Manufacturing of biologicals requires the use of living cells. Growing in bioreactors, they are vulnerable for viral contaminations of various origins. A contamination could have devastating consequences, from the shut-down of a manufacturing facility, the limitation of drug supplies for patients, the loss of revenues or in worst-case to illness of patients. Viral safety testing is therefore of utmost importance for the detection and identification of adventitious agents. Several in vivo and in vitro methods, like detection of pathogens animal host systems or in cell culture are applicable but limited. They may not be sufficient to identify the virus or to detect very low virus levels. Some methods can only detect specific virus agents that are assumed as possible contaminant.

A possible approach to overcome the limitations of these assays is the Next Generation Sequencing (NGS), also designated as Massively Parallel Sequencing, High Throughput Sequencing or Deep Sequencing, which is an evolving field that is steadily gaining acceptance in the biopharmaceutical industry. This high throughput method is characterized by a high sensitivity and the potential to identify both known and unknown contaminants from a variety of sample matrices.
The following study describes the development of Next Generation Sequencing for the detection of adventitious viruses in Chinese Hamster Ovary (CHO) cells. A virus panel containing Minute Mice virus, Reovirus 3, Encephalomyocarditis virus, Parainfluenca virus 3 and Vesicular Stomatitis virus was selected, representing the biophysical diversity in term of genome form and orientation and their ability to infect CHO cells.
The focus was on the analysis of viral RNAs produced during viral replication to identify contaminations with DNA and RNA viruses by means of RNA-Seq transcriptome analysis. NGS was compared with in vitro virus test in terms of sensitivity and reproducibility. Therefore, different numbers of virus infected cells were harvested 24 hours after infection and diluted with non-infected cells. Furthermore, infected cell samples taken at various time points after infection were analyzed with both methods.
Sequencing of the full virus genomes and the correct identification of the respective contaminant was possible for the whole virus panel with a database containing virus reference sequences. It could be shown that NGS was at least as sensitive as the in vitro virus test.