UNIT 9.2 Assessment of Cell Viability

 

Unit Introduction

The method used to determine cell viability (and to a degree, the definition of viability) is often related to the phenomenon studied. Frequently, cell viability is thought of in somewhat negative terms. That is, one needs to exclude dead cells because they generate artifacts as a result of nonspecific binding and/or uptake of fluorescent probes. In addition to simple enumeration of live or dead cells present, there is a broad range of biologically relevant cytometric procedures that are related to the physiological state of the cells measured. For example, one may wish to measure cell morbidity in apoptosis, cell survival as a result of cytotoxicity, the potential of bacteria or microalgae to survive environmental insult and grow once normal conditions have been restored, or the potential of intracellular protozoa to undergo division cycles as part of the infective process.

Cell viability may be judged by morphological changes or by changes in membrane permeability and/or physiological state inferred from the exclusion of certain dyes or the uptake and retention of others. Here methods are presented for staining nonviable cells by dye exclusion (indicative of an intact membrane) using the fluorescent, DNA-binding probes propidium iodide (PI) (see Basic Protocol) and 7-amino actinomycin D (7-AAD; see Alternate Protocol 1). These probes may also be used in cells labeled with phycoerythrin (PE)-conjugated antibodies (see Alternate Protocol 2). The next two protocols present different aspects of physiological state that can be used to assess viability; one is based on esterase activity (see Alternate Protocol 3) and the other on mitochondrial membrane potential (see Alternate Protocol 4). For fixed cells, the state of viability prior to fixation can be determined using DNA-binding probes either before (see Alternate Protocol 5) or after (see Alternate Protocol 6) fixation. Each of the above protocols requires basic understanding in cell handling and flow cytometry for which they are designed. In contrast, the final method presented (see Alternate Protocol 7) is a dye exclusion procedure for microscopy using trypan blue and a hemacytometer (appendix 3A & 3B).

 

Assessment of Cell Viability Using Probes for Membrane Integrity

Live cells with intact membranes are distinguished by their ability to exclude dyes that easily penetrate dead or damaged cells. Staining of nonviable cells with propidium iodide (PI) has been performed on most cell types. Its broad application is most likely due to ease of use: the procedure is very simple, and the stained cells are bright red and easy to identify. Alternatively, the longer-wavelength emission of 7-amino actinomycin D (7-AAD) makes it easier to use simultaneously with phycoerythrin (PE) as a surface marker, as detailed in Alternate Protocol 2, or with fluorescein isothiocyanate (FITC). Although 7-AAD is not as bright as PI, permeable cells are easily distinguishable from live cells.

Basic Protocol:  Propidium Iodide Staining of Nonviable Cells

 Materials

• 2 mg/ml propidium iodide (PI) in PBS (store wrapped in foil £1 month at 4°C)
• Cell suspension
• PBS (appendix 2A)
• 13 × 100–mm polystyrene culture tubes

CAUTION: Propidium iodide is a suspected carcinogen and should be handled with care. In particular, be careful of particulate dust when weighing out the dye. Use gloves when handling it.

 1.

Add 1 µl of 2 mg/ml propidium iodide (2 µg/ml final) to approximately 10^6 washed cells suspended in 1 ml PBS in 13 × 100–mm polystyrene culture tubes..

To use this procedure with cells that are also labeled with PE-conjugated antibodies, see Alternate Protocol 2.

 2.

Incubate ³5 min in the dark on ice.

 3.

Analyze on flow cytometer with excitation at 488 nm and emission collected at >550 nm.

PI is easily excited at 488 nm. The dye has a broad fluorescence emission and can be detected with photomultiplier tubes (PMTs) normally used for phycoerythrin (~585 nm) or at longer wavelengths (³650 nm).

Amplify the PMT signal logarithmically to distinguish populations of permeable (and presumed dead) cells from viable cells. Adjust the PMT high voltage such that bright cells are well separated from dim, viable cells. It is often easy to identify nonviable cells on a bivariate plot of forward light scatter (see Fig. 9.2.1).

Figure 9.2.1 Identification of nonviable cells with propidium iodide (PI). Nonviable cells are more than two decades brighter than the unstained, viable cells. Gating on a one-parameter histogram is sufficient to identify the viable population. Region 1: viable cells; region 2: nonviable cells.

Dead cells can be live-gated, but unless one is absolutely sure of the viable population, it is always much better to collect ungated listmode data and perform gating after the raw data files are collected. Populations that may not be obvious on the flow cytometer display will be seen during subsequent data analysis. Moreover, any gating can be done and redone without losing cells.

Alternate Protocol 1:  7-AAD Staining of Nonviable Cells

 Additional Materials (also see Basic Protocol)

• 1 mg/ml 7-amino actinomycin D (7-AAD; see recipe)

 1.

Add 1 µl of 1 mg/ml 7-AAD (1 µg/ml final) to approximately 10^6 washed cells suspended in 1 ml PBS in 13 × 100–mm polystyrene culture tubes.

To use this procedure with cells that are also labeled with PE-conjugated antibodies, see Alternate Protocol 2.

 2.

Incubate ³30 min in the dark on ice.

 3.

Analyze on flow cytometer with excitation at 488 nm. Collect fluorescence emission with a 650-nm long-pass or a 670 ± 20–nm band-pass filter.

7-AAD is easily excited at 488 nm. The fluorescence emission of the dye has a peak at ~670 nm.

Use logarithmic amplification to distinguish permeable and bright cells from nonpermeable cells. Adjust the PMT high voltage to resolve a population of viable cells in the first decade of the 7-AAD fluorescence histogram as shown in region 1 of Figure 9.2.2A; nonviable cells are in region 2 of Figure 9.2.2A.

Figure 9.2.2 Effects of gating and compensation with 7-AAD. (A) Gating discriminates live cells. One-parameter histogram of logarithmically amplified 7-AAD fluorescence using a 650-nm long-pass filter. Mouse spleen cells are labeled only with 7-AAD. Note the peak of dead cell population in region 2 at a relative brightness between 100 and 200 (about ten-fold dimmer than what is expected for propidium iodide). The live cell population that occupies the first decade in the histogram (region 1) is 7-AAD negative and constitutes the majority of the cells in the population. (B) Uncompensated phycoerythrin fluorescence in the presence of 7-AAD. A bivariate plot of mouse spleen cells labeled with 7-AAD and a PE-labeled antibody to a cell surface antigen. Note the two small populations of dead cells in regions 2 and 4 and the large population of live cells that occupies region 2. If gating was performed before adequate compensation was achieved, then most of the viable PE-positive cells could be lost. (C) Compensation of PE with 7-AAD. Distribution of cells from same sample as in B. Note the dead cells are in the same location as in B, but the live cells are now clearly resolved from 7-AAD positive populations. At this point, the live cell gate defined in region 2 of A is valid.

Alternate Protocol 2:  Use of PI or 7-AAD for Cells Labeled with PE-Conjugated Antibodies

 Additional Materials (also see Basic Protocol)

• PE-labeled cell suspension (unit 6.2)

1.

Stain and incubate PE-labeled cells with PI (see Basic Protocol, steps 1 and 2) or 7-AAD (see Alternate Protocol Alternate Protocol 1, steps 1 and 2).

2.

Analyze on flow cytometer with excitation at 488 nm. Use a 585 ± 20–nm band-pass filter for detection of PE fluorescence and a 650-nm long-pass filter for PI or 7-AAD. PI, 7-AAD, and PE are all excited by 488-nm light. Despite the separation by two detection filters, there is substantial overlap between the PE and PI or 7-AAD, requiring compensation between the two detectors. This problem is illustrated for PE and 7-AAD in Figure 9.2.2 (parts B and C), though the same procedure is used with PI. Note the typical discrimination of viable/nonviable cells. If one were to set a gate region on the viable cells in region 1 and use that for analysis in the absence of proper compensation, most PE-positive cells would be eliminated as nonviable (see Fig. 9.2.2B). Using a bivariate histogram of log 7-AAD versus log PE fluorescence, adjust the PE-7-AAD compensation such that PE-positive, 7-AAD-negative cells are above the PE-negative, 7-AAD-negative population (see Fig. 9.2.2C). Once proper compensation of the long-wavelength component of PE is subtracted from the output of the detector for 7-AAD, the problem disappears.

 

Assessment of Cell Viability Using Probes of Physiological State

These protocols describe the use of probes that require a specific cellular function in addition to an intact membrane. Alternate Protocol Alternate Protocol 3 describes the use of fluorescein diacetate (FDA), which requires cellular esterase activity, and Alternate Protocol Alternate Protocol 4 describes the use of rhodamine 123 as a probe for mitochondrial membrane potential.

Alternate Protocol 3:  Fluorescein Diacetate Staining of Viable Cells

Cell viability can be assessed directly through the presence of cytoplasmic esterases that cleave moieties from a lipid-soluble nonfluorescent probe to yield a fluorescent product. The product is charged and thus is retained within the cell if membrane function is intact. Hence, viable cells are bright and nonviable cells are dim or nonfluorescent. Typical probes include fluorescein diacetate (FDA, described here), carboxyfluorescein, and calcein. Variations in uptake or retention of the dye among individual cells or under different conditions affect the efficacy of particular probes.

 Additional Materials (also see Basic Protocol)

• 1 mg/ml fluorescein diacetate (FDA; prepare fresh in acetone in a 13-mm glass culture tube and cover with foil)
• Cell suspension in culture medium appropriate for the cell type

1.	Add 2 µl of 1 mg/ml FDA (2 µg/ml final) to approximately 10^6 cells in 1 ml medium in a 13 × 100–mm polystyrene culture tube.
2.	Vortex to mix and incubate 15 min at 37°C.
3.	Analyze on flow cytometer immediately with excitation at 488 nm. Collect fluorescence using a 530 ± 20–nm band-pass filter. FDA is excited by 488-nm light and fluoresces green. Filters used for measuring fluorescein (e.g., 530 ± 20–nm band-pass) are sufficient to visualize the nonfluorescent cells on the same scale. Use logarithmic amplification of the PMT output. Cells that take up and retain free fluorescein are very bright (approximately two decades brighter on a logarithmic scale) and should be easily distinguishable from nonfluorescent, nonviable cells. Unless one is absolutely sure about the location of viable cells in the histograms, it is preferable to collect listmode data files of all samples and perform gating after the raw data files are collected, to avoid the danger of inadvertent loss of viable cells.

Alternate Protocol 4:  Rhodamine 123 Staining of Viable Cells

Another property of viable cells is the maintenance of electrochemical gradients across the plasma membrane. Functional subsets of this general phenomenon include the maintenance of pH and other ion gradients as well as the capacity for energy-yielding metabolism in mitochondria. These physiological processes can be exploited to distinguish viable from nonviable cells. One of the most commonly used probes for identifying viable cells is rhodamine 123, a cationic lipophilic dye that partitions into the low electrochemical potential of mitochondrial membranes. Active mitochondria in viable cells are stained bright green; loss of gradients within nonviable cells results in loss of the dye.

 Additional Materials (also see Basic Protocol)

• 1 mg/ml rhodamine 123 (prepare fresh in distilled water)
• Cell suspension in culture medium appropriate for the cell type

 1.

Add 5 µl of 1 mg/ml rhodamine 123 (5 µg/ml final) to approximately 10^6 cells in 1 ml medium in a 13 × 100–mm polystyrene culture tube.

 2.

Vortex to mix and incubate 5 min at 37°C; return to room temperature.

 3.

Analyze immediately with excitation at 488 nm. Collect fluorescence using a 530 ± 20–nm bandpass filter.

Rhodamine 123 absorbs 488-nm light and fluoresces green. Collect fluorescence through a filter that transmits at ~530 nm, as for FITC. PMT output should be amplified logarithmically.

Viable cells are brighter than nonviable cells, though with some cell types there may be some overlap (see Fig. 9.2.3A). A bivariate plot of forward scatter versus rhodamine 123 fluorescence can be useful to distinguish viable from nonviable cells (see Fig. 9.2.3B).

Figure 9.2.3 Effects of gating with rhodamine 123. (A) Identification of live cells after gating. Rhodamine 123 may not always completely resolve viable from nonviable cells as indicated in the ungated histogram (dotted line). Gating on forward light scatter versus rhodamine 123 fluorescence helps separate both populations. Note the histogram of the gated population of viable cells (solid line) overlaid on the ungated population. (B) A bivariate plot of forward light scatter versus rhodamine 123 fluorescence helps to resolve live (rhodamine 123–bright) and dead (rhodamine 123–dim) populations. Debris is gated out at the same time.

Unless one is absolutely sure about the location of viable cells in the histograms, it is preferable to collect listmode data files of all samples and perform gating after the raw data are collected, to avoid the danger of inadvertent loss of viable cells.

 

Assessment of Cell Viability in Fixed Cells

For reasons of safety or convenience, it is frequently necessary to fix cells prior to analysis. Data analysis is less ambiguous if nonviable or damaged cells can be eliminated, but the methods discussed above will not work, because fixation will render all cells permeable. There are, however, DNA probes that penetrate and stain dead or damaged cells and that can withstand fixation. Ethidium monoazide (EMA) is positively charged and penetrates the membranes of dead or damaged cells but not live ones. EMA can be photochemically cross-linked with short exposure to visible light; after the excess dye is washed away, the cells are fixed. Another DNA-binding fluorochrome used for staining nonviable cells is laser dye styryl-751 (LDS-751). The procedure is somewhat more simple, as staining is done after fixation and does not require cross-linking.

Alternate Protocol 5:  Ethidium Monoazide Staining of Nonviable Cells Prior to Fixation

 Additional Materials (also see Basic Protocol)

• 50 µg/ml ethidium monoazide (EMA; see recipe)
• 1% (w/v) paraformaldehyde in PBS (see appendix 2A for PBS recipe; store mixture £1 week at 4°C and discard if precipitate forms)
• 40-W fluorescent light

1.	Add 10 µl of 50 µg/ml EMA (~5 µg/ml final) to approximately 10^6 washed cells suspended in 100 µl PBS in a 13 × 100–mm polystyrene culture tube. Preparations of EMA dye vary. The exact concentration needed may range from 1 to 5 µg/ml.
2.	Place tubes on ice ~18 cm beneath a 40-W fluorescent light for 10 min.
3.	Wash and fix cells by adding 1 ml µl of 1% paraformaldehyde to the cell pellet. Incubate 1 hr at room temperature.
4.	Analyze on flow cytometer with excitation at 488 nm. Collect fluorescence emission using a ³630-nm long-pass filter; amplify PMT output logarithmically. EMA excites with 488-nm light and fluoresces well into the red region of the spectrum. EMA does not fluoresce as brightly as propidium iodide, so discrimination of nonviable, EMA-bright cells from viable cells may be less obvious. A bivariate plot of forward light scatter versus EMA fluorescence may aid in distinguishing viable (EMA-negative) cells. Dead cells can be live-gated, but unless one is absolutely sure of the viable population, it is always better to collect ungated listmode data and perform gating after the raw data files are collected. Populations that may not be obvious on the flow cytometer display will be seen during subsequent data analysis. Moreover, any gating can be done and redone without losing cells.

Alternate Protocol 6:  LDS-751 Staining of Previously Nonviable Cells After Fixation

 Additional Materials (also see Basic Protocol)

• 1% (w/v) paraformaldehyde in PBS (see appendix 2A for PBS recipe; store mixture £1 week at 4°C and discard if precipitate forms)
• 2 µg/ml LDS-751 (laser dye styryl-751) working solution (see recipe)

1.	Wash approximately 10^6 cells in PBS in a 13 × 100–mm polystyrene culture tube.
2.	Fix cells by adding 1 ml µl of 1% paraformaldehyde to the cell pellet. Incubate 1 hr at room temperature.
3.	Add 10 µl of 2 µg/ml LDS-751 working solution to 1 ml fixed cells at a concentration of approximately 10^6 cells per ml.
4.	Incubate overnight at room temperature in the dark.
5.	Analyze on flow cytometer with excitation at 488 nm. Collect fluorescence emission using a 650-nm long-pass filter. LDS-751 is excited with 488-nm light and emits in the red portion of the spectrum; the 650-nm long-pass filter is adequate to separate red fluorescence from other fluorochromes and scatter laser light. Amplify the PMT output logarithmically to distinguish bright, nonviable cells from dim, viable cells and from nonfluorescent red cells or debris. A bivariate plot of forward scatter versus log LDS-751 fluorescence can help identify populations. Dead cells can be live-gated, but unless one is absolutely sure of the viable population, it is always much better to collect ungated listmode data and perform gating after the raw data files are collected. Populations that may not be obvious on the flow cytometer display will be seen during subsequent data analysis. Moreover, any gating can be done and redone without losing cells.

 

Assessment of Cell Viability by Microscopy

Alternate Protocol 7:  Using Trypan Blue Staining

Assessment of cell viability may be accomplished with a microscope, using dyes that mark nonviable cells by dye exclusion. The most commonly used dye is trypan blue, but others may be used as well; see Background Information for details on other useful stains. Viable cells have intact membranes and exclude the dye; nonviable cells are labeled with the dye and are visible with brightfield optics. As well as being useful as a means of assessing functional integrity, trypan blue exclusion is widely used as an objective method of determining viable cell count prior to using cells; a simple protocol for this application is presented, along with other basic cell culture techniques, in appendix 3B.

 Additional Materials (also see Basic Protocol)

• 0.4% (w/v) trypan blue in PBS (store up to 1 year at room temperature in the dark; filter if a precipitate forms; for PBS recipe, see appendix 2A)
• Serum-free culture medium (appendix 3B optional)
• Additional materials for cell counting with a hemacytometer (appendix 3A)

1.	Add an equal volume of 0.4% trypan blue to a cell suspension at a concentration of approximately 10^6 cells per ml. Use PBS or a serum-free medium for the cell suspension. Serum proteins may stain with trypan blue, resulting in falsely depressed viable counts.
2.	Incubate at room temperature for ~3 min and load into a hemacytometer. Using brightfield optics, count cells in three separate fields (see appendix 3A for use of hemacytometer). Count nonviable, deep blue cells as well as viable, clear cells.
3.	Calculate viability: % viable = (number viable cells/number total cells) × 100.

 

Reagents and Solutions

Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see appendix 2A; for suppliers, see suppliers appendix.

 7-AAD (7-amino actinomycin D), 1.0 mg/ml

Dissolve 1.0 mg 7-AAD in 50 µl of dimethyl sulfoxide (DMSO), and then add 950 µl of PBS (appendix 2A). Store £1 month in the dark at 4°C.

As 7-AAD is not water soluble, an organic solvent is required. Although the mutagenicity of 7-AAD is unknown, caution should be exercised when handling the dye.

 EMA (ethidium monoazide), 50 µg/ml

Dissolve ethidium monoazide at 5 mg/ml in PBS (appendix 2A). Dilute to 50 µg/ml (working strength) and divide into 0.5-ml aliquots. Wrap in aluminum foil and store £6 months at –20°C. Thaw immediately before use. Discard unused working-strength dye.

EMA is very light sensitive; keep wrapped in foil or in the dark.

 LDS-751 (laser dye styryl-751) working solution, 2 µg/ml

Stock solution: Dissolve LDS-751 (Exciton Corp.) at 0.2 mg/ml in methanol. Store up to 1 month at 4°C in the dark.

Working solution: Add 0.5 ml stock solution to 50 ml PBS (appendix 2A; 2 µg/ml final). Store up to 1 week in the dark at 4°C.

 

Commentary

 

Background Information

Assessments of viability depend on one or both of two cellular properties: (1) the intactness of the cell membrane, and (2) the physiological state of the cell.

Dye exclusion methods are based on the fact that only intact membranes are impermeable to large or charged molecules. Intact membranes also maintain cytoplasmic gradients with respect to the surrounding medium, thus retaining intracellular concentrations of ions and small molecules. This latter property also reflects the physiological state of the cells in that energy is required to maintain gradients. Thus, methods that assay physiological properties of the cell also are dependent upon and indicative of an intact membrane.

 

Probes for Membrane Integrity

Permeability of the cytoplasmic membrane is commonly exploited to mark cells that are moribund or dead. The reagents most often used for this purpose are dyes such as trypan blue or a variety of fluorochromes that will penetrate only damaged, permeable membranes of nonviable cells. These are then easily identified visually by the presence of blue color (with trypan blue) in a simple brightfield microscope, or by bright fluorescence seen by fluorescence microscopy or flow cytometry.

Fluorescent probes include a wide range of dyes that label DNA of membrane-damaged cells. Tetrabromofluorescein (eosin Y) is a fluorescein derivative that has been used to identify nonviable Candida blastospores (Costantino et al., 1995). In addition, the intracellular penetration of enzymes such as DNase or trypsin can indicate the loss of membrane integrity and thus nonviability (Frankfurt, 1990; Darzynkiewicz et al., 1994; Johnson, 1995). Along the same lines, the penetration of probes for cytoplasmic markers (actin, tubulin, or cytokeratin) has been used to identify cells with damaged plasma membranes (O'Brien and Bolton, 1995).

The most widely used group of fluorescent probes are those that label nucleic acids (for further discussion of nucleic acid stains, see unit 4.3). The most straightforward labeling methods use propidium iodide (PI) or 7-amino actinomycin D (7-AAD) to identify dead cells, which are hundreds or thousands of times brighter than viable cells. Propidium iodide is in widespread use with many mammalian cell types (Jacobs and Pipho, 1983; Massaro et al., 1989; Coco-Martin et al., 1992; Darzynkiewicz et al., 1994; Stewart and Stewart, 1994; O'Brien and Bolton, 1995), bacteria (Vesey et al., 1994; Nebe-von Caron and Badley, 1996), and protozoa (Armstrong et al., 1991; Humphreys et al., 1994). 7-AAD is a useful alternative to PI. Like PI, 7-AAD penetrates only dead cells, but 7-AAD fluorescence is both less intense and at a longer wavelength (~670 nm, versus ~610 nm for PI). These latter two properties make 7-AAD preferable as a viability marker when FITC and PE are used to label surface antigens (Schmid et al., 1992). It has been reported that the dye can be used for fixed cells as well (Fetterhoff et al., 1993); see section on viability assays of fixed cells for further discussion.

SYBR-14, a recently introduced DNA stain from Molecular Probes, penetrates viable cells (Garner et al., 1994). The dye can be used in conjunction with propidium iodide to unambiguously differentiate viable from dead or moribund sperm cells (Garner et al., 1994). Another dye that can differentiate apoptotic and nonapoptotic cells is SYTO 16 (Molecular Probes), of the SYTO series of dyes (Frey, 1995). The dyes YOYO-1 and TOTO-1 (Molecular Probes; both derivatives of thiazole orange) have also been used successfully in viability assays (Becker et al., 1994). As with other membrane exclusion/DNA binding probes, the dyes do not penetrate viable cells, and remain nonfluorescent until they bind to DNA. Variants of these dyes, YO-PRO-1 and TO-PRO-1 (Molecular Probes), have even higher affinities for nucleic acids and have been used successfully with mammalian and bacterial cells (Vesey et al., 1994; O'Brien and Bolton, 1995; see Haugland, 1994, for details of dyes). The use of YO-PRO-1 for the identification of apoptotic cells seems to have the advantage of preserving the proliferation capacity of living cells (Idziorek et al., 1995).

Ethidium bromide (EB) used with low concentrations of acridine orange (AO) identifies normal (AO at high fluorescence level, EB low), early apoptotic (AO low, EB low), and late apoptotic/necrotic (AO low, EB high) cells (Liegler, et al., 1995; Olivier, 1995). Ethidium bromide staining due to loss of membrane integrity identifies the population of necrotic cells; the mechanism of decreased AO staining is not clear but may be related to loss of DNA integrity (Liegler et al., 1995).

 

Probes of Physiological State

Viable cells can be identified directly using fluorescent probes that identify properties of normal cells. Two principle properties are the integrity of the plasma membrane and the presence of metabolic processes. Cell viability can be assessed based on the presence of cytoplasmic esterases that cleave moieties from a lipid-soluble, nonfluorescent probe to yield a charged fluorescent product that is retained within the cell if membrane function is intact. Hence, viable cells are bright and nonviable cells are dim or nonfluorescent. The most common of these probes is fluorescein diacetate (FDA). It has been used with bacteria (Diaper et al., 1992; Diaper and Edwards, 1994; Vesey et al., 1994; Nebe-von Caron and Badley, 1996), protozoa (Armstrong et al., 1991; Humphreys et al., 1994), phytoplankton (Yentsch and Pomponi, 1994), plants (Galbraith, 1994; Brigham et al., 1995; Kodama and Komamine, 1995), and a variety of mammalian cells (Coco-Martin et al., 1992; Darzynkiewicz et al., 1994; Johnson, 1995). Once FDA diffuses into cells, nonspecific esterases in the cell cytoplasm generate free fluorescein. The dye works well in some instances, but the rate at which fluorescein diffuses out of cells varies greatly. To circumvent this problem, dye variants such as BCECF (Molecular Probes) and carboxyfluorescein diacetate, which require energy-dependent efflux of the fluorescent dye (Massaro et al., 1989; Breeuwer et al., 1994), have been developed.

Another approach to minimize dye loss is to use the acetoxymethyl ester of calcein (calcein AM). The ester group facilitates uptake of the dye and is cleaved in the cytoplasm to give free calcein. The fluorochrome has an increased retention time (a 3-hr half-life is reported) and less sensitivity to pH (Haugland, 1994; Holló et al., 1994). Molecular Probes provides the dye as a kit in conjunction with ethidium homodimer (EukoLight). Only live cells retain calcein and are labeled green, while dead cells are labeled red because their permeable membranes allow the ethidium homodimer to penetrate and label DNA. In addition, Molecular Probes has produced a variety of kits for determining viability in bacteria (BacLight) or fungi (FungoLight, FunLight).

Dihydroethidium is taken up by viable cells and cleaved by esterases to generate ethidium monomer, which binds to DNA and is retained in the nucleus (Bucana et al., 1986). Dead cells do not produce the monomer and remain nonfluorescent. Viable intracellular parasites (Babesia bovi) can be identified with flow cytometry (Wyatt et al., 1991). The dye has been used in conjunction with carboxyfluorescein diacetate to identify viable populations of sperm from frozen samples (Ericsson et al., 1989). Questions regarding the potential toxicity of the dye were raised when the dye was found to inhibit the oxygen uptake of sperm cells stained with dihydroethidium (Downing et al., 1991).

Another dye that seems to have promise is Vita Blue (Becton Dickenson; Lee et al., 1989). This dye is excited by red light, thereby making possible the simultaneous use of green and orange emitting dyes that are excited by 488-nm light.

Viable cells maintain electrochemical gradients across the plasma membrane. Functional subsets of this general phenomenon include maintenance of pH and other ion gradients as well as the capacity for energy-yielding metabolism in mitochondria. These physiological processes can be exploited to distinguish viable from nonviable cells. Fluorochromes useful in viability assays include those used to measure membrane potential and intracellular pH. These probes are typically lipophilic, charged molecules that preferentially partition into cells having a negative potential difference with respect to the surrounding environment, so that the dye becomes concentrated in the cytoplasm or internal organelles.

One of the most commonly used probes for identifying viable cells is rhodamine 123. This is a cationic lipophilic dye that partitions into mitochondria because of their low potential. Active mitochondria in viable cells are stained bright green; loss of gradients within the cell results in loss of the dye. Rhodamine 123 can indicate viable cells among bacteria (Diaper et al., 1992; Diaper and Edwards, 1994; Vesey et al., 1994; Porter et al., 1995a, b; Nebe-von Caron and Badley, 1996) or a variety of mammalian cells (Darzynkiewicz et al., 1994; Johnson, 1995). The presence of aliphatic side chains on fluorochromes may strongly influence retention within the cells. For example, side chains of the cyanine dye DiOC₆(3) cause the probe to partition into the mitochondrial membranes, resulting in increased concentration and brightness—the brightness related to better fluorescence in a lipid environment (Sims et al., 1974). A possible disadvantage is that increased affinity for membranes may also slow the loss of the dye if mitochondria lose their metabolic capacity.

Any of the assays employing dye uptake to label viable cells can be coupled with assays using dyes to label nonviable cells. Thus, one may use rhodamine 123 with, for example, PI or ethidium bromide.

Instead of viable cells being labeled with cationic dyes, nonviable cells can be labeled with the lipophilic anionic dye oxonol; this has been done in bacteria (Deere et al., 1995; Nebe-von Caron and Badley, 1996) and protozoa (Humphreys et al., 1994). Loss of negative potential with respect to the outside causes accumulation of oxonol within dead cells. It has been reported, however, that starved bacteria can contain populations of oxonol-positive bacteria that are PI-negative (Nebe-von Caron et al., 1996).

In some cells, dyes that accumulate in live cells may be pumped out by the glycoprotein pump (Holló et al., 1994; Shapiro, 1995). Thus, cells that do not stain brightly with a dye such as rhodamine 123 may still be viable. It has also been reported that some nonspecific staining of rhodamine 123 in natural particulate environments may complicate the identification of live bacteria (Porter et al., 1995b). Conversely, cells in the presence of glutathione may have hyperpolarized mitochondria and hence enhanced uptake of dye into mitochondria (Pieri et al., 1992). Valinomycin has been used to hyperpolarize bacteria to enhance their dye uptake (Porter et al., 1995b). There is some indication that the toxicity of rhodamine appears to be low (Downing et al., 1991).

 

Probes for Fixed Cells

Often cells must be fixed or permeabilized before analysis. In both cases analysis of other markers is less ambiguous if nonviable or damaged cells can be identified and rejected. Several methods have been successful in the labeling of dead or damaged cells prior to fixation. A modification of Hoechst/PI staining methods permits the use of ethanol as a fixative (Pollack and Cianco, 1990). Prior to fixation, dead cells are labeled with PI. After ethanol fixation, all cells label with Hoechst, but the PI in dead cells quenches Hoechst fluorescence; viable cells are reported to have insignificant PI fluorescence. Other DNA probes that penetrate and stain dead or damaged cells brightly are ethidium monoazide (EMA) and laser dye styryl-751 (LDS-751).

Ethidium monoazide is positively charged and penetrates the membranes of dead or damaged cells but not live cells. EMA intercalates into DNA and can be photochemically cross-linked with short exposure to visible light (Riedy et al., 1991). In contrast, LDS-751 penetrates both damaged and live cells, but labels the DNA of damaged or dead cells much more brightly (Terstappen et al., 1988).

A novel use of 7-AAD for labeling nonviable fixed cells involves addition of a molar excess of actinomycin D (AD) while cells are being fixed (Fetterhoff et al., 1993). It is thought that the higher concentration of AD prevents binding of 7-AAD to viable cells; for the short term, nonviable cells retain bound 7-AAD.

A different approach is based on the penetration of large molecules into dead or damaged cells. Streptavidin-Tricolor (SA-TR; Caltag) irreversibly penetrates nonviable cells, staining them bright red; subsequent washing, fixation, and permeabilization do not result in significant dye loss (Levelt and Eichmann, 1994). The mechanism for this retention is not clear. This technique permits prelabeling damaged cells prior to labeling intracellular antigens.

 

Methods for Microscopy

In determining the viability of cells, one should not overlook the obvious. Much information can be obtained by direct observation of the cells in a microscope. In many cases, obviously misshapen or bloated cells or cells that have lost refractility in phase contrast indicate severe problems that obviate more sophisticated approaches. In other cases, morphological changes are very useful in following physiological processes. For example, certain morphological changes are hallmarks of apoptosis (Darzykiewicz et al., 1994). In plant cells, changes in shape are easily detectable, and the loss of metabolically driven processes such as cytoplasmic streaming can indicate the loss of viability (Brigham et al., 1995).

Assessment of cell viability under the microscope can be accomplished with stains that mark nonviable cells. Dyes include the very common trypan blue (McGahon et al., 1995), nigrosin (Johnson, 1995), and erythrosin B (Bochner et al., 1989). Viable cells have intact membranes and exclude the dyes. Nonviable cells are labeled and are visible with brightfield optics.

Fluorescent probes in common use for flow cytometry can also be used in the microscope. These include FDA (Humphreys et al., 1994), YO-PRO-1 (Idziorek et al., 1995), and dihydroethidine (Bucana et al., 1986).

 

Critical Parameters and Troubleshooting

 

Using PI or 7-AAD with Phycoerythrin (PE)

Although the emission spectrum of PE overlaps the shorter-wavelength end of PI and 7-AAD emission spectra, PE can be used in conjunction with PI or 7-AAD for dead cell discrimination. Using a gate for PI- or 7-AAD-negative cells allows for the examination of PE label on presumed viable cells.

It is important that there be appropriate compensation between PE and PI or 7-AAD detectors, as both fluorochromes will be detected by both detectors. (For illustrations of compensation, see Fig. 9.2.2 and comments in Alternate Protocol 2.) To get compensation for dual-labeled cells, prepare a tube without PI or 7-AAD. Adjust the compensation of PE into the PI/7-AAD channel such that PE-positive cells are in the range of PI- or 7-AAD-negative or viable cells (see Fig. 9.2.2C).

If expected nonviable cells are not found, demonstrate the efficacy of the dye by heating. Take one tube of cells ready for analysis and heat 10 min in a 45°C water bath. Cool to room temperature and reanalyze. All cells are now nonviable and should be a brighter red.

7-AAD should also be usable with dyes that emit in the longer-wavelength region if they are excited by a second laser that is not colinear with the 488-nm laser. For example, a helium-neon (HeNe) laser emitting at 633 nm, a krypton (Kr) laser emitting at 647 nm, or a diode laser operating in the same region would not excite 7-AAD. One may also be able to use a HeNe laser that is colinear with a 488-nm beam if the filters in front of the PMT that detects 7-AAD emission exclude the 633-nm laser line.

 

Probes of Physiological State

Cells must be kept under optimal conditions for assays that reflect their physiological state. That is, some cells may survive well in PBS or HBSS, but others may require serum supplementation or other factors to remain healthy.

Because FDA can leak from cells, it is important to analyze immediately after the incubation period. When FDA is used together with cell surface markers labeled with PE, the FDA fluorescence overlaps the PE emission range. Thus, setting proper compensation is crucial. Compensation should be checked with samples labeled with each fluorochrome.

The glycoprotein pump in some cells may pump rhodamine 123 out of live cells (Holló, 1994; Shapiro, 1995). Thus, cells that do not stain brightly may be viable. When in doubt, try an alternate reagent for nonviable cell identification, such as PI or 7-AAD.

 

Probes for Fixed Cells

Although EMA emission can be used in conjunction with surface-labeled antibodies, the emission of PE overlaps that of EMA; hence, appropriate compensation is required (see Alternate Protocol 2 for further discussion of compensation). LDS-751 stains all nucleated cells, so all will be positive. Dead or damaged cells are about ten-fold brighter than viable cells. If nonviable cells are present, then two populations should be present; mature red cells will be negative (see Terstappen et al., 1988, for illustrations). Artifacts in cell surface labels have been observed in formaldehyde-fixed cells labeled with LDS-751 (McCarthy et al., 1994). If alteration of the surface labeling pattern is suspected, compare the staining pattern of unfixed, labeled cells stained with LDS-751.

 

Methods for Microscopy

After staining, count cells within 5 min. On sitting, some viable cells may become permeable and take up dye, appearing as false nonviables.

 

Anticipated Results

Probes for membrane integrity. Nonviable cells are stained bright red and viable cells are nonfluorescent. With PI, there should be about a 2-log difference in brightness between viable and nonviable cells; 7-AAD-positive cells are dimmer. Discrimination of viable cells, nonviable cells, and debris can be done easily on a bivariate plot of forward scatter (typically linear scale) versus log PI or 7-AAD fluorescence (see Fig. 9.2.1).

Probes of physiological state. FDA and rhodamine 123 are both excited by 488-nm light and fluoresce green. Filters used for measuring fluorescein (e.g., 530 ± 20–nm band-pass) are adequate to detect fluorescence of either dye. With logarithmic amplification of the fluorescence signal, viable FDA-positive cells are very bright, but viable rhodamine 123–positive cells will be dimmer. A bivariate plot of forward scatter versus rhodamine 123 fluorescence may make it simpler to distinguish rhodamine 123–positive viable cells from nonviable cells (see Fig. 9.2.3B).

Probes for fixed cells. Because the fluorescence of EMA is not as bright as that of propidium iodide, discrimination of nonviable, EMA-bright cells from viable cells may be less obvious. A bivariate plot of forward light scatter versus EMA fluorescence may aid in distinguishing viable (EMA-negative) cells. When stained with LDS-751, viable cells can be identified on a bivariate plot of light scatter and red fluorescence as a population of intermediate brightness. Dead or damaged cells stain more brightly, and enucleate red blood cells are unstained.

Methods for microscopy. The procedure is very straightforward, but phase contrast optics can aid in the identification of viable cells that do not stain with trypan blue. Nonviable cells are blue and phase dense; viable cells are phase bright.

 

Time Considerations

Probes for membrane integrity. Cell staining should take <5 min with PI or 30 min with 7-AAD. Instrument setup (and compensation, if employed) should take 5 min. If data are collected as listmode files, then subsequent analysis may require another 5 to 10 min depending on the complexity of the gating and the degree of automation of the analysis software employed.

Probes of physiological state. Cell staining should take 5 min with rhodamine 123 or 15 min with FDA. Instrument setup (and compensation, if employed) should take 5 min. If data are collected as listmode files, then subsequent analysis may require another 5 to 10 min depending on the complexity of the gating and the degree of automation of the analysis software employed.

Probes for fixed cells. EMA labeling of cells takes ~10 min to stain and cross-link the dye, followed by ~30 mins for washing and fixing. Labeling of cells with LDS-751 takes ~20 min for staining and incubation. Instrument setup (and compensation, if employed) should take 5 min. If data are collected as listmode files, then subsequent analysis may require another 5 to 10 min depending on the complexity of the gating and the degree of automation of the analysis software employed.

Methods for microscopy. Staining and counting cells can be done in 5 to 10 min.