1. Introduction
Chinese hamster ovary (CHO) cells (Kao & Puck, 1968) play a key role as heterologous hosts for the commercial-scale production of protein pharmaceuticals (Greber & Fussenegger, 2007; Hacker, De Jesus, & Wurm, 2009; Omasa, Onitsuka, & Kim, 2010). The adaptivity of CHO cells to high cell density processes has ensured their successful application for this purpose. In the biopharmaceutical industry, a large number of products have been introduced into the market using high cell density processes such as batch and fed-batch cultures and different cell culture strategies for reducing the production cost and improving product quality (Pollock, Ho, & Farid, 2013; Tapia, Vazquez-Ramirez, Genzel, & Reichl, 2016). Cell culture times are generally long (12 days for fed-batch cultures) and advances are further prolonging culture times; perfusion continuous bioprocessing (60 days) is widely recognized as the next-generation biomanufacturing platform for achieving superior product quality and reducing manufacturing costs (Konstantinov & Cooney, 2015; Zydney, 2015).
A potential issue when using long-term cultures for therapeutic protein production is the increasing cellular heterogeneity over time (Altschuler & Wu, 2010). Cellular proliferation and gene expression are critical cell processes but genomic mutations arise during DNA replication (Furusawa, 2014) and RNA transcription (Bachl, Carlson, Gray-Schopfer, Dessing, & Olsson, 2001) inherent to cell division and gene expression, respectively. The limited knowledge regarding cellular heterogeneity in high cell density processes poses a barrier to solving this issue.
It has been proposed that insisting on “true” clonality is scientifically unjustifiable (O’Callaghan et al., 2015; Wurm, 2013), with cell line heterogeneity being included in the FDA “Challenges to Potency Assay Development for CGT products” industry guidelines (Guidance for Industry: Potency Tests for Cellular and Gene Therapy Products) (FDA, 2011). We can only hope that the trend toward a modified gene pool can be minimized by maintaining cell populations in stable environments (Wurm, 2013). Deeper characterization such as single-cell sequencing (Tang et al., 2009) of the heterogeneity within cell line cultures would enable the study of cooperative and competitive interactions between cell populations and the mechanisms underlying related cell functions (Ben-David et al., 2018).
In this study, we performed single-cell transcriptome sequencing of serum-free suspension-adapted CHO K1 cells in shake flasks. The degree of cellular heterogeneity and the existence of subpopulations were determined based on gene expression analysis and whole mitochondrial genome sequencing. To facilitate comparisons of our data to those from other single-cell transcriptome studies, we also performed single-cell transcriptome sequencing of adherent CHO K1 cells in culture dishes. Additionlly, we developed clonal markers to validate the clonality of the cells based on their mitochondrial genome sequences.