SOME ASPECTS OF ESTIMATING VIRUS ANTIGEN CONCENTRATIONS BY ELISA

To estimate the depletion of virusantigen (Cymbidium mosaic virus - CyMV) by immuneadsorption, crude sap preparations were re-collected from the ELISA plate and tested repeatedly in daily intervals for 8 days. After 8 repetitions the extinction value has decreased to about 15 % of the control value, demonstrating that in each test only a small part of the available antigen is bound by antibodies.

After scanning a density gradient of partially purified CyMV, two UV-absorbing peaks were obtained. The two peaks contained predominantly broken and normal particles, respectively. Testing of all density gradient fractions by ELISA revealed high antigen content in many fractions independend from the content of UV-absorbing material or detectable virus particles. By electronmicroscopy short (broken) or normal virus particles were observed in most of those fractions, but with highest concentration in the UV-absorbing fractions. ELISA readings seemed not to depend on intact or broken particles but on total available and reacting virus antigen including subunits.

Attempts to outline a standard curve for ELISA absorbancy readings in relation to different virus concentrations of CyMV (estimated from UV-absorbance of a purified preparation), revealed that a linear proportionality between log virus concentration and log absorbance was found only in the small range of concentrations between 15 and 480 ng/ml. At higher virus concentrations absorbance values showed saturation. Immuno electronmicroscopy (IEM) made visible that this saturation was not due to covering the reactive surface with virus particles. It seemed likely that small antigen, not detectable in the electron microscope (EM) must have saturated the surface of ELISA plates or EM grids, respectively. This hypothesis was supported by the finding that freshly prepared fractions of CyMV particles from a sucrose density gradient (presumed to contain very little breakdown products of particles) could densely cover the surface of EM grids with virus particles.

In consequence we have to realize, that from ELISA readings reference to total antigen concentration, but not to virus particle concentration should be made. For routine assay by ELISA, broad reactivity against virus aggregates, particles and subunits may be useful, for strain differentiation high specificity against intact virusparticles may be advantageous.