A collaborative study evaluating the detection of virus-infected cells by transcriptomic high-throughput sequencing

ABSTRACT High-throughput sequencing (HTS) is an agnostic virus detection technology that can replace or supplement the conventional adventitious virus assays used for safety testing of biologics. Transcriptomic HTS is particularly suitable for the detection of replicating viruses. We have evaluated the sensitivity of virus detection by transcriptomic HTS using targeted and non-targeted bioinformatics analysis of infected cells spiked into a background of uninfected cells, mimicking material representing a test sample from product manufacturing. Eight laboratories tested a mix of Raji cells (expressing Epstein-Barr virus [EBV] RNAs without virus production) and EBV-negative Ramos cells at ratios ranging from 0.5 to 0.00005. Seven laboratories performed the analysis using different HTS protocols, while the eighth laboratory quantified EBV RNA recovery in all extracted samples by quantifying EBNA-1 by RT-ddPCR assay. All participants detected EBV transcripts at a 0.001 ratio with the targeted analysis and a 0.01 ratio with the non-targeted analysis (i.e., one infected cell in 100 uninfected cells), with four laboratories achieving detection at the 0.0001 ratio for both analysis. The results of the spiking studies demonstrated the capabilities of transcriptomic HTS for adventitious virus detection in cell lines used for the production of biologics, and highlight that optimization in the workflow can improve HTS virus detection. IMPORTANCE High-throughput sequencing (HTS) is currently recognized as a powerful technology for broad virus detection. While HTS can be considered an alternative method to replace or supplement the conventional adventitious virus detection assays for testing biologics, its routine implementation has been challenging due to the complexities of the HTS workflow, including the matrix of the test sample, selection of HTS methodology (viromic, genomic, or transcriptomic), and variability in the protocols (from sample preparation to bioinformatic analysis). The study demonstrated the performance and capabilities of the transcriptomic HTS approach for the detection of virus-infected cells using different workflows presented by seven laboratories. The study indicated that regardless of the protocol used, all participants detected viral RNA at the 0.01 ratio; however, certain HTS protocols demonstrated better sensitivity of viral detection.