ABSTRACT
Robust and precise quantitation of influenza virus is a premise for the efficient development of vaccine production processes. In this article, revised assays for the determination of hemagglutination (HA) and neuraminidase (NA) activity are presented. Bias of traditional discontinuous HA assays and operator dependency was overcome by introduction of a regression procedure. At little effort, a continuous assay result is obtained with repeatability as good as +29%/-22% in the best case (95% confidence intervals reported). Similarly, neuraminidase activity determined in microtiter plates resulted in repeatability better than +/-20%. NA activity decreased almost linearly for pH ranging from 5.8 to 7.8 and was enhanced by the addition of Ca2+. Non-linearity of the assay (due to unspecific adsorption) was overcome by addition of BSA. Using 4-methylumbelliferyl-alpha-D-N-acetylneuraminic acid as substrate, Michaelis-Menten constants of 30 and 460 microM were determined for strains A/PR/8/34 (H1N1) and A/Equi 2/NM/1/93 (H3N8), respectively. The error introduced by approximation of Michaelis-Menten kinetics (zero and first order) was minimized by limiting substrate consumption to about 10%. Linearity of both assays was verified in dilution experiments. Applicability was demonstrated in three cases: virus propagation in mammalian cell culture, ultrafiltration and precipitation of nucleic acids.
