Our results suggest that A. taylori produces a novel proteinaceous hemolytic exotoxin.     

RNA and a cell wall component of Enterococcus faecalis IC-1 are required for phagocytosis and interleukin 12 production by the mouse macrophage cell line J774.1

Enterococcus faecalis is a resident lactic acid bacterium in the human intestine. Its immunostimulatory action was reported to be enhanced by heat sterilization. To investigate its beneficial actions, we evaluated the ability of 10 E. faecalis strains to induce interleukin-12 (IL-12) production in a mouse macrophage cell line, J774.1 and found that the strain, E. faecalis IC-1, had a potent IL-12-inducing ability. Furthermore, we investigated the underlying mechanism by treating IC-1 cells with RNase or lysozyme. Its activity almost disappeared and an antagonist of Toll-like receptor (TLR) 7 inhibited this activity. Moreover, lysozyme-treated IC-1 bacteria were not phagocytized by J774.1 cells, and did not induce IL-12 production. Based on our results, we propose that macrophages recognize the cell wall components of IC-1, leading to phagocytosis. The IC-1 RNA is then recognized by TLR7, which induces the production of IL-12.     

Use of monoclonal antibodies against horse immunoglobulin in an enzyme immunoassay of bacterial toxins and toxoids

Immunization of BALB/c mice by horse antiserum against diphtheria made it possible to obtain IgG₁ monoclonal antibodies (MoAbs) 2B7E4 specific for light chains of horse immunoglobulin (Ig). Unlike commercial preparations of anti-horse immunoglobulin antibodies, which are specific for the whole Ig molecule or its Fc-fragment, the peroxidase (HRP) conjugate of the MoAb, 2B7E4-HRP did not interact with human, mouse, rabbit, and sheep Igs, or horse albumin. The conjugate obtained was used with MoAbs against bacterial toxins and commercial horse antitoxins, as a universal reagent in sandwich enzyme immunoassay (ELISA) for bacterial toxins and toxoids. The detection sensitivity of diphtheria toxin/toxoid equaled 0.0005 Lf/ml; tetanus toxin and toxoid were detected with sensitivities of 20 LD₅₀/ml and 0.005 UI/ml, respectively. A similar sandwich ELISA for botulinum toxoids (group measurement) allowed types A, B, and E to be detected at 0.02, 0.002, and 0.001 UI/ml, respectively; selective measurement was only possible in the case of type E toxoid (0.001 UI/ml).     

The opportunistic pathogen Enterococcus faecalis resists phagosome acidification and autophagy to promote intracellular survival in macrophages

While many strains of Enterococcus faecalis have been reported to be capable of surviving within macrophages for extended periods, the exact mechanisms involved are largely unknown. In this study, we found that after phagocytosis by macrophages, enterococci‐containing vacuoles resist acidification, and E. faecalis is resistant to low pH. Ultrastructural examination of the enterococci‐containing vacuole by transmission electron microscopy revealed a single membrane envelope, with no evidence of the classical double‐membraned autophagosomes. Western blot analysis further confirmed that E. faecalis could trigger inhibition of the production of LC3‐II during infection. By employing cells transfected with RFP‐LC3 plasmid and infected with GFP‐labelled E. faecalis, we also observed that E. faecalis was not delivered into autophagosomes during macrophage infection. While these observations indicated no role for autophagy in elimination of intracellular E. faecalis, enhanced production of reactive oxygen species and nitric oxide were keys to this process. Stimulation of autophagy suppressed the intracellular survival of E. faecalis in macrophages in vitro and decreased the burden of E. faecalis in vivo. In summary, the results from this study offer new insights into the interaction of E. faecalis with host cells and may provide a new approach to treatment of enterococcal infections.     

Antibodies to Aeromonas salmonicida Lipopolysaccharide and cell wall antigens in rabbit and salmon immune sera

Atlantic salmon (Salmo salar L.) immunised with A-layer positive or A-layer negative strains ofAeromonas salmonicida did not produce antibodies reactive with proteinase K-digested LPS in the low molecular weight area corresponding to the core-region of LPS. The salmon produced antibody titres as high as those produced by rabbit when assayed against whole bacteria or LPS in ELISA. The salmon antibodies against the A-layer positive strain of A. salmonicida lysed rabbit erythrocytes sensitised with LPS from the A-layer positive strain of A. salmonicida. This was in contrast to the non-haemolytic activity of the salmon antibodies against the A-layer negative strain, indicating differences in epitopes between the two strains.     

Species differences in membrane susceptibility to lipid peroxidation

The susceptibility of liver microsomes to lipid peroxidation was evaluated in seven species: rat, rabbit, trout, mouse, pig, cow, and horse. Lipid peroxidation was measured as thiobarbituric acid reactive substances formed in the presence of either FeCl₃-ADP/ascorbate or FeCl₂/H₂O₂ initiating systems. For rat, rabbit, and trout microsomes, the order of susceptibility to peroxidation was rat > rabbit >> trout. The lack of peroxidation in trout microsomes could be explained by high microsomal vitamin E levels. Membrane fatty acid levels differed between species. Docosahexaenoic acid predominated in the trout, arachidonic acid in the rat, and linoleic acid in the rabbit. The contribution of individual fatty acids to lipid peroxidation reflected the degree of unsaturation with docosahexaenoic > arachidonic >>> linoleic. For all species except trout, the predicted susceptibility to peroxidation, based on the response of individual fatty acids, agreed well with directly measured microsomal peroxidation. With the exception of the trout, vitamin E content ranged from 0.083–0.311 nmol/mg microsomal protein between species, and low levels did not influence susceptibility to peroxidation. Trout microsomes peroxidized only after vitamin E depletion by prolonged incubation. The data indicate that below a vitamin E threshold, species differences in membrane susceptibility to peroxidation can be reasonably predicted based only on content of individual peroxidizable fatty acids.     

Purification, molecular cloning, and application of a novel sphingomyelin-binding protein (clamlysin) from the brackishwater clam, Corbicula japonica

A novel sphingomyelin-binding protein (clamlysin) was purified from the foot muscle of a brackishwater clam, Corbicula japonica. The purified 24.8-kDa protein lysed sheep, horse and rabbit erythrocytes and the hemolytic activity was inhibited by sphingomyelin, but not other phospholipids or glycosphingolipids. The open reading frame of the clamlysin gene encoded a putative 26.9-kDa protein (clamlysin B) which showed high sequence similarity with the actinoporin family. A surface plasmon resonance assay confirmed that clamlysin B specifically bound to sphingomyelin. Furthermore, two cDNA variants of clamlysin, encoding putative 31.4 kDa (clamlysin A) and 11 kDa (clamlysin C) proteins, were isolated. Only the 31.4-kDa variant was found to exhibit sphingomyelin-binding activity. Clamlysin A and B, but not C, shared a sequence (domain II) conserved in all known sphingomyelin-binding proteins. Domain II fused with a glutathione S-transferase bound to sphingomyelin. Horse erythrocytes, mouse melanoma B16 and GM95 cells, and Chinese hamster ovary CHO-K1 cells, but not the same cells treated with bacterial sphingomyelinase, were immunostained with clamlysin B. These results indicate that clamlysin B binds to the sphingomyelin of living cells and thus would be useful as a molecular probe to detect sphingomyelin.     

Francisella strains express hemolysins of distinct characteristics

Historically, Francisella strains have been described as nonhemolytic. In this study, we show by use of solid and liquid hemolysis assays that some Francisella strains have hemolytic properties. The Francisella novicida type strain U112 is hemolytic to horse erythrocytes and Francisella philomiragia type strain FSC144 is hemolytic towards both human and horse erythrocytes. The F. novicida strain U112 released a protein (novilysin A) into the culture supernatant which cross-reacted with antiserum against Escherichia coli HlyA whereas there was no similar protein detectable with this cross-reactive property from the supernatant of the F. philomiragia strain.     

Colonization factors andEscherichia coli belonging to enterotoxin-associated serotypes

Hemagglutination (HA) tests using human and bovine erythrocytes and microagglutination tests using pili-specific antisera (PSA) were performed to examine 168 strains ofEscherichia coli belonging to enterotoxin-associated serotypes for colonization factors (CFs). Seventy-one (42%) of these 168 strains possessed at CF, but only 10 (6%) were found positive by both HA and PSA tests. Groups of test strains from different sources (feces, urine, blood, and wounds) were not found to contain statistically different percentages of CF-positive strains. Strains producing heat-stable enterotoxin alone were less frequently associated with a CF than were other enterotoxigenic and nonenterotoxigenic strains. Strains showing heat-labile hemolytic activity and belonging to serotype O6: H—were less likely (P=0.014, Fisher's exact probability) to contain a CF than were similarly hemolytic strains belonging to other serotypes.