Castanospermine (1,6,7,8-tetrahydroxyoctahydroindolizine) is a plant alkaloid that inhibits alpha- and beta-glucosidase in fibroblast extracts [Saul, R., Chambers, J. P., Molyneux, R. J., & Elbein, A. D. (1983) Arch. Biochem. Biophys. 221, 593-597]. In the present study, castanospermine also proved to be a potent inhibitor of glycoprotein processing by virtue of the fact that it inhibits glucosidase I. Thus, when influenza virus was raised in the presence of castanospermine, at 10 micrograms/mL or higher, 80-90% of the viral glycopeptides were susceptible to the action of endoglucosaminidase H, whereas in the normal virus 70% of the glycopeptides are resistant to this enzyme. The major oligosaccharide released by endoglucosaminidase H from castanospermine-grown virus migrated like a hexose10GlcNac on a calibrated Bio-Gel P-4 column. This oligosaccharide was characterized as a Glc 3 Man 7 GlcNAc on the basis of various enzymatic treatments, as well as by methylation analysis of the [2-3H]-mannose-labeled or [6-3H]galactose-labeled oligosaccharide. The presence of three glucose residues in the oligosaccharide was also confirmed by periodate oxidation studies of the [6-3H]galactose-labeled hexose10GlcNAc. Castanospermine did not inhibit the incorporation of [3H]leucine or [14C]alanine into protein in MDCK cells at levels as high as 50 micrograms/mL. In addition, influenza virus produced in the presence of this alkaloid were fully infective and apparently produced in similar amounts to that of control cells, as determined by plaque counts. Castanospermine did, however, cause considerable changes in cell surface properties, since MDCK cells grown in 10 micrograms/mL castanospermine were able to bind twice as much [3H]concanavalin A as were control cells.

MDCK cells (a line of stable canine kidney cells) infected with influenza A/NWS/33 virus (a neurotropic variant of the Wilson Smith strain) were tested with 18 selected bacterial species to determine whether mammalian cells become susceptible to bacterial adherence as a result of virus infection. Cell monolayers were washed and examined microscopically for adherence. Bacteria of only two of 18 species were seen to adhere to the infected cells: a group B Streptococcus and Streptococcus sanguis. Control monolayers were negative for adherence. Pretreatment of virus-infected cultures with mouse ascitic fluid containing antibody to influenza A virus completely blocked adherence of the bacteria. Further testing with the strains representative of the five serotypes of group B Streptococcus disclosed that adherence occurred with types Ia, Ic, and II, but not with types Ib and III.

Madin-Darby canine kidney cells infected with influenza A virus (strains PR8, FM1, Jap 305, and Tex 1) were tested with nine strains of Staphylococcus aureus and group B Streptococcus type Ic to determine whether mammalian cells become susceptible to bacterial adherence as a result of virus infection. Bacterial adherence to virus-infected cells varied depending on the virus strain and on the strain of bacteria tested. A quantitative radioassay was developed to study the parameters of adherence. Attachment of 3H-labeled S. aureus grown in chemically defined or biologically complex medium to FM1 virus-infected cells was significantly increased (P < 0.0005) compared with attachment to control cells. Adherence coincided with the appearance of hemadsorption, which is a marker of the presence of virus-induced glycoproteins on the cell surface. Adherence was temperature dependent, increased with a decrease in hydrogen ion concentration, and was not affected by the presence of K+, Mg2", or Ca2". Adherence was blocked when 3H-labeled S. aureus was pretreated with trypsin but not when cells were pretreated with neuraminidase.

A ferret model was used to study bacterial adherence in animals with influenza. Ferrets were inoculated intranasally with influenza A3/Hong Kong/ 1 /68 virus. Antiviral serum antibodies were apparent by Day 5. On Days 3, 5, 7, 9, and 1 I , three virus-inoculated and two uninoculated controls were anesthetized, exsanguinated, and decapitated, and the lower jaw was removed. Each animal was inoculated intranasally with a 1-ml suspension containing 20 mg (dry wt) of either 3H-labeled Staphylococcus aureus or 3H-labeled group B Streptococcus type Ia and incubated for 45 min at ambient temperature. In animals challenged with staphylococci, 80% of the original inoculum remained free in suspension; of the remaining 20%, the distribution in the upper respiratory tracts of virus-infected and control animals was significantly different. Of the staphylococci remaining in the nasopharynx of control animals, 74% was present in mucinous plugs, 1 1 % was bound to host cells present in washes of the nasal cavity, and 15% was released by protease treatment of the nasopharynx. Of the staphylococci remaining in the upper respiratory tract of virus-infected ferrets, 36% was recovered in plugs, 24% was bound to cells in nasal washes, and 40% was released by enzyme treatment. Overall, adherence-positive staphylococci represented 64% of recoverable bacteria in virus-infected ferrets versus 26% in controls. Adherence was increased twofold (Days 5 and 7) to threefold (Days 3, 9, and 11) in virus-infected ferrets compared to uninfected controls. In contrast, only 7% of the original streptococcal inoculum was recovered from virus-infected and uninfected control animals and virus infection did not enhance streptococcal adherence except for an approximately threefold increase that was seen on Day 11.

An enzyme-linked lectinsorbent assay (ELLA) was developed for quantification of in situ biofilm produced by Staphylococcus epidermidis in polystyrene 96-well tissue culture plates with phosphatase-labeled wheat germ agglutinin (WGA) as a specific probe for the GlcNAcbeta-1,4n component of exocellular matrix material (EMM) that is responsible for intercellular adhesion and accumulation. The ELLA and the modified Christensen dye assay were used to test 13 laboratory strains of coagulase-negative staphylococci and 10 clinical isolates of S. epidermidis. Four biofilm-positive laboratory strains of S. epidermidis were positive by both tests, and six biofilm-negative strains were negative by both. One strain of S. hemolyticus was positive by the ELLA only. Two of the 10 clinical isolates of S. epidermidis were positive by both assays, two were negative by both, and the remaining were positive by the ELLA only. The ELLA was objective, reproducible, specific, sensitive, and useful for screening strains for their capacity to adhere to plastic, produce EMM, and form biofilm.