Improved cell counting technology

1st April 2013

For years, cell counts in life science research laboratories have been performed using a light microscope and hemacytometer. This method is slow and prone to error. Errors multiply when hemacytometer counts and flow-cytometric (FC) phenotypic percentages are combined to determine cell subset numbers. Digital flow cytometers, utilising laminar-flow fluidics, allow fast, multiparametric data collection (up to 10 000 events per second) on a wide range of cell types but require addition of counting beads to each sample to calculate cell-subset concentration.

Alternatively, flow cytometers with syringe-driven fluidics can deliver absolute count measurements without the addition of counting beads to samples, but are often limited by lower data acquisition rates (<1000 events per second), diminished fluorescence and light-scatter resolution, and a propensity for flow-cell clogging. Flow cytometers with unique peristaltic pump-driven, laminar-flow fluidics systems combine the advantages of hydrodynamically focused cell sampling (high data-acquisition rates, good light scatter and fluorescence resolution) with the ability to report absolute counts for any identified population in a sample.

Two applications of direct cell-concentration determination are presented that make use of the fluorescence and light-scatter measurements possible with a flow cytometer to identify sub-populations of interest. Counting bead or hemacytometer data is included for comparison.

Viability assessment

Initial cultures of U937 and Jurkat cell lines were seeded so that the final analysis cell concentration was dense or light. After one week of culture, cells were resuspended and 1mL transferred into a 1.5mL tube. 7-Amino-actinomycin D (7-AAD, Cayman Chemical), a fluorescent dye which is excluded by viable cells, was used as a marker of cells with compromised outer membranes.

5.0µL of a 1mg/mL 7-AAD solution, along with 50µL of AccuCount Fluorescent Particles (Spherotech; ACFP-50-5) were added to each sample tube and mixed thoroughly. Tubes were kept at room temperature (RT) in the dark, and sampled between five and 30 minutes after addition of 7-AAD, with gentle mixing immediately prior to analysis.

Microscopic cell counts were performed with a hemacytometer. Appropriate dilutions of samples were made into phosphate-buffered saline containing Trypan Blue. Triplicate counts of at least 100 non-blue cells were performed.

The Accuri C6 Flow Cytometer requires only a single 2D density plot of forward light scatter (FSC-A) versus 7-AAD fluorescence (7-AAD FL3-A) for data analysis. A negative control sample for each cell type (no added 7-AAD) defined the viable cell gate. This gate included events with high FSC-A, and defined the FL3-A background fluorescence.

The dead cell gate was set with a cell sample containing 7-AAD. Data was copied into a spreadsheet program to determine the standard deviation (SD) and coefficient of variation (CV) for triplicate measurements (Fig. 1 A-D).

Small particle measurements

Whole blood aliquots (1-2µL), collected in sodium citrate tubes, were diluted 1:10 into HEPES buffered saline with 1 per cent formaldehyde.

Twenty microliter aliquots of diluted blood were incubated in 1.5mL tubes at RT, 20 minutes, with 20µL of CD41-PE antibody (DAKO clone 5B12).

Samples were then diluted with 1 mL HEPES-buffered saline with 1 per cent formaldehyde. 5µL of RFP-50-5 beads (Spherotech) were added to allow counting method comparison. Samples were well mixed and read on the C6.

Data collection was triggered by the positive fluorescence signal of CD41-PE labeled platelets (FL2-H), in order to improve discrimination of platelets from debris and to increase counting accuracy.

The FL2-H threshold channel was determined by first triggering on FSC-H, determining where CD41-PE+ events fell relative to the negatives, and setting this value (FL2-H channel=1000) as the primary threshold.

All samples were collected using this threshold. Platelet counts per µL of sample were copied into a spreadsheet program, and dilution correction factors were applied to determine the platelets per µL of original whole-blood sample (Fig. 2 A-C).


No statistically significant differences were found between the volumetric C6 cell concentration measurements and those obtained with a hemacytometer and counting beads. However, the precision of the cell-count data obtained by C6 volume measurement was significantly better than that obtained by hemacytometer (p<.004) or counting bead (p<.04) methods (Fig. 3). The average CV for triplicate cell counts with the C6 was 1.2 per cent, that for counting bead counts was double (2.8 per cent) and for hemacytometer counts 20 times higher (25 per cent).

The sources of error inherent in hemacytometer counts are well known. The use of counting beads in flow-cytometry experiments has largely replaced combining hemacytometer counts with flow-cytometry data. But even this approach is likely impacted by sources of error such as pipetting technique and calibration, variability in bead-stock concentrations, and the subjective setting of a bead gate in the flow cytometer data file.

Obtaining event counts per µL directly from the C6 data is easy to do. No complicated back calculations to determine the volume sampled based on bead number collected are required, nor are subjective decisions about how to gate on the singlet bead population.


- Clare Rogers and MaryAnn Labant are with Accuri Cytometers, Ann Arbor, MI, USA





To receive the Scientist Live weekly email NewsBrief please enter your details below

Twitter Icon © Setform Limited