A comparison of methods for the estimation of retroviral burden

The presence of retroviral contamination is of vital concern in the manufacture of cell culture-derived biopharmaceuticals. These cell lines usually have A- or C-type retrovirus-like particles which are visible by transmission electron microscopy (TEM) even when infectivity (IF) or reverse transcriptase activity (RTA) cannot be demonstrated. The supernatant of the post-production cell cultures, therefore, also needs to be evaluated by TEM for viral burden. A major question, however, is how to establish a quantitative viral load estimate for the evaluation of a purification process. The FDA recommends that a purification process for viral contaminants remove or inactivate 3-5 logs over the estimated viral burden. Viral particles are difficult to identify and quantify, however, by conventional negative staining. We present a comparison of infectivity assay, reverse transcriptase assay, negative staining, and thin sectioned TEM. These assays were performed on four samples. Ultracentrifuged sediments of cleared cell-culture media were measured, fixed and processed. Thin sections were evaluated by TEM and the number of viral particles estimated by morphometric derived quantification. Retrovirus particles were easily identified and quantified when examined by TEM as compared to negative staining and correlated with the other viral assays (IF, RTA). These results demonstrate that the TEM thin section method was a superior technique to negative staining for estimating viral particle load in cell-culture supernatant. To validate further the plastic embedding with thin sectioning, we evaluated cell culture supernatants (pellets) for retroviral burden at various dilutions, from two cell lines. Morphometric determinations were made of the number of viral particles present per unit volume and compared to results obtained by infectivity assay. Since the morphometric calculation for viral density assumes even distribution of viral particles, we also evaluated and calculated viral counts on multiple thin sections taken throughout selected pellets.