(1221-B) Assessment of cloning efficiency and error rates of various single-cell dispensing technologies

Abstract: Current regulatory guidance for the biotechnology industry calls for reasonable assurance of monoclonality for production cell lines. Cloning procedures should be fully documented, including details of imaging techniques where applicable. In recent years, the advent of new technologies has significantly shortened the cell line development workflow for novel clinical assets, compared to time-consuming and labor-intensive limiting dilution cloning, which requires two or more rounds of single cell isolation. However, each novel technology has pros and cons in terms of ease-of-use, cost of implementation, efficiency of seeding and cloning, and – most importantly – error rates. The error rate provides a measure of how often a cell is missed by the cloning process across an entire cloning experiment. Both single-cell deposition and single-cell imaging modalities have associated error rates which compound to impact the probability of clonality. Here, we assess the error rates of different established and emerging technologies by direct side-by-side comparison, including i) impedance, ii) microfluidics, iii) nanodroplet, and iv) gravity-based deposition. We contrast these findings to conventional limiting dilution cloning (LDC).
To assess variability across different CHO backgrounds we generated stably transfected clonal cell lines (two different commercially available expression platforms) expressing red or green fluorescent proteins (RFP, GFP, respectively) for high-sensitivity detection. RFP and GFP expressing cell lines of the same host background were matched for doubling time, fluorescence intensity, and outgrowth characteristics. Prior to single cell cloning, RFP and GFP expressing clones were mixed at equal ratio and co-cultured overnight to allow for potential cell-to-cell interaction and aggregation. The co-cultured cells were single-cell dispensed into four 96-well plates for each of the four different dispensing technologies. In addition, twelve 96-well plates were seeded manually using LDC at 0.5 cells/well. After 14 days of incubation, plates were imaged on a high-resolution automated microscopic whole well imager using both red and green fluorescence channels. Then, each well was manually inspected to identify any mixed populations of cells in a well. Outgrowth (cloning efficiency) and the rate of undetected errors from each technology was determined across both CHO expression system backgrounds.
We made several key observations: i) We found differences in outgrowth rates across all evaluated technologies, with the best results being comparable to those from limiting dilution cloning. ii) All tested dispensing modalities show superior outcomes with regards to cloning error rates when compared to limiting dilution cloning. iii) We observe meaningful differences in the error rates between impedance, microfluidics, nanodroplet and gravity-based deposition technologies.
Taken together, we outline a rigorous comparative approach for the assessment, qualification, and selection of different single cell cloning technologies in different CHO genetic backgrounds.