Haemophilus ducreyi is the causative agent of chancroid, a sexually transmitted genital ulcer disease shown to be involved in the heterosexual transmission of the human immunodeficiency virus. Heterogeneity of H. ducreyi lipooligosaccharide (LOS), complex sugars present in the bacterial membrane, was investigated to explore a potential basis for a classification system. A panel of clinical H. ducreyi isolates could be classified into 7 groups (1 to 7) according to the number and intensity of the LOS bands after electrophoresis, and grouped into 5 categories (A through E) according to the specific reaction and cross-reaction of antisera raised to outer membrane proteins (OMP). Carbohydrate and mass spectrometry analyses confirmed that all strains studied expressed many different LOS glycoforms at the cell surface, and that strains belonging to serologic categories B and D expressed truncated LOS molecules and lacked the characteristic DD-heptose found only in H. ducreyi strains. The vaccine potential of the native (rHgbA) and recombinant forms (rHgbA) of the hemoglobin receptor of H. ducreyi as well as the recombinant form of the outer membrane protein D15 (rD15) were evaluated in the temperature-dependent rabbit model of chancroid . Rabbits were immunized twice, 4 weeks apart, with 100 mug of the immunogens (rHgbA, rHgbA, rD15, PBS and rFetA) in Freund's adjuvant, and challenged 4 weeks after the booster immunization with strains 35000 (homologous) and V1157 (heterologous) in doses ranging from 10 3 to 103 CFU. rHgbA vaccination modified the course of a homologous challenge at an inoculation of 103 CFU: fewer rabbits developed any ulcers, which were culture positive for a shorter period of time than in controls. After heterologous infection with 104 CFU of broth-grown strain V1157, none of the lesions of nHgbA-vaccinated rabbits developed into ulcers and lesion aspirates were sterile, while 2/3 of controls had ulcers. The rHgbA vaccine also modified the course of an experimental homologous infection, although results were very inconsistent from experiment to experiment. Indeed, across all experiments only the duration of ulcers was significantly reduced in rHgbA-vaccinated animals compared to controls at a 104 CFU inoculum. After a heterologous challenge with 104 CFU of H. ducreyi strain V1157, no lesions ulcerated in rHgbA-vaccinated animals. Modest protection similar to the rHgbA vaccine was obtained when animals were immunized with rD15. In terms of protection, there appeared to be different performance indifferent experiments: although rD15 provided partial protection against a homologous challenge infection in some experiments, it failed to protect animals against disease in others. After a homologous challenge with 104 CFU of H. ducreyi strain 35000, only the duration of ulcers were reduced in rD15-vaccinated animals compared to controls. However, across all experiments, better protection in rD15-vaccinated animals was associated with a greater humoral response to vaccine. In conclusion, rHgbA seems to be the best of these vaccine candidates against chancroid, while more experiments are needed to determine the potential of rHgbA and rD15 vaccines.

We were interested in generating an in vivo retroviral gene therapy vector based on the commonly used Moloney murine leukemia virus (MoMLV). This was accomplished by pseudotyping with an engineered influenza A hemagglutinin. Point mutations were introduced to abrogate hemagglutinin's wild type binding and a single chain variable domain antibody fragment (scFv) was added to its amino terminus to provide new binding specificity. The engineered hemagglutinin was able to mediate binding of pseudotyped retrovirus to a scFv specific peptide but was unable mediate infection of target cells. To rescue the infectivity of the pseudotyped retrovirus the role of lipid rafts in the lifecycle of the MoMLV was examined. Lipid raft isolation from transfected cells and virus particles revealed that retroviral proteins were not associated with lipid rafts. Using a panel of hemagglutinin mutants with reduced lipid raft affinity we also determined that this parameter did not affect pseudotyping efficiency.

Our hypothesis proposed that Trichomonas vaginalis would induce apoptosis in host cells as part of its pathogenic assault. This hypothesis was tested in an established co-culture system; previous research demonstrated that the parasite destroyed McCoy cell monolayers. The TUNEL assay was used to identify cells undergoing apoptotic DNA fragmentation as a result of co-incubation with T. vaginalis. Early experiments suggested that apoptosis may have been occurring in the McCoy monolayers, but this was later disproved with the help of a sandwich ELISA to detect DNA/histone fragments that characteristically result from apoptosis. Instead, the cells that had originally been identified as apoptotic were found to be adherent trichomonads. While McCoy cell apoptosis in response to co-culture with T. vaginalis was not observed, T. vaginalis did, over time, cause detachment of almost all monolayer cells, and also caused irreparable damage in a portion of these cells. Induction of McCoy monolayer detachment was determined to be contact-independent, given that trichomonad culture supernatants had a similar effect. (Abstract shortened by UMI.)

This study was carried out to determine the frequency and levels of occurrence of legionellae in groundwaters from a variety of sources in U.S. and Canada. A limited number of water samples from cooling towers were also tested because municipally treated waters are considered to be the main sources of Legionella in artificial habitats such as cooling towers, air conditioners, whirlpools, hot tubs and plumbing systems (Krammer and Ford, 1994). Conventional procedures, including cultivation and polymerase chain reaction (PCR), were used to determine the presence of Legionella spp. in the water and biofilm samples during this study. A novel approach, using immuno-magnetic separation (IMS) in combination with cultivation or PCR, was also explored as an improved and rapid detection method for Legionella over conventional procedures. Because the IMS technique was continually being improved over the course of this study, all the samples received did not undergo the same procedure. (Abstract shortened by UMI.)

Bacterial isolates from bean-sprouts were screened for anti-Listeria monocytogenes bacteriocins using a well-diffusion method. Thirty-four isolates inhibited the growth of L. monocytogenes. Ten strains, which produced the biggest zones of inhibition, were characterized. One strain, isolate 80, which exhibited the strongest inhibition against L. monocytogenes, was selected for subsequent in-depth analysis. Both ribotyping and DNA sequencing of 16S ribosomal RNA demonstrated that the isolate was Lactococcus lactis subsp. lactis. Primary characterization showed that the anti-L. monocytogenes compound was proteinaceous and the antimicrobial effects were not caused by pH, a phage or $\rm H\sb2O\sb2,$ indicating that it was a bacteriocin. A plasmid cured derivative of isolate 32 was obtained. This plasmid-free strain still produced a bacteriocin, and had the same antibiotic resistance pattern as the wild type strain, indicating that the antibiotic resistance and bacteriocin genes were most likely to be chromosomally mediated. These bacteriocin-producing isolates produced nisin Z. Tests indicate that lactic acid starter cultures, possibly in combination with other hurdles, can be used to control the growth of L. monocytogenes in fresh-cut vegetables. (Abstract shortened by UMI.)

Concentrations of Cryptosporidium oocysts in source waters, especially surface sources, are of great concern to the drinking water supply industry. One of the objectives of this study was to assess if differences in water chemistry, biology or seasonal variations affect oocyst survival. Three watersheds, the Grand River, the St. Lawrence River, and the Carp River, all in Ontario, Canada, were selected to assess the survival of oocysts in waters with different geography. Water samples were collected from each river at two locations, one upstream from the urban district, the other at or downstream from the urban district. Water was sampled from each location within each watershed during different seasons of the year and seeded with $5\times10\sp5$ Cryptosporidium oocysts/mL. Synthetic hard water (100 ppm as CaCO$\sb3,$ pH 7.0) seeded with an equivalent number of oocysts served as the control for each experiment. Survival of seeded oocysts was determined by a standardized in vitro excystation assay combined with total oocyst counts; bacterial heterotrophic plate counts (HPC) at the start of each experiment were determined on R2A agar and counted after 5 days of incubation at room temperature. All water samples were sent to a commercial laboratory for chemical analysis. Water samples collected in the summer were incubated at 20$\sp\circ$C and 30$\sp\circ$C, winter water samples were incubated at 4$\sp\circ$C and 20$\sp\circ$C and spring water samples were also incubated at 4$\sp\circ$C and 20$\sp\circ$C. (Abstract shortened by UMI.)

This project outlines a novel method for testing the survival and inactivation of viral pathogens on viable human tissue carriers. The two topically applied chemical agents selected for testing in the ex vivo model were chlorhexidine digluconate (CD) and benzalkonium chloride (BC). The ex vivo method was developed to use animate carriers: pieces of human skin excised during routine plastic surgery for the stratum corneum and umbilical cords as mucous membranes. With both specimen types, the tissue was mounted on a cutting board and 2.3 cm diameter patches were then made with a cutting tool. Each patch was mounted over the top of a specially designed holder and secured with an O-ring. The animate carriers were then ready for use in the ex vivo protocol. The ex vivo model was evaluated with Human Herpesvirus 2 (HHV 2) strain 333 and Adenovirus 4 (AD 4) strain RI-67 with parallel tests run on the tissues and stainless steel disks to compare virus survival and inactivation on animate and inanimate surfaces. HHV 2 and AD 4 virus pools were made by infecting confluent Vero and 293 cell monolayers, respectively. Throughout the study infectivity for both viruses was measured by plaque assay in Vero cell monolayers. (Abstract shortened by UMI.)

Dientamoeba fragilis is a protozoan parasite first described in 1918. It resides in the human large intestine. D. fragilis has been associated with diarrhoea and abdominal pain in some patients. An indirect fluorescent antibody (IFA) assay was used to examine the presence of D. fragilis trophozoites in preserved faecal specimens. Antiserum to D. fragilis trophozoites was raised in a rabbit with a dixenic culture from the American Type Culture Collection (ATCC 30948). After absorption with Klebsiella pneumoniae and Bacteroides vulgatus, the immune rabbit serum was used for IFA examination. A total of 155 clinical samples were tested: 42 with no parasites, 9 with D. fragilis and 104 with other parasites. The IFA assay detected D. fragilis in 7/9 known positive samples. There were no false-positive IFA readings. Monoclonal antibodies (Mabs) against the parasite was produced in mice and characterized. A MAb (8B2) was obtained against an antigen released when D. fragilis was frozen and thawed. MAb 8B2 reacted with cultured D. fragilis in indirect immunofluorescence but did not react with clinical D. fragilis strains and it did not immunoblot. Another MAb (1H11) was generated against a sodium acetate-acetic acid-formalin fixed antigen. It reacted with D. fragilis isolates in patient specimens and identified a D. fragilis 39kDa protein. Susceptibility testing was performed on D. fragilis ATCC 30948. D. fragilis was cocultured with the bacteria in ATCC medium 1171. The activities of antiparasitic drugs were assessed by counting viable D. fragilis trophozoites with a haemacytometer by trypan blue exclusion. The minimal amoebicidal concentrations of the following four drugs were determined: iodoquinol at 128 $\mu$g/ml, paromomycin at 16 $\mu$g/ml, tetracycline (questionably) at 32 $\mu$g/ml, and metronidazole at 32 $\mu$g/ml. In this seroprevalence study sera from 3 symptomatic patients (F 10 y, M 12 y and F16 y), 12 age and sex-matched controls, and 189 randomly-selected healthy individuals (age range 6 months to 19 years) were tested by an indirect immunofluorescence (IIF) assay for antibodies. All 3 symptomatic patients infected with D. fragilis gave positive IIF titres of 80, and all 12 matched controls were also positive (titre ranges from 20 to 160, geometric mean titer (GMT) was 48). Of the 189 healthy children, 172 (91.0%) were positive at a serum dilution of 1 in 10, or higher. The specificity of the IIF assay was confirmed by immunoblotting 20 representative serum samples against D. fragilis (3 negative and 17 positive sera). All 17 IIF-positive serum samples identified the 39 kDa band, suggesting that this may be an important immunodominant protein of D. fragilis. (Abstract shortened by UMI.)

Haemophilus ducreyi is the etiologic agent of chancroid, a classic sexually transmitted genital ulcer disease. In endemic areas, HIV and chancroid interact in a complex, bidirectional epidemiology and biologic manner so as to amplify the prevalence of each other. Chancroid outbreaks can be controlled with the help of efficacious antibiotic treatment. Single doses of ceftriaxone, azithromycin and fleroxacin have been used successfully. However, because of HIV associated immune disease, an unacceptable failure rate of over 20% has been observed with concurrent HIV infections, and this, in the absence of H. ducreyi drug resistance. In this study, two-drug combinations of ceftriaxone, streptomycin, azithromycin and rifabutin were assayed in vitro by the checkerboard agar dilution technique for their activity against H. ducreyi strains proven susceptible by limiting agar dilution minimum inhibitory concentration determination. We conclude that a supra-additive effect observed between ceftriaxone and streptomycin in vitro by the checkerboard technique was corroborated by a synergistic effect observed between the same antimicrobial agents in the temperature-dependent rabbit model of infection. This combination may be evaluated as a single-dose in the treatment of chancroid in all patients, particularly those with HIV coinfection.

Hepatitis B Virus surface (HBsAg) and core antigens (HBcAg), and a variety of derivatives, were expressed from recombinant baculoviruses in insect cells cultured in monolayer. It was initially established that HBsAg was secreted from insect cells at low efficiency. HBsAg self-assembles into 22 nm spherical lipoprotein particles. Full-length and (four) truncated HBsAg genes were fused in-frame to segments encoding portions of the rabies virus glycoprotein (rabies G). Three different rabies G segments were utilized in various combinations with HBsAg. The products were analyzed for their ability to form chimeric 22 nm particles for potential vaccine use. No fusion proteins were secreted, and all were found to be insoluble in nonionic detergent. Expression of some fusion proteins in mammalian COS cells did not alleviate the inhibition of secretion. To further investigate the assembly and secretion of HBsAg in insect cells, a variety of mutants and fusions were designed to test the importance of the first hydrophobic transmembrane (TM) domain (domain I) of HBsAg. Domain I appeared to increase the rate of membrane insertion at the endoplasmic reticulum (E.R.). Also, a negatively-charged residue immediately preceding domain I was important to particle secretion. It was found that domain I could be replaced by a heterologous TM domain provided the heterologous domain was proceeded by a negatively-charged residue. HBcAg self-assembles into 27 nm protein particles. HBcAg is released from infected insect cells by an unknown mechanism. The normally secretory HBV e antigen (HBeAg), expressed from the core plus precore gene, was not released. We found that processing of the precore signal peptide (SP) was defective in insect cells. This SP could be functionally replaced by the SP from honeybee prepromellitin. Proteolytic cleavage appeared to be occurring subsequent to mellitin SP cleavage, but its nature is unknown. Phosphorylation of HBcAg was normal, while phosphorylation of mellitin SP fusion proteins suggested restriction to fully-cleaved species. A segment of the precore domain which normally remains after SP cleavage during HBeAg synthesis was important to phosphorylation and cleavage patterns. Electron microscopy revealed no apparent mechanism for the release of HBcAg which was present as spherical particles in the cytoplasm. Mellitin SP fusion proteins were observed as spherical particles in distended rough E.R. A gene encoding three peptides derived from HIV-1 env and gag proteins was fused in-frame to both the N- and C-termini of HBcAg. Chimeric 27 nm particles were formed from both fusions, and were demonstrated to be cross-reactive to both HIV-1 p17 and gp120 proteins. Morphological and structural features suggested that the HIV-1 peptide was interior to the particles for the C-terminal fusion and exterior to the particle for the N-terminal fusion.