Molecular cloning and characterization of the genes encoding the L1 and L2 components of hemolysin BL from Bacillus cereus.

Hemolysin BL, which is composed of a binding component, B, and two lytic components, L1 and L2, is the enterotoxin responsible for the diarrheal food poisoning syndrome caused by strains of Bacillus cereus. To further characterize the toxin, we sought to clone and sequence the genes encoding the L1 and L2 proteins. A genomic library was screened with polyclonal antibody to the L1 and L2 proteins to identify recombinant clones containing the genes. Five clones reacted with the antibody to L2, but none reacted with the antibody to L1. Southern hybridization analysis with oligonucleotide probes designed from the N-terminal amino acid sequences of the L1 and L2 proteins, in conjunction with immunoblot and nucleotide sequence analysis, revealed that the recombinant plasmid from one of the clones contained two genes, hblC and hblD, which encode L2 and L1, respectively. The two genes are arranged in tandem and are separated by only 37 bases. The gene which encodes the B component of hemolysin BL (hblA) is located immediately downstream from the gene encoding the L1 protein. Northern blot analysis of B. cereus RNA showed a 5.5-kb transcript which hybridized with DNA fragments internal to, or including a portion of, the coding sequences of the B, L1, and L2 genes, suggesting that the clustered genes which encode the components of hemolysin BL are cotranscribed and constitute an operon.

The effect of elastase-specific monoclonal and polyclonal antibodies on the virulence of Aspergillus fumigatus in immunocompromised mice.

Elastase has been implicated as a potential virulence factor involved in the invasion process of the opportunistic pathogen, Aspergillus fumigatus. Monoclonal and polyclonal antibodies, known to inhibit elastase in vitro, were employed in an immunocompromised mouse model of invasive aspergillosis to determine if the antibodies could protect mice from fatal infection. Individual monoclonal antibodies, known to inhibit elastase partially (13 to 23%), or combinations of monoclonal antibodies, known to inhibit elastase 70 to 100%, were tested in the mouse model. No individual nor combination of monoclonal antibodies protected immunosuppressed, infected mice in the doses tested. Similarly, elastase-specific polyclonal antibodies, raised in mice or rabbits, did not exhibit a protective effect, nor did immunization of mice with elastase prior to immunosuppression and infection. Histological examination of the lungs of immunosuppressed, infected mice showed no amelioration of fungal invasiveness by treatment with elastase-specific monoclonal or polyclonal antibodies. However, immunocompetent mice, instilled with a spore inoculum much higher than used in the preceding studies and treated with antibodies, survived, while control mice not treated with antibodies were overwhelmed by the massive spore dose and died. Nevertheless, overall evidence suggests that elastase may not be the primary virulence factor involved in invasive pulmonary aspergillosis.

Molecular cloning and characterization of the hblA gene encoding the B component of hemolysin BL from Bacillus cereus.

Previous evidence suggests that hemolysin BL, which consists of a binding component, B, and two lytic components, L1 and L2, is the enterotoxin responsible for the diarrheal form of gastroenteritis caused by food-borne strains of Bacillus cereus. To prove that hemolysin BL and the enterotoxin are the same requires large amounts of these components free of other B. cereus proteins. For this purpose, we sought to clone the gene encoding the B component and to express it in Escherichia coli. A partial genomic library was constructed and a 29-base, 1,152-fold-degenerate oligonucleotide probe, designed from the N-terminal amino acid sequence of the B component, was used to identify recombinant clones containing the gene. Detection of gene products reactive with a monoclonal antibody specific for the B component and analysis of the nucleotide sequence confirmed that isolated clones, reactive with the oligonucleotide probe, did contain the gene encoding the B component. The protein, expressed in E. coli, apparently from the B. cereus promoter, produces a ring-shaped zone of hemolysis when combined with purified L components from B. cereus, a reaction typical of hemolysin BL. Northern (RNA) blot analysis of B. cereus RNA showed a large (5.1-kb) transcript which hybridized with a 500-bp probe internal to the B-component-coding sequence, suggesting that the hblA gene encoding the B component may be transcribed as part of a polycistronic message, possibly including the structural genes for the two lytic components. Higher levels of expression and disruption of the hblA gene are being pursued to resolve whether hemolysin BL is indeed the enterotoxin.

Purification and properties of the elastase from Aspergillus fumigatus.

Elastase, a potential virulence factor from the opportunistic pathogen Aspergillus fumigatus, was purified 220-fold from culture broth by fast-performance liquid chromatography employing anion exchange (Q Sepharose fast flow), cation exchange (S Sepharose fast flow), and gel filtration (Superose 12). Purified to near homogeneity, the elastase had an apparent molecular mass of 32 kDa by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (silver stain) but a mass of about 19.1 kDa as determined by gel filtration on Superdex 75. The elastase is not glycosylated and is positively charged at neutral pH, having a pI of 8.75. Inhibition by 0.2 mM phenylmethylsulfonyl fluoride (100%) and 0.21 mM leupeptin (60%) implies that the elastase is a serine protease. However, the enzyme is also inhibited by 5 mM EDTA (100%) and 10 mM 1,10-orthophenanthroline (30%), suggesting a requirement for divalent cations. The enzyme acts optimally at pH 7.4 and 45 degrees C in 50 mM sodium borate buffer, but in Tris HCl, the pH optimum shifts to 8.8.

Inhibition of Aspergillus fumigatus elastase with monoclonal antibodies produced by using denatured elastase as an immunogen.

In preparing monoclonal antibodies to the elastase from Aspergillus fumigatus, we found that the enzyme was weakly immunogenic in BALB/c mice. Antiserum titers were only 1:1,000 to 1:5,000, and hybridomas secreted nonspecific immunoglobulin M (IgM). Denaturing the elastase in 0.5% sodium dodecyl sulfate at 80 degrees C for 10 min prior to injection increased titers of antiserum against the nondenatured (native) enzyme 10-fold. Of eight hybridomas selected following immunization with the denatured enzyme, seven produced IgG reactive with the native enzyme and one produced nonspecific IgM. The nondenatured immunogen tested again yielded mainly IgM producers. Immunoblots and enzyme-linked immunosorbent assay showed that the IgG monoclonal antibodies were reactive with both the denatured and nondenatured fungal elastases; none cross-reacted with human neutrophil elastase, porcine pancreatic elastase, or Pseudomonas elastase. Elastase-specific polyclonal antibody produced in mice inhibited elastase activity beginning at a molar ratio (antibody to elastase) of 4:1, and activity was completely inhibited at 14.5:1. Some individual monoclonal antibodies partially inhibited elastase, but certain pairs, at a molar ratio of each antibody to elastase of 5.4:1, acted synergistically to inhibit the activity completely.

Characterization of the components of hemolysin BL from Bacillus cereus.

Previously we described the partial purification of a novel hemolysin from Bacillus cereus and showed that hemolytic activity required the combined action of at least two components, called B and L to signify their cell-binding and cell-lytic roles in this activity. On further purification, as described in the present article, a combination of anion-exchange chromatography and polyacrylamide gel electrophoresis separated three proteins, B, L1, and L2 (35, 36, and 45 kDa, respectively). Individually, these proteins were inactive in hemolytic and vascular permeability assays, and combinations of B and L1 or B and L2 were either inactive or slightly active. Combinations of all three moieties produced the unique ring-shaped zone of hemolysis, a previously described characteristic of hemolysin BL, as well as edema and bluing in the vascular permeability assay. Since the vascular permeability assay is known to correlate with enterotoxicity, these results suggest that hemolysin BL is enterotoxigenic. Furthermore, the molecular weights and isoelectric point values of the hemolysin BL components are consistent with those described by others for the multicomponent diarrheal enterotoxin of B. cereus. Immunofluorescent staining of B-treated erythrocytes confirmed that B binds to cells as an initial step required before the L components can act to cause cell lysis.

Double-antibody sandwich enzyme-linked immunosorbent assay for cellobiohydrolase I.

A double-antibody sandwich enzyme-linked immunosorbent assay was developed for quantifying cellobiohydrolase I (CBH I) in crude preparations of the cellulase complex from Trichoderma reesei. The other enzymes (endoglucanase and beta-glucosidase) in this complex and other ingredients in culture broth did not interfere with this assay. The antibody configuration that resulted in the highest specificity for the assay of CBH I employed a monoclonal antibody to coat wells in polystyrene plates and peroxidase-labeled polyclonal antibody to detect cellobiohydrolase bound to the immobilized monoclonal antibody. Previously, procedures have not been available for the direct assay of CBH I activity in the presence of the other enzymes in the complex, and current indirect procedures are cumbersome and inaccurate. The direct procedure described here is highly specific for CBH I and useful for quantifying this enzyme in the range of 0.1 to 0.8 mug/ml.

A novel bicomponent hemolysin from Bacillus cereus.

A procedure combining isoelectric focusing (Sephadex IEF) and fast protein liquid chromatography (Superose 12; Mono-Q) removed hemolytic activity (presumably a contaminant) from partially purified preparations of the multicomponent diarrheal enterotoxin produced by Bacillus cereus. However, when the separated fractions were recombined, hemolytic activity was restored, suggesting that hemolysis is a property of the enterotoxin components. Combined fractions exhibited a unique ring pattern in gel diffusion assays in blood agar. During diffusion of the hemolysin from an agar well, the erythrocytes closest to the well were not lysed initially. After diffusion, hemolysis was observed as a sharp ring beginning several millimeters away from the edge of the well. With time the cells closer to the well were also lysed. This novel hemolysin consists of a protein (component B) which binds to or alters cells, allowing subsequent lysis by a second protein (component L). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis, isoelectric focusing, and Western blot analysis showed that hemolysin BL has properties similar to those described previously for the enterotoxin and that both components are distinct from cereolysin and cereolysin AB.

Effect of feeding rate on monoclonal antibody production in a modified perfusion-fed fermentor.

Previously we described a perfusion system for production of high yields of monoclonal antibodies in a fermentor. This system incorporated a cylindrically shaped, stainless steel filter mounted around the stirring shaft for retention of cells within a 1 liter fermentor. Modification of this filter by increasing the pore size from 5 micron to 10 micron decreased its tendency to clog and allowed continuous operation for about 3 weeks. Fresh culture medium, containing 6.5 mg glucose/ml and 3% horse serum, was supplied continually at two different perfusion rates, 850 and 1100 ml/day. Spent culture medium containing monoclonal antibody was harvested concomitantly. Highest cell density (5 X 10(7)/ml) and best antibody yield (1.7 g/l culture per day) were obtained at the higher feeding rate.

Factors affecting monoclonal antibody production in culture.

Factors that affect production of monoclonal antibodies (McAb) by a murine cell line were investigated. The goal was to estimate the efficiency of large scale production in stirred reactors. It was found that in batch cultures most McAb was produced after the log growth phase; final yield was 100-200 micrograms/ml. Yields of McAb were increased to 290 micrograms/ml by feeding cells glucose and glutamine. Lactic acid, which was produced in culture as a result of glucose metabolism, had no toxic effect on cells, while another waste product, ammonium ion, was probably accumulated at toxic levels during late stages of cell growth. The hybridoma cell line was propagated in four different systems: fed-batch, semi-continuous, two stage and perfusion. These systems were compared to batch cultures for their effect on cell viability and antibody production. Daily addition of fresh medium (fed-batch propagation) increased antibody productivity from 15 (batch culture) to 27 mg/l of culture/day. In the semi-continuous culture productivity was raised to 34 mg/l/day. Further increase in productivity to a level of 62 mg/l/day was achieved by applying a second batch stage to the semi-continuous culture. A perfusion culturing method was the most effective for production of McAb. Average concentrations of 2.2 X 10(7) live cells/ml and 390 micrograms of antibody/ml corresponding to productivity of 660 mg/l/day were achieved. Serum concentration in the medium was reduced to 0.125% resulting in a specific activity of 0.4 mg of McAb per mg of protein in the cell-free culture broth.

Factors affecting cell growth and monoclonal antibody production in stirred reactors.

Environmental and cultural factors that could affect growth and cell viability of mouse-mouse hybridoma cells in culture were investigated. The aim was to determine conditions that could prolong viability and result in increased yields of monoclonal antibodies in stirred reactors. Factors studied included temperature, level of dissolved oxygen, nutrient depletion, and waste product accumulation. Growing cells at temperatures 3-9 degrees lower than optimum (37 degrees C) increased viability but monoclonal antibody production was lowered. A low level of dissolved oxygen (25% air saturation compared to 60% for optimum growth) prolonged cell viability and increased the monoclonal antibody yield by about 50%. Feeding cultures daily to maintain the glucose level above 1 mg/ml and at the same time feeding cells glutamine (150 micrograms/10(6) cells per day) maintained the level of viable cells at 1.7 X 10(6)/ml for at least 9 days and resulted in an antibody yield of 290 micrograms/ml, about a 70% increase over cultures fed either glucose or glutamine alone. Ammonium ion, added to cell populations at levels produced in cultures, stopped cell growth and decreased antibody production. Another waste product, lactic acid, had no toxic effect when added to media at levels found in cultures. These results agree with our suggestion that monoclonal antibody production is enhanced by maintaining cell viability over a prolonged period and provide a base for investigating modes of hybridoma cell propagation in fermentors.

Kinetics of monoclonal antibody production in low serum growth medium.

Factors affecting growth and monoclonal antibody production in vitro by a mouse-mouse hybridoma cell line have been investigated in a series of studies. The goal was to maximize antibody yields and demonstrate that antibodies can be produced efficiently on a large-scale in fermentors. This initial report describes (i) development of a radial immunodiffusion assay for accurate determination of antibody levels in culture, (ii) a culture medium formulation that allowed for reduction in the amount of fetal bovine serum required for good cell growth, and (iii) the kinetics of cell growth and monoclonal antibody production in low-serum media. The radial immunodiffusion assay, employing rabbit anti-mouse IgG antibodies in the immobile phase and the monoclonal antibody (an IgG2a to Rhizobium japonicum cells) as the antigen in the mobile phase, was more reproducible and reliable for determining antibody levels in culture broth than was an indirect enzyme-linked immunosorbent assay. Addition of 0.25% Primatone RL and 0.01% Pluronic F-68 to Dulbecco's modified Eagle medium allowed cells to adapt to growth in medium containing as little as 1% fetal bovine serum; without these additives, 5% serum was the lowest level attained. For the kinetic studies, cells were grown in the low-protein medium in 3 liter spinner flasks. Antibody production occurred during the growth phase, however, significant amounts were also produced during later phases when the cells had stopped growing. Final titers were 100-200 micrograms/ml. It was concluded that maintenance of cell viability is more important than growth rate in production of antibody. This conclusion, confirmed in other studies, has developed into the major underlying strategy employed in subsequent investigations to maximize antibody production in stirred reactors.

Production of monoclonal antibodies in culture.

Factors that affected the production of monoclonal antibodies by a mouse-mouse hybridoma cell line, propagated in vitro in stirred vessels, were investigated. The purpose of the research was to estimate the efficiency of this system for large scale production of monoclonal antibodies. The antibody produced by these hybridoma cells was an IgG2a, specific for a surface antigen on Rhizobium japonicum NR-7 cells. Antibody content in the culture supernatant was determined by a radial-immunodiffusion assay using rabbit anti-mouse IgG antibodies in the immobile phase and mouse IgG (the monoclonal antibody) as the antigen in the mobile phase. This method was found to be more reproducible and reliable compared with an ELISA method. Cells were adapted to grow in an inexpensive, low protein content medium based on Dulbecco's Modified Eagle Medium (DMEM) supplemented with 0.25% Primatone RL, 0.01% Pluronic polyol F-68 and fetal bovine serum as low as 1%. Doubling time for the cells averaged 24 hrs, and final yields reached 2 X 10(6) cells per ml. The hybridoma cells were grown in the newly developed medium in 3 liter fermentors. Monoclonal antibody was produced during the early growth phase (3 days), however, most of the antibody was produced during a later growth phase (3-10 days) when 30 to 90% of the cells were dead. Final antibody yields were estimated to be 100-200 micrograms/ml. A low level of dissolved oxygen (25% air saturation) in the culture was found to increase the amount of antibody produced as compared with cells propagated at 60% air saturation (the optimal level for cell propagation) since the cells were kept alive for longer periods at the lower dissolved oxygen concentration. Using a fed-batch propagation method we were able to keep cells alive for long periods (up to 1 month) at a concentration of about 1 X 10(6) cells per ml, and thus to increase further monoclonal antibody production. Yields of 300-400 micrograms/ml were obtained.

Development of a monoclonal antibody capable of interacting with five serotypes of Staphylococcus aureus enterotoxin.

A monoclonal antibody capable of binding to determinants shared in common by staphylococcal enterotoxin serotypes A, B, C1, D, and E was developed. To accomplish this, BALB/c mice were immunized by alternating injections of serotypes A and D to enrich for spleen lymphocytes programmed to produce antibody to possible common determinants. These cells, fused with mutant myeloma cells (P3-X-63Ag8.653) produced hybrids that formed monoclonal antibodies to either serotype A only or to both serotypes A and D. A cloned hybridoma from the latter group produced an antibody (subclass immunoglobulin G1) which interacted with five serotypes. Highest affinity was to B and C1. An immunomatrix consisting of this antibody cross-linked to protein A-Sepharose CL-4B with dimethyl pimelimidate was capable of binding enterotoxin. Bound toxin was eluted with diethylamine. Because of its ability to interact with all five serotypes, the monoclonal antibody should prove useful in the development of a rapid method for screening foods for the presence of staphylococcal enterotoxin.

Correlation of elastase production by some strains of Aspergillus fumigatus with ability to cause pulmonary invasive aspergillosis in mice.

Seventy-five strains of Aspergillus fumigatus were screened for production of elastase in liquid and agar media containing elastin in yeast carbon base buffered with 0.05 M borate, pH 7.6. Of 71 strains which cleared elastin in agar plates, 33 produced elastase in liquid medium, as measured spectrophotometrically with elastin-Congo red. Six strains producing elastase and four nonproducers were tested for ability to cause invasive aspergillosis in immunocompromised mice (six mice per strain). All 36 mice exposed to elastase-producing strains died within 48 to 96 h. Lung tissue from dead mice showed hyphae and necrosis of the alveoli. Lungs of mice exposed to spores of strains not producing elastase showed few germinated spores and no destruction of alveoli. These results indicate that elastase may be significant in the invasion process.

Evaluation of glutaraldehyde and hydrogen peroxide for sanitizing packaging materials of medical devices in sterility testing.

External surfaces of packaging materials used for sterile medical devices may introduce contaminants into working areas used for sterility testing. Light wiping with tissues moistened with alkaline 2% glutaraldehyde (Cidex) or 3% hydrogen peroxide effectively reduced counts on 5 X 8 cm strips of packaging material (Tyvek) inoculated with 10(7) spores of Bacillus subtilis. The ability of antimicrobial agents to penetrate packaging material and kill contaminants on the medical device was tested by inoculating filter membranes with ca 100 cells of Pseudomonas aeruginosa or Staphylococcus aureus or ca 100 spores of Bacillus subtilis. A sterile square of test packaging material placed over the inoculated membrane (direct method) or 0.5 cm above the membrane (indirect method) was wiped with the antimicrobial agent. Except for polyethylene film (3 mil), all materials tested, including glassine and several types of coated and uncoated Tyvek, were penetrated by the agents, killing cells on the inoculated membranes. Death rates varied, depending on the organism, packaging material, and testing method. It is suggested that penetration tests be performed before using antimicrobial agents for sanitizing packaging materials during sterility tests.

Production and ecological significance of yeast cell wall-degrading enzymes from oerskovia.

Motile actinomycetes capable of degrading walls of viable yeast cells were isolated from soil and identified as Oerskovia xanthineolytica. A lytic assay based on susceptibility of enzyme-treated cells to osmotic shock was developed, and 10 of 15 strains of O. xanthineolytica, Oerskovia turbata, and nonmotile Oerskovia- like organisms from other collections were found to possess yeast lytic activities. All lytic strains produced laminaranase and alpha-mannanase, but the amounts, determined by reducing group assays, were not proportional to the observed lytic activities. The Oerskovia isolates demonstrated chemotactic, predatory activity against various yeast strains and killed yeasts in mixed cultures. Of 15 carbon sources tested for production of lytic enzyme, purified yeast cell walls elicited the highest activity. Glucose repressed enzyme production and caused cells to remain in the microfilamentous and motile rod stages of the Oerskovia cell cycle. Crude lytic activity was optimal at pH 5.6 to 7.0 and inactivated by heating for 6 min at 50 degrees C. Partial purification by isoelectric focusing showed that all lytic activity was associated with four beta-(1-->3)-glucanases. The absence of protein disulfide reductase, N-acetyl-beta-d-hexosaminidase, and phosphomannanase in crude preparations indicated that the principal enzyme responsible for yeast wall lysis was a beta-(1-->3)-glucanase that produced relatively little reducing sugar from yeast glucan.