Rousselle et al. 1998

Rousselle C, Robert-Nicoud M, Ronot X. 1998. Flow cytometric analysis of DNA content of living and fixed cells: a comparative study using various fixatives. Histochemical Journal 30: 773-781.

These authors tested a range of fixatives for their effects on cell parameters measured in flow cytometry. Two different fluorochromes, Hoechst 33342 and Propidium iodide were tested, with a standard fixation procedure that varied little between treatments. Fixatives were derived from the relevant literature, including static rather than flow cytometric experiments.

All fixative treatments were applied after initial cell processing steps, including isolation of individual cells using a series of Phosphate-buffered-saline (PBS) washes and Trypsin. No detergent was used to permeabilize or remove cell membranes, though there is some discussion of the use of saponin as a permeabilizing agent in other studies.

Hoechst 33342 will penetrate living cells and stain the DNA in a stoichiometric fashion, allowing these authors to compare fixatives to living cells with this stain. Under these treatments, 68% or 70% ethanol (the figure disagrees with the text on this detail) provides excellent resolution of nuclear DNA content and cell cycle (G0/1 vs. S vs. G2 phases, ratio of G2 to G0/1). Other fixatives such as 85% methanol and Acetone also provide good results, with poor and inconsistent results from Carnoy and Boehm-Sprenger treatments.
The results for Propidium iodide were similar, though because Propidium iodide does not penetrate living cells, ethanol was used as the standard for comparison.

Cell size and cell granularity were measured under all treatments, using forward scatter for size and orthogonal (presumably equivalent to side-scatter) scatter for granularity. It is not clear in this paper what the precise definition of “granularity” might be, but it seems to be related to membrane permeabilization and DNA staining stoichiometry. In any case, all fixatives except 1% formaldehyde caused significant cell shrinkage, and all fixatives caused increased granularity. Though they do not discuss it, ethanol fixation produced relatively non-variable cell shrinkage compared to the other fixatives, shrinking cells to 0.74 +- 0.03 of living cell volume vs. 0.81 +- 0.05 for methanol.

This study did not find that alcohols increased cell aggregations, in contrast to an earlier study by Schimenti and Jacobberger (1992) in which 67%, 81% and 90% ethanol fixation caused cell clumping and increased debris. Alcohols increase cell membrane permeabilization, presumably allowing greater access to the DNA by stains. In this study, acetone treatment was basically not good compared to other fixatives.

This study did not examine the effects of long-term storage in fixatives such as ethanol. Fixatives were applied for 30 minutes, then washed away by centrifugation of cells to a pellet, removal of fixative supernatant, and resuspension in PBS. PBS seems to be the fluid of choice for many flow cytometry studies, and these authors note that separated cells can be stored for months in cold ethanol before flow cytometry, though they do not describe a procedure to replace ethanol with PBS or other solutions.