Comparison of Manual Cell Counts and Fluorescence Plate Readings in Determining Neuronal Cell Density

Neurons are one of the two principal cell types in the mammalian nervous system. These highly specialized cells consist of a cell body (soma), an axon, and several to many dendrites. Neurons transmit electrochemical impulses through specialized nerve endings called synapses. During nervous system development many neurons die by a process called programmed cell death or "apoptosis," and this is important for regulating brain size and organization. However, with rare exceptions, neurons in the adult brain are post-mitotic and cannot be replaced if killed. Thus, neuronal apoptosis in the mature nervous system can lead to severe neurological dysfunction. Research has determined that members of the bcl-2 and caspase gene families regulate apoptosis in many cell types, including neurons. To determine which specific family members regulate neuronal apoptosis, our laboratory studies the effects of targeted gene disruptions on neuron survival both in vivo and in vitro.

In a typical experiment, neuronal cell cultures, derived from the embryonic mouse brain, are used to determine if a specific targeted gene disruption causes increased or decreased susceptibility to neuronal apoptosis. The neuronal cells are cultured in 48-well plates and death-promoting agents can be added at various concentrations. To quantitate the effect of these agents, the cells are stained with bisbenzimide, a fluorescent dye that binds to DNA in all cell nuclei, and with fluorescein-linked microtubule associated protein 2 antibodies that label neuronal cell bodies and dendrites. Following labeling, each well must be examined microscopically and the number of nuclei and of neurons quantitatively determined. This method is potentially biased by the observer, relies on assessment of non-randomly selected "representative" fields, and is extremely time-consuming. The purpose of my project was to determine if the fluorescence plate reader FLx800 (Biotek) could be used instead of manual counting to produce accurate, rapid, and non-biased quantitation of cell nuclei and neurons, thus replacing the researcher in this task.

The fluorescence plate reader FLx800 uses a photomultiplier to measure the fluorescence of the cellular assay on the plate. A 3.0-mm diameter probe moves to the center of each well and takes ten fluorescence readings, which are averaged to determine the fluorescence value for that well. Excitation and emission filter wheels are programmed to read specific fluorophores such as bisbenzimide and fluorescein. Wells with high densities of stained cells receive higher fluorescence readings than wells with low densities of stained cells.

To determine if a fluorescent plate reader produces cell density values comparable to those determined by manual counting, wild type mouse embryo neurons were cultured on a 48 well plate for 24 hours. Cells were then subjected to AraC, a nucleotide analogue that induces neuronal apoptosis, in doses of 0, 1, 10, and 100 µM for 48 hours. Antibodies against the neuron-specific protein, MAP2, and immunocytochemical techniques were used to fluorescently label neurons in the cultures. The culture was then stained with bisbenzimide to detect DNA in cell nuclei.

Two separate researchers each counted bisbenzimide and fluorescein stained cells using a fluorescence microscope, a process that took about three hours. The researchers determined cell density by counting all the bisbenzimide stained nuclei and MAP2 immunoreactive neurons in two representative viewing fields of each well. The average cell density values for each AraC dose were then calculated. The FLx800 determined fluorescein and bisbenzimide fluorescence values for all 48 wells in about two minutes. The fluorescence plate reading values were then compared with the two human-obtained values to determine the relationship between quantitation methods. A clear AraC dose-response was observed with both methods, and the average dose density values, when plotted, showed similar slopes. By plotting individual well counts against fluorescence readings, regression lines were calculated which illustrated a high correlation (R² = .91) between manual counts and fluorescence readings.

This work concludes that fluorescence plate readings can be used to accurately determine cell densities. The readings are completed in less time than manual counting, and the possible bias of researcher-chosen representative viewing fields is eliminated. Hopefully, the replacement of manual cell counting with fluorescence plate readings will contribute to a more productive lab environment by allowing for mechanically produced, quick, accurate, and unbiased cell density values.