Dr O. Hérault Laboratoire d'Hématologie, JE 1993, Faculté de Médecine, 2 bis boulevard Tonnellé, 37032 Tours Cédex, France.
A recently reported cytometric method described the possibility of discriminating apoptotic from necrotic cells using FITC-labelled annexin V and propidium iodide (PI). Nevertheless, the brightness of PI-staining and its extensive spectral emission overlap with phycoerythrin (PE) does not permit the study of a subset of a heterogenous cell population with single laser instrumentation. The surface staining of a subset with PE in a heterogenous cell population therefore requires another exclusion dye to detect necrotic cells. We used 7-amino-actinomycin D (7-AAD) that can be excited by the 488 nm argon laser line. 7-AAD emits in the far red range of the spectrum and 7-AAD spectral emission can be separated from the emissions of FITC and PE. The fluorescence parameters allow characterization of necrotic (7-AAD+ annexin V-FITC+ cells), apoptotic (7-AAD− annexin V-FITC+ cells) and viable cells (7-AAD− annexin V-FITC− cells) in a subset of PE+ cells. The value of this method was demonstrated by measuring apoptosis and necrosis in a model of HL-60 cells exposed to different inducers of cell death. The method was validated by fluorescent cell sorting in combination with morphologic examination of the sorted cells. The technique we present is particularly valuable in a clinical setting because it enables rapid multiparameter analysis of necrosis and early apoptosis in combination with cell surface phenotyping with a single laser. We present the effects of haemopoietic growth factor deprivation on myeloid progenitor CD34+ cells as an example of its application.
Apoptosis is an important mechanism for the selective elimination of mammalian cells which is distinct from the process of cell death by necrosis (Wyllie et al, 1984). A recently reported flow cytometric method (Vermes et al, 1995) enables discrimination of these two cell death processes using annexin V and propidium iodide (PI). Nevertheless it does not permit the study of a subset of a heterogenous cell population with a single laser.
The anticoagulant annexin V is a member of a family of proteins that are structurally related and exhibit Ca2+-dependent phospholipid-binding properties. Annexin V can bind to various phospholipid species and shows its highest affinity for phosphatidylserine (Andree et al, 1990). In normal blood cell phosphatidylserine is situated on the inner layer of the plasma membrane. When cell death occurs, phosphatidylserine is translocated to the outer layer of the membrane, i.e. the external surface of the cell (Fadok et al, 1992). This occurs in the early phase of apoptosis during which the cell membrane itself remains intact. Thus, fluorescein isothiocyanate (FITC)-labelled annexin V is usually used for the quantification of apoptotic cells (Homburg et al, 1995; Koopman et al, 1994). Necrosis is a form of cell death that differs from apoptosis (Kerr et al, 1972; Majno & Joris, 1995; Wyllie et al, 1984) and includes the loss of cell membrane integrity. Measurement of annexin V binding, performed simultaneously with a dye exclusion test, is commonly used to detect early apoptotic cells (Vermes et al, 1995).
The exclusion test is performed with fluorescent nonvital dyes forming fluorescent complexes with the DNA which emit fluorescence in the orange range of the spectrum. PI has been widely utilized in assays using annexin V FITC. Nevertheless, the brightness of its staining and its extensive spectral emission overlap with phycoerythrin (PE) (Schmid et al, 1992). Thus, in a heterogenous cell population, surface staining of a subset with PE requires another exclusion dye to detect necrotic cells.
Actinomycin is a molecule of very high molecular weight with two penta-peptides attached to the phenoxazone fluorophore that is non-charged at neutral pH (Gill et al, 1975). 7-Amino-actinomycin D (7-AAD) is a G-C base-specific DNA intercalator (Cowden & Curtis, 1981) that can be excited by the 488 nm argon laser line (Zenelin et al, 1984) and emits in the far red range of the spectrum. 7-AAD spectral emission (λ emmax: 655 nm) can be separated from the emissions of FITC (λ emmax: 517 nm) and PE (λ emmax: 578 nm) and this renders it suitable for dead cell discrimination (Schmid et al, 1992).
The visualization and correlation of all fluorescence parameters are simultaneously collected on the cell preparation to characterize necrotic (7-AAD+ annexin V-FITC+ cells), apoptotic (7-AAD− annexin V-FITC+ cells) and viable cells (7-AAD− annexin V-FITC− cells) in a subset of PE+ cells. 7-AAD has been previously used in combination with reagents conjugated to FITC or PE for cell cycle analysis or dead cell discrimination of subsets of activated cells (Jordan et al, 1996; Schmid et al, 1992). However, 7-AAD has not been used to date with annexin V-FITC for detection of early apoptotic cells. We document here the successful utilization of this method (with a single laser) to discriminate these cells in heterogenous populations.
The value of this method was shown by measuring apoptosis and necrosis in a model of HL-60 cells exposed to inducers of cell death. The results were correlated with the typical changes in cellular morphology detected on the basis of light scatter measurements. This method was validated by fluorescent cell sorting in combination with morphologic examination of the sorted cells. The technique we present is particularly valuable in a clinical setting because it enables rapid multiparameter analysis of necrosis and early apoptosis in combination with cell surface phenotyping with a single laser. We applied it following the appearance of apoptosis on normal marrow CD34+ cells deprived of haemopoietic growth factors.
MATERIALS AND METHODS
The human myeloid cell line HL-60, kindly provided by Professor C. Chomienne (St Louis Hospital, Paris, France), was cultured in RPMI 1640 medium with 20 μML-glutamine (Gibco/BRL, Grand Island, N.Y., U.S.A.) supplemented with 15% heat-inactived fetal bovine serum (Gibco/BRL), 100 units/ml penicillin G and 100 μg/ml streptomycin (Boehringer-Mannheim, Mannheim, Germany). Cultures were maintained under a fully humidified atmosphere of 95% air/5% CO2 at 37°C and were passaged twice weekly. Exponentially growing cells were used in all experiments. Cell viability was assessed by the ability to exclude ethidium bromide (Sigma, St Louis, Mich., U.S.A.). CD34+ bone marrow cells were obtained from bone marrow aspirates of 10 healthy donors who had given informed consent. Mononuclear cells were separated by centrifugation over a Ficoll-Hypaque gradient and enriched in CD34+ cells by direct immuno-separation using magnetic beads (Dynal CD34 Progenitor Cell Selection SystemTM, Dynal A.S., Oslo, Norway) according to the manufacturer's instructions. CD34+ cells were incubated at a concentration of 2 × 104 cells/ml for 7 d at 37°C, 5% CO2 with or without stem cell factor (SCF) (10 ng/ml, kindly provided by Amgen, Thousand Oaks, Calif., U.S.A.) in a serum-free medium consisting of Iscove's medium supplemented with 1% deionized bovine serum albumin (Sigma), 0.3 mg/ml human transferrin (Boehringer-Mannheim), 7.8 μg/ml cholesterol (Sigma), 30 μg/ml L-α-phosphatidyl cholin (Sigma), 0.25 U/ml human insulin (OrgasulineTM, Organon, St Denis, France) and 0.1 mM 2-mercaptoethanol (Sigma).
Induction of apoptosis and necrosis
Necrosis of HL-60 cells was induced by exposure to 50°C heat shock for 90 min in RPMI (Martin et al, 1995) and apoptosis was induced by exposure to camptothecin (Sigma), a topo-I-inhibitor, dissolved in dimethylsulphoxide (DMSO) (Sigma). Camptothecin stock solution was diluted to ensure a final concentration of <0.03% for DMSO (Palissot et al, 1996). Cells were seeded at 2 × 105 cells/ml in medium with camptothecin at 1 μM. Cells were then returned to a 5% CO2 incubator at 37°C for 1, 2 and 4 h. Control cultures were treated with an equivalent volume of DMSO in RPMI which did not induce apoptosis or differentiation of HL-60 cells. Haemopoietic cells rapidly die in the absence of appropriate haemopoietic growth factors (HGF) (Williams et al, 1990; Zauli et al, 1994), which directly promote myeloid CD34+ cell survival by suppressing apoptosis. In these experiments selected CD34+ cells were cultured in a serum-free medium and their survival was promoted by SCF as previously described (Keller et al, 1995; Zauli et al, 1994).
HL-60 cell surface antigens were stained by addition of 10 μl of PE-conjugated anti-CD13 monoclonal antibody (mAb) (PE-conjugated SJ1D1, Immunotech, Marseille, France) in 400 μl of phosphate-buffered saline (PBS without Ca2+ and Mg2+, BioMérieux, Marcy l'Etoile, France) to 106 cells followed by incubation for 30 min at 4°C. Background staining was determined by incubation of cells with 10 μl mouse IgG1 PE (Dako, Glostrup, Denmark). The supernatant was removed after two washes with PBS by centrifugation for 5 min at 250 g, and the cell pellet was resuspended in Ca2+-binding buffer. Then 5 μl of annexin V- FITC (Immunotech) and 1 μg of 7-AAD (Sigma) were added, followed by incubation for 10 min at 4°C in the dark. The feasibility of apoptosis characterization in a minor subpopulation was demonstrated using camptothecin-exposed HL-60 cells labelled with anti-CD13 PE (5%), mixed with unlabelled and untreated HL-60 cells (95%). These mixed cells were stained using annexin V-FITC and 7-AAD. Selected CD34+ cells were stained with PE-conjugated anti-CD34 mAb (PE-conjugated-HPCA2 mAb, Becton Dickinson, San Jose, Calif., U.S.A.). Immediately after cell selection and after 3 d and 7 d liquid culture, labelled CD34+ cells were stained using annexin V-FITC and 7-AAD as previously described. Concomitantly, the PE-conjugated anti-CD41 mAb (PE-conjugated 5B12 mAb, Dako) was used with annexin V-FITC and PerCP-conjugated anti-CD34 mAb (PerCP-conjugated HPCA2 mAb, Becton Dickinson) as previously described. Unlabelled fixed (with 1% paraformaldehyde) cell preparation was used to verify the annexin V-FITC positivity.
Samples were analysed on a FACScaliburTM flow cytometer (Becton Dickinson) equipped with a 15 mW air-cooled 488 nm argon-ion laser. The green fluorescence was collected after passing through a 530/30 nm band pass (BP) filter. PI and PE emissions were detected by filtration through a 585/42 nm BP filter. 7-AAD emission was collected after passing through a 650 long pass filter. This standard filter combination is provided by Becton Dickinson and cannot be changed by the user. Electronic compensation was used in the fluorescence channels to remove residual spectral overlap. Spectral compensation between FL1 (FITC channel) and FL2 (PE channel) was set using cells exposed to camptothecin and then labelled with annexin V-FITC alone. Cells labelled with anti-CD13 PE or anti-CD34 PE alone were used to eliminate PE fluorescence entering FL1 and FL3 (7-AAD channel). An unlabelled fixed (with 1% paraformaldehyde) cell preparation stained with 7-AAD was used to compensate 7-AAD fluorescence entering FL2. Fluorescence data were displayed on a four-decade log scale. A minimum of 3 × 104 events was collected on each sample. Analysis of the multivariate data was performed with CELLquestTM software (Becton Dickinson).
Cells were sorted using the FACScaliburTM. Sorted cells were collected into 50 ml conical centrifuge tubes precoated with bovine serum albumin as recommended by Becton Dickinson. Cells were concentrated by centrifugation, most of the supernatant was removed, and the cells were resuspended in the remaining sheath fluid. Aliquots of this suspension were processed for morphologic analysis.
Before and after camptothecin or heat shock exposure, sorted HL-60 cells were centrifuged at 6000 rpm using a cytocentrifuge (CytospinTM, Shandon-Southern, Eragny, France). The cells were fixed in 100% methanol, stained with May-Grünwald-Giemsa (MGG) and examined for morphological and nuclear changes consistent with apoptosis (Arends et al, 1990; Del Vecchio et al, 1991; Sarraf & Bowen, 1988).
Identification of apoptotic and necrotic cells
Flow cytometry analysis of HL-60 cells treated with camptothecin showed that apoptotic cells can be distinguished among CD13-PE-labelled cells by the appearance of an annexin V+ peak in 7-AAD− cells. After 4 h of incubation with camptothecin, >20% of the cells within the gated region (CD13+ 7-AAD−) were stained by annexin V-FITC (Fig 1). Changes in the morphology of cells undergoing apoptosis affected their light scattering properties. Due to condensation and nuclear fragmentation, apoptotic cells usually produce lower forward scatter and higher side scatter signals than viable cells (Darzynkiewicz et al, 1992). Fig 2 shows that apoptotic CD13+ cells obtained after camptothecin treatment had a lower FSC signal, in contrast to necrotic cells. 7-AAD allowed characterization of necrotic cells after 50°C heat shock exposure as demonstrated by the appearance of an annexin V+ 7-AAD+ peak in the CD13+ population. No annexin V+ 7-AAD− cells were found as heat shock induced the prompt loss of the membrane integrity, without initial apoptosis followed by secondary necrosis.
It was possible to discriminate the apoptotic/necrotic status of each subpopulation using annexin V-FITC and 7-AAD in a heterogenous cell suspension of camptothecin-exposed HL-60 cells labelled with anti-CD13-PE mAb (5%) mixed with unlabelled and untreated HL-60 cells (95%) (Fig 3). The proportion of apoptotic cells in the gated CD13+ cell subpopulation (camptothecin treated) was increased in comparison with untreated CD13− cells: 23% and 6% respectively.
The same type of analysis on haemopoietic growth factor-deprived CD34+ cells (Fig 4) showed a massive decrease in strictly viable CD34+ cells (annexin V− 7-AAD−), whereas CD34+ cell survival was promoted by SCF, indicated by a decrease in CD34+ annexin V+ cells. As annexin V adheres to activated platelets, we also studied CD34+ CD41+ cells. All the CD34+ annexin V+ cells were involved in the apoptotic/necrotic process: although a few CD34+ CD41+ cells were found (exclusively after CD34+ cell selection) indicating platelet adhesion or the presence of megakaryocytic progenitors, no annexin V+ cells were observed in this cell population (Fig 5).
Analysis of changes in cellular morphology
Apoptotic cells are characterized by condensation or fragmentation of the nucleus. MGG staining of cytospin preparations from sorted HL-60 populations confirmed that CD13+ annexin V− 7-AAD−, CD13+ annexin V+ 7-AAD− and CD13+ annexin V+ 7-AAD+ cells had very distinct morphological features which represented viable, apoptotic and necrotic states respectively (Fig 6). The apoptotic state included nuclear condensation and fragmentation, and finally cell fragmentation, whereas necrotic cells did not exhibit condensed chromatin or fragmented nuclei but the shapes of the nuclei were diffuse and irregular, associated with early loss of cell membrane integrity.
Many attemps have been made to find a protocol capable of discriminating the apoptotic stage and the phenotype in homogenous and heterogenous cell populations using a dual laser cytometer (Chrest et al, 1993; Swat et al, 1991) or fixed cells (Homburg et al, 1995). We propose here a rapid and simple method to study unfixed cells using a single laser cytometer.
The methods used to discriminate cells with annexin V-FITC staining also used PI to identify the dead cells. Unfortunately, PI emits a red-orange fluorescence, limiting the use of a single laser cytometer for the discrimination of other parameters such as surface antigens. We set up a cytometric method that enables detection of the presence of antigens in apoptotic subpopulations. As this method does not require permeabilization of the cell membrane, concurrent staining for cell surface antigens is possible, thus permitting the use of this method in heterogenous populations. The use of unfixed cells avoids problems related to antigen loss or accessibility after the fixation process.
Using 7-AAD, which emits only in the red region of the spectrum when excited by the blue light of an argon laser tuned to the 488 nm wavelength, in combination with annexin V-FITC, we were able to distinguish apoptotic cells (7-AAD− annexin V+) from late apoptotic or necrotic cells (7-AAD+ annexin V+) among HL60 cells triggered to apoptosis or necrosis using camptothecin or heat shock, respectively. It was possible to stain HL-60 cells simultaneously with monoclonal antibodies labelled with PE directed against different surface antigens.
This approach allows the discrimination of early apoptotic cells in heterogenous cell populations and the evaluation of the level of expression of different surface antigens in live and apoptotic cells. It is thus possible to follow the development of apoptosis using others parameters (e.g. scatter signal) (Darzynkiewicz et al, 1992) and stage by stage to check the presence of certain antigens on apoptotic subpopulations during the whole apoptotic process. For instance, a decrease in some cell surface antigens such as human lymphocytes CD45 antigen (Carbonari et al, 1994), thymocyte CD4 and CD8 antigens (Carbonari et al, 1994; Swat et al, 1991) or various neutrophil antigens (Dransfield et al, 1994, 1995) has been seen in apoptotic cells.
As 7-AAD was utilized for dead cell exclusion, a low dye concentration (1 μg/ml) was used for 10 min to minimize potential dye uptake by live cells. This low 7-AAD concentration gave good resolution of late apoptotic and necrotic cells from cells with intact membranes, and with annexin V-FITC early apoptotic cells were distinguished from necrotic cells and viable cells. The use of 7-AAD alone was recently described to identify apoptosis (Philpott et al, 1996; Schmid et al, 1994), but using 20–200-fold greater concentrations of 7-AAD for 20 min. In these experimental conditions the authors described increased 7-AAD fluorescence of apoptotic cells, perhaps because of alterations in membrane physiology occurring during apoptosis (e.g. changes in membrane transport pumps). 7-AAD, which is a cytotoxic agent, may at high doses interfere with membrane physiology and the apoptotic process. Conversely, the two-parameter flow cytometric method described in this paper permits the detection of the early phases of apoptosis before the loss of cell membrane integrity.
We chose HL-60 cells for this study because of their characteristic easily detectable apoptotic morphology (Blair et al, 1985; Lennon et al, 1991; Martin et al, 1990). We verified that the induction of apoptosis by camptothecin and of necrosis by heat shock could be evaluated specifically by the combination of light-scatter measurement with both 7-AAD and annexin V-FITC on the PE-conjugated anti-CD13 mAb stained cells. The modifications observed were correlated with the FSC properties of apoptotic and necrotic cells (Darzynkiewicz et al, 1992, 1997). The microscopic examination of cell morphology, representing one of the most reliable parameters for confirmation of apoptotic changes, confirmed the flow cytometry results after cell sorting for apoptotic and necrotic states. We extended the experiment with 7-AAD and annexin V-FITC staining on normal human cells, using myeloid CD34+ progenitors stained with PE-conjugated anti-CD34. We demonstrated the opposite effects of SCF and deprivation of haemopoietic growth factors on apoptosis of such cells using the above method.
In conclusion, our results demonstrated the possibility of discriminating apoptotic and necrotic cells in a heterogenous cell population using a single laser. This method, which is quick, simple and reproducible, will be a valuable tool to study the role of apoptosis in cells such as normal and malignant haemopoietic cells.
We are grateful to Professor Michel Marchand for providing samples from patients undergoing cardiac surgery. We thank Marie-Thérèse Georget and Nathalie Héraud for their technical assistance. This work was supported by grants from the Conseil Régional de la Région Centre (grant number 99500 R3 0005), the Comités du Cher et de l‘Indre-et-Loire de la Ligue Nationale contre le Cancer and the Ministère de l'Enseignement Supérieur et de la Recherche.