Search

A single radial hemolysis test was developed for quantitation of specific antibody to non-hemagglutinating viruses. With the human coronaviruses as models, this test utilizes the binding properties of the chromic cation to attach viruses to glutaraldehyde-treated sheep erythrocytes. The most satisfactory system consisted of stabilizing washed sheep erythrocytes with 0.0073% glutaraldehyde for 15 min at 23°C, binding a high concentration of virus to a 25% erythrocyte suspension with 0.0016% chromic chloride for 20 min at 23°C, stopping the reaction with phosphate-saline, and finally mixing the treated, rewashed cells with complement and agarose at 45°C to prepare a slide gel. The gel mix, which was dispensed in plastic plates (23 by 73 mm) in 3-ml volumes, consisted of 1% agarose, 0.1% sodium azide, 5% reconstituted complement, and 0.82% treated cells. Wells 2 mm in diameter were loaded with 5 µl of antiserum, incubated for 18 h at 4°C for diffusion of antiserum and fixation of complement, and then incubated for 8 to 24 h at 37°C for development of hemolysis zones. The diameter of a zone was linearly related to antibody concentration, as determined by conventional serological tests. This single radial hemolysis test was applicable to human and animal coronaviruses and to selected serotypes of the adenovirus, picornavirus, rhabdovirus, and rotavirus groups.

Virus particles morphologically resembling adenovirus were found in fecal specimens from infants and were examined for cultivability with standard cell culture techniques and for characteristics of human adenoviruses. Specimens from 13 of 15 infants could not be cultivated in cell cultures. The two adenoviruses that were cultivated, types 1 and 31, reacted in the expected manner in all tests. Counterimmunoelectrophoresis with group-specific anti-hexon serum confirmed that the observed particles in the 15 specimens were human adenoviruses. The buoyant density in sucrose of five of the noncultivable adenoviruses in original stool suspensions averaged 1.335 g/cm3and that of the two cultivable ones averaged 1.332 g/cm3; both groups had typical adenovirus morphology by electron microscopy. Treatment of the specimens and of a variety of tissue culture cells with proteolytic and other enzymes did not improve cultivability. Examination of partially purified virus by immunoelectron microscopy did not reveal evidence of immunoglobulin A, G, or M coating on the particles, an indication that coproantibody inhibition was not the cause of noncultivability. Fluorescent-antibody studies with an antihexon conjugate and counterimmunoelectrophoresis studies of serially passaged noncultivable viruses indicated that the viruses are infecting cells but are not undergoing effective replication. Antisera to three of the noncultivable viruses demonstrated homologous reactions in counterimmunoelectrophoresis with the respective immunizing antigens but showed only low levels of hemagglutination-inhibiting and neutralizing activity to a few of the known human adenoviruses. We concluded that the noncultivable viruses in these infant diarrhea cases were indeed human adenoviruses, were not defective particles, were not bound to coproantibody, were infectious but incapable of effective relication in conventional cell cultures, were serologically related to types 11, 17, 32, and 33, and should be considered a new, distinct subgroup.