Validation of NucleoCounter, an Automated Cell Counter

Abstract

IntroductionThe most commonly used cell counting method is hemocytometer (HC) trypan dye exclusion due to its low cost and versatility. However, alternative cell counting methods are available for use when large numbers of samples are required to be analyzed rapidly.The NucleoCounter (NC) is an automated cell counter with an optimal counting range of 5E4–5E6 total cells/mL. It utilizes cassettes containing two different DNA‐intercalating fluorescent dyes to stain viable cells green and non‐viable cells blue.Developing a validation method is crucial to introducing an automated cell counter into industrial processes. ICH Q2 guidelines require accuracy and other parameters to be addressed when validating a new method. However, it is difficult to execute accuracy in the absence of a known reference standard. Therefore, a novel approach to augment accuracy determination was employed.While other reports have evaluated accuracy for trypan blue automated cell counters using bead suspensions, a suitable fluorescent bead reference standard was not available for the NC. Therefore, the addition of parallelism was considered an appropriate approach to augment the percent recovery calculations.Parallelism studies were carried out by spiking target cell numbers on top of an existing cell standard curve to create a new target cell standard curve. The two curves generated (original and spiked) would be expected to have similar slopes if the counting method is accurate across the range tested.MethodsPreliminary studies conducted compared HC cell counts to the NC across the 5E5–5E6 total cells/mL range.To execute parallelism, a stock single‐cell suspension of VTL C3A cells (a human hepatoblastoma‐derived cell line) was used to create a 5‐point (5.2E4–4E6 cells/mL) cell concentration standard curve and was counted on the NC. The measured standard curve concentrations and dilution factors were used to calculate the average target cell stock concentration. Next, a target spike of 0.75E6 cells/mL was added to each standard sample and the spiked curve was counted. The percent recovery (actual spiked/target spiked concentration) was calculated at each concentration. A similar parallelism approach was utilized to assess % viability accuracy, in which a non‐viable cell stock (created by treatment in 1% Saponin) was mixed with a viable cell stock in order to generate a reduced viability 5‐point (50–100% viability) standard curve.ResultsPreliminary studies comparing HC to NC cell counts demonstrated that NC exhibits better precision, linearity, and was slightly more accurate at higher concentration of cells.Comparison of the original and spiked standard curves displayed parallelism as evidenced by non‐statistically different slopes (p=0.694) (Figure 1). The % recovery across all concentrations was between 80–130%, which is within acceptable ranges for automated cell counting methods. The standard curves for % viability accuracy assessments (original and reduced viability) were also shown to be parallel to one another with similar slopes (p=0.338).ConclusionsThe results from these studies demonstrated that the NC automated cell counting instrument could be validated using the novel approach of parallelism to augment accuracy determinations in the absence of a reference standard.This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.