Reprinted from American Biotechnology Laboratory May 2003
BY COSTAS IOANNIDES
The skill of culturing mammalian cells successfully requires consistent, accurate measurements of cell growth and viability. Cell counts must be taken daily, a task that is not only tedious and time consuming, but also prone to error because of operator subjectivity. To that end, a simple device has been developed that uses well-documented fluorescence microscopy techniques to eliminate operator variances. The automated system detects fluorescent signals bound to cell nuclei, providing accurate cell counts in less than 60 sec. This report details the technology behind the NucleoCounter (New Brunswick Scientific Co., Inc., Edison, NJ; figure 1) and its operation, including data from a run of Chinese hamster ovary (CHO) cells. Alternate counting methodologies, both manual and automated, are also discussed.
Today, most laboratories count cultured cells manually, using a hemocytometer under a microscope to determine total cell number using the Trypan Blue exclusion method to determine cell viability. Because the Trypan blue dye only enters the walls of impaired plasma membranes and is excluded from viable cells, both stained and nonstained cells can be visualized and counted in a light microscope. One drawback, however, is the propensity for staining artifacts, which can lead to higher counts. A second drawback is the proportional increase of stained cells as the length of incubation time between sampling and analysis increases. Furthermore, the naked eye can only differentiate pigment intensities in a limited concentration range of the stain in the hemocytometer chamber. Combining these drawbacks with the potential for cell aggregation, it is easy to see why variation in cell counting is commonplace using this standard methodology. Moreover, it is a common occurrence to acquire cell counts from various cell-counting s pecialists with wide inconsistencies in total cells or cell viability. Such a variation from specialist to specialist can lead to complications in downstream applications, especially when accuracy is critical for the validity of experimentation.
The fully automated NucleoCounter system is optimized to count mammalian cells cultured in research and production applications, such as when using T-flasks or bioreactors. The system features an integrated fluorescence microscope designed to detect signals from a fluorescent dye, propidium iodide (PI), which intercalates to DNA in the cell nuclei (figure 2). Excitation of PI (figure 3) occurs at ~540 nm (green light) and emission (fluorescence) occurs strongly at ~600 nm (red light). The signal intensity is enhanced by a factor of 2030 when PI is bound to DNA, which improves the signal-to-noise ratio. The system detects PI-stained nuclei rather than individual cells. Because nuclei are virtually uniform in size regardless of cell type, no calibration for varying cell size or morphology is required; heterogeneous cell lines such as a-dipocytes and CHO cells have been measured successfully without any adjustment to the instrument. Combined with a charge-coupled device (CCD) camera and integrated image analysis, the compact microscope is designed to permit small and large cell culture facilities to perform fast, efficient, and reproducible cell counts.
Determination of total cell count
The determination of the total cell count involves sample preparation and sample analysis. During sample preparation, a representative cell sample from the cell suspension (200 μL) is mixed with an equal volume of lysis/disaggregation buffer (reagent A-100) and vortexed. A volume of stabilization buffer (reagent B) equal to the initial sample (200 μL) is added to the cel l lysate and vortexed. This subsequent cell lysate is loaded into a Nucleo- Cassette(New Brunswick Scientific Co., Inc., figure 4), where the nuclei are stained with PI, by immersing the tip of the cassette into the sample mixture and depressing the loading piston; this loads a predefined volume (~50 μL) of the sample mixture into the cassette. The Nucleo- Cassette is then placed into the NucleoCounter for analysis. The lysate dissolves the Propidium Iodide, which is immobilized in the cassette, and a mechanical drive in the instrument then transfers the sample into the measuring area of the cassette (figure 4). During analysis the fluorescent signal is registered and correlated to a total cell count (figure 5). The total cell concentration in the NucleoCassette is presented in the NucleoCounter display as cells per milliliter. Optionally, the data can be presented on a PC using the NucleoView software (New Brunswick Scientific Co., Inc.).
Determination of viability
The total cell determination is achieved by deliberately disrupting the plasma membranes of all cells in a sample by pretreatment with reagent A-100 and reagent B. All nuclei are therefore accessible to PI staining, independent of whether cells initially were viable or nonviable prior to cell lysis. However, by also measuring a cell sample without pretreatment, (no addition of Reagent A- 100 or B), only cells with pre-impaired plasma membranes are PI-stained (figure 6). Analysis in the system then gives an estimate of the concentration of nonviable cells in the original cell suspension. By using these two measurements, the viability can be automatically calculated and displayed via the NucleoView software. Alternately, a manual calculation can be applied.
where V=Viability; C=Cell concentration; M=Multiplication (dilution) factor; t=Total cell count; nv=Non-viable cell count.
When the count and viability analysis is complete, the NucleoCassette can be simply and safely disposed of; and the next sample can be run without delay, and without the need to clean, re-calibrate or maintain the NucleoCounter.
Comparison using CHO hIR Cells
CHO cells overexpressing the human insulin receptor (CHO hIR) with known cell concentrations were counted manually and compared with readings from identical samples with the NucleoCounter (figure 7). Triple concentration determinations indicate improved precision was attained with the system compared with the standard manual method. As indicated by the error bars, the standard deviations for the microscopic method are indicative of the high variability of the technique, whereas the system readings maintained a greater uniformity over a wider range of cell counts. Furthermore, it was shown that the NucleoCounter readings generated coefficients of variation (CV) of <5% under normal conditions.
Additionally, samples were acquired from T-flasks at figure 6 PI is excluded from viable cells. This is utilized in the NucleoCounter to estimate the concentration of nonviable cells and the concentration of total cells in a suspension. figure 7 Precision comparison of NucleoCounter versus manual counting methods. figure 8 Comparison of growth curves observed using either the NucleoCounter or manual methods. various time points, trypsinized and resuspended in Dulbecco's modified Eagle's medium (DMEM). The cells were lyzed with 2% Triton X-100 (Sigma Chemical Co., Perth, Australia) and counts were obtained using both manual method and the NucleoCounter (figure 8). The comparison of the data indicates almost identical growth patterns with either method. However, the time required to acquire the data using the system was a fraction of that required by the manual method.
Other automated methods of cell counting currently in use have not solved the critical points of reproducibility and rapidity to satisfaction. To date, automated viability determinations have been based on either the Trypan blue exclusion method or on differentiation between cells sizes using impedance. Many of these systems require frequent calibration for cell size or morphology prior to use; often, these requirements compromise the accuracy of the final result. As indicated in the example above, the NucleoCounter delivers reproducible precision and is a low-maintenance unit that does not require calibration for cell type or morphology.
Mr. Ioannides is an employee of New Brunswick Scientific Co., Inc., Box 4005, 44 Talmadge Rd., Edison, NJ, 08818-4005, U.S.A.; tel.: 877-723-3314; fax: 732-287-4222; e-mail: email@example.com; home page: www.nbsc.com/ABL7.htm. NucleoCounter , NucleoCassette, and NucleoView software are registered trademarks of ChemoMetec A/S, Allerd, Denmark, and are distributed exclusively in North America by New Brunswick Scientific Co., Inc.