Real-time flow cytometry for the kinetic analysis of oncosis


  • G. Warnes,

    Corresponding author
    1. Flow Cytometry Core Facility, The Blizard Institute of Cell and Molecular Science, Barts and The Royal London School of Medicine and Dentistry, London University, London E1 2AT, United Kingdom.
    • The Flow Cytometry Core Facility, The Blizard Institute of Cell and Molecular Science, Barts and The Royal London School of Medicine and Dentistry, London University, 4 Newark Street, London E1 2AT, United Kingdom
    Search for more papers by this author
  • S. Martins

    1. Flow Cytometry Core Facility, The Blizard Institute of Cell and Molecular Science, Barts and The Royal London School of Medicine and Dentistry, London University, London E1 2AT, United Kingdom.
    Search for more papers by this author


The standard method of distinguishing apoptotic and oncotic cells has been by microscopic analysis of nuclei and cell membrane morphology. Thus a rapid test for analyzing large numbers of cells in the study of cell necrobiology has not been possible until the recent advent of the Amnis Image-stream and real-time Lab-on-a-Chip technologies. An interesting difference between apoptosis and oncosis is that they are ATP dependent and independent processes, respectively. Here we describe an assay measuring real-time kinetic changes in the potential differences of the inner mitochondrial membrane (mmp) and the plasma membrane (pmp) in cells immediately before and after the addition of the inducing agent. Live cells were loaded with carbocyanine dye DiIC1(5) and bis-oxonol (DiBAC4(5)) to measure mmp and pmp in conjunction with annexin V-FITC and DAPI labeling for gating out annexin V binding cells and dead cells respectively. Live cells gave specific membrane signatures in response to apoptotic or oncotic reagents in real-time. Apoptosis showed little change in mmp and pmp signals over the course of 25 min, the mitochondria only showed a slight hyperpolarization. In contrast chemical treatment with oxidative phosphorylation blocker, sodium azide (SA) caused an immediate hyperpolarization spike followed by a complete abrogation of mmp over a 25 min time course. Treatment with SA (1%) also caused plasma membrane depolarization. Likewise detergent (0.01% Triton X-100) treatments also caused abrogation of mmp and depolarization of pmp. Whereas heat shock (42°C) treatment showed only a slight mitochondrial membrane potential depolarization. These flow cytometric observations were confirmed by confocal microscopy. This novel real-time kinetic assay measuring mitochondrial and plasma membrane potential changes has important implications in the field of cell necrobiology in that it allows the researcher to differentiate apoptotic and oncotic processes in an immediate manner for the first time. © 2011 International Society for Advancement of Cytometry

“Cell Necrobiology” is defined as “various modes of cell death: the biological changes, which predispose precede and accompany cell death; as well as the consequences and tissue response to cell death” (1–3), this therefore includes the broad range of modes of cell death including classic apoptosis, caspase-independent apoptosis-like programmed cell death (PCD), autophagy, mitotic catastrophe, necrosis-like PCD, necrosis or accidental cell death and cell senescence. Flow cytometry assays have been used to study classic apoptosis (4–6) and to a more limited extent in recent years cell necrosis by accidental cell death (7). This study takes a fresh look at classic necrosis using a standard flow cytometric approach employing a combination of multiplexed assays and a real-time assay to the study of accidental cell necrosis.

Necrosis or more specifically dead cells derived from oncosis is characterized as accidental cell death as a result of a gross environmental insult, causing the release of inflammatory contents after the cell and organelles increase in size (1–3, 8, 9). This is in contrast to cell death resulting from apoptosis in which an active or programmed process results in cell shrinkage by nuclear and cytoplasmic condensation and fragmentation, membrane blebbing and the formation of apoptotic bodies. Necrosis is the term used when cells lose membrane integrity no matter how the cell reached that point, so there are two types of necrotic cells those derived from apoptosis or oncosis (9). Interestingly oncosis is an ATP independent process whilst apoptosis is an ATP dependent process (10).

Until recently the flow cytometry community has searched for a means to distinguish between cells undergoing apoptosis or oncosis by standard flow cytometry (6, 11–14). There are a plethora of assays for the study of apoptosis several hours after induction, by measurement of mitochondrial membrane potential, caspases, annexin V binding through to cell death or necrosis as measured by lack of viability and DNA fragmentation. These assays include the use of Hoechst 33342 (Ho33342) and propidium iodide (PI; 15–17) and annexin V binding to externalized phosphatidylserine (PS) in the presence of a viability dye such as PI to show the presence of apoptotic and necrotic cells derived from apoptosis (4). Classically caspases become activated during apoptosis; this activation can be detected by fluorogenic caspase peptide substrates (PhiPhiLux) and Fluorochrome-Labeled Inhibitors of Caspases or FLICA, which become fluorogenic when acted on by active caspases (18, 19). These fluorescence signals resulting from activation of caspases can be detected flow cytometrically and multiplexed with annexin V, mitochondrial potential and reactive oxygen dyes further subdivisions of apoptosis and necrosis can be studied (5, 6, 20–22). Another approach has been the use of SYTO dyes, originally thought of as nucleic acid binding dyes that could be used to detect apoptosis (23–26). SYTO dyes have been used as a cheaper replacement for annexin V (7, 26). Interestingly, SYTO16 was shown not to significantly reduce during sodium azide or heat shock induction of oncosis, the cells rapidly lost membrane integrity, a classic profile of oncosis (7). A few earlier studies have tried to find a means to show a difference between cells undergoing apoptosis and oncosis without much success by standard flow cytometry (12, 13, 22). One study by Waring et al. (11) has shown that thymocytes undergoing oncosis have a higher level of annexin V binding than cells undergoing apoptosis, however this is not the case for all cell lines (12). Interestingly changes in mitochondrial function at hourly time points have been reported in cells undergoing apoptosis and necrosis by the use of Etoposide and sodium azide (30 mM or 0.2%; 22). However it was concluded that there were no differences in changes in mitochondrial function in cells undergoing apoptosis or oncosis. Measurement of mitochondrial function was reported more recently to show differences in cells undergoing apoptosis and oncosis in a drug dose response study after 24 h exposure (13). Although most earlier studies concluded that it is not possible to distinguish between cells generated from an apoptotic or oncotic process the recent use of SYTO16 does afford a means to show differences in cells undergoing apoptosis and oncosis by standard flow cytometry in a time dependent manner (7, 12, 22).

More recently a major advance in the real-time study of apoptosis developed by Wlodkowic and colleagues has shown that a range of viability dyes, including PI and SYTO16 are not toxic to cells and can be used in Lab-on-a-Chip-based real-time imaging systems in the study of the induction of apoptosis by real-time fluorescence and morphological analysis (2, 27–29). In a similar vein a polarity-sensitive annexin-based biosensor (pSIVA) with switchable fluorescence states allows the real-time detection of apoptotic cells by time-lapse imaging from 5–42 h (30). Here we describe a standard flow cytometric assay that distinguishes between apoptosis and oncosis based on real-time kinetic measurement of potential differences of the plasma and mitochondrial membranes directly after the addition of agonists by use of the fluorescent dyes bis-oxonol and carbocyanine dye, DiIC1(5) (31).

Necrotic cells derived from apoptosis and oncosis were also further investigated by use of bis-oxonol in conjunction with DNA quantification and confocal microscopy to determine if it was possible to distinguish between early and late necrosis.


Cell Lines

Jurkat T-cell line were grown in RMPI-1640 with L-Glutamine (Cat No 21875-034, Invitrogen, Paisley, UK) supplemented with 10% Foetal Bovine Serum (FBS, Cat No 16000-044, Invitrogen, Paisley, UK) and penicillin and streptomycin (Cat No 15140-122, Invitrogen, Paisley, UK) in the presence of 5% CO2 at 37°C.

Induction of Cell Death

Jurkat cells were treated with 1 μM Staurosporine (STS; Cat No S6942, Sigma Chemicals, Poole, UK) for up to 30 h to induce apoptosis. Oncosis was induced by various approaches, first by heat treatment at 42°C for 1.5 h and then incubated at 37°C for up to 30 h; or with 1% Sodium Azide (Cat No S8032 Sigma Chemicals, Poole, UK) for up to 30 h; or with 0.01% Triton X-100 (Cat No X-100, Sigma Chemicals, Poole, UK) up to 30 h. Time points analyzed were 0,1, 2, 3, 4, 5, 20, 24, and 30 h, (n = 3) see cell labeling section below.

Cell Labeling

Before or after induction of apoptosis or necrosis cells were loaded with DiBAC4(5) bis-oxonol (100 nM, Cat No B436, Invitrogen, Paisley, UK) and carbocyanine dye DiIC1(5) (40 nM, Cat No H14700, Invitrogen, Paisley, UK) by incubating cells with dyes for 15 min at 37°C. Cells were then washed in PBS buffer (Cat No D8537, Invitrogen, Paisley, UK) and resuspended in 100 μl calcium-rich buffer with Annexin-FITC (2.5 μl) (Cat No 556547 Becton Dickinson, San Jose, CA). Cells were then incubated at Room Temperature (RT) for 15 min. DNA viability dyes, DAPI (200 ng/ml) (Cat No D9542, Sigma Chemicals, Poole, UK) or 7-Amino-actiomycin D (7-AAD) 25 μg/ml (Cat No 559925 Becton Dickinson, San Jose, CA) was added just before flow cytometric analysis.

Flow Cytometry

Single color controls for annexin V-FITC, bis-oxonol, DilC1(5) and DAPI or 7-AAD were used to set compensations. Annexin V-FITC was detected in the 530/25 nm channel on the argon laser octagon (BD LSRII 500 Volts; BD FACSCanto II 357 V; BD FACSAria I 676 V). Bis-oxonol was detected in the 575/26 nm (BD FACSCanto II 495 V; BD FACSAria I 901V) or 610/10 nm channel (BD LSRII 610 V) on the argon laser octagon. 7-AAD was detected in the 670LP detector on the argon laser octagon (voltage 516 linear). DilC1(5) was detected in the 660/20 nm channel on Red HeNe trigon (BD LSRII 350 V; BD FACSCanto II 253 V; BD FACSAria I 291 V). DAPI was detected in the 440/40 nm channel on the violet diode trigon (BD LSRII 400 V; BD FACSCanto II 319 V; BD FACSAria I 577 V). The compensation matrix used on the BD FACSCanto II without 7-AAD was PE-FITC 59.5%; DAPI-FITC 0.12%; FITC-PE 3.12%; DAPI-PE 0.44%; APC-PE 1.5%; FITC-DAPI 0.08%, and APC-DAPI 0.06%. The compensation matrix used on the BD FACSCanto II with 7-AAD was PE-FITC 66.7%; DAPI-FITC 0.12%; FITC-PE 4.12%; DAPI-PE 0.44%; APC-PE 0.2%; PE-7-AAD 25%; FITC-DAPI 0.08%, and APC-DAPI 0.06%. The compensation matrix used on the BD LSR II was bis-oxonol-FITC 34.29%; APC-FITC 0.01%; FITC-bis-oxonol 0.74%; APC-bis-oxonol 6.59%; bis-oxonol-APC 0.33%; DAPI-APC 1.41%; FITC-DAPI 0.32%; bis-oxonol-DAPI 0.54% and APC-DAPI 0.02%. The compensation matrix used on BD FACSAria I was bis-oxonol-FITC 25%; APC-FITC 0.01%; FITC-bis-oxonol 1.01%; APC-bis-oxonol 1.25%; bis-oxonol-APC 5.5%; FITC-DAPI 0.125%; APC-DAPI 0.5%. The bis-oxonol dye measures pmp depolarization by showing an increase in fluorescence whilst cells undergoing hyperpolarization show a decrease in fluorescence. In contrast the carbocyanine dye, DiIC1(5) acts in a reverse manner to bis-oxonol, depolarization of mmp giving a decrease in fluorescence whilst hyperpolarization of mmp shows an increase in fluorescence.

Cells (100,000) were analyzed or sorted on a Becton Dickinson FACSCanto II or FACSAria I cell sorter fitted with a 488 nm Ti-Sapphire Argon laser, Red HeNe 633 nm diode and violet diode 405 nm with FACSDiva Software ver 6.1.2. Kinetic experiments lasting 25 min were performed on a BD LSRII fitted with a 488 nm Ti-Sapphire Argon laser, Red HeNe 633 nm diode, UV laser (350–360 nm) and violet diode 405 nm with FACSDiva Software ver 6.1.2. All data was analyzed on FlowJo (Treestar Inc, CA) in the form of list-mode data files version FCS 3.00 using the default bio-exponential transformation. Optical filters and mirrors in The BDFACS Aria I and LSRII were installed in 2005 and in 2008 for the BD FACSCanto II, respectively.

DNA Fragmentation Determinations

Cells undergoing the various apoptotic and oncotic treatments including STS, heat shock 42°C, 1% sodium azide and 0.01% Triton X-100 were analyzed for DNA fragmentation after 20 h incubation (n = 2). To characterize dead cells undergoing necrosis further the annexinV+ve/DAPI+ve population were analyzed for bis-oxonol intensity, with low and high populations being analyzed for DNA content by the addition of 7-AAD at 25 μg/ml. Sub G1 analysis was performed on a Becton Dickinson FACSCanto II using the 670LP channel set to linear and the width parameter used to allow doublet discrimination.

Sub G1 Analysis

Cells undergoing apoptosis or oncosis treatments as described above were sampled at 24 h for DNA fragmentation (n = 3). Cell preparations were fixed in 70% ice-cold ethanol and left on ice at least for 30 min. Cells were washed twice (at 2,000 rpm) in PBS buffer. Cells were then incubated with 100 μg/ml RNAse (Cat No R5125, Sigma Chemicals, Poole, UK) at 37°C for 15 min. Cells were resuspended in 50 μg/ml propidium iodide (Cat No P4170, Sigma Chemicals, Poole, UK). Samples were analyzed (20,000 events collected) on a Becton Dickinson FACS Canto II cytometer using the 576/25 nm channel (288 V) from the argon laser to detect PI in a linear manner with the width parameter used to exclude doublets of cells. Histogram analysis of the propidium iodide signal allowed the determination of the percentage of cells in Sub G1, G1, S phase, and G2m phases. Data was analyzed by FACSDiva software ver 6.1.2 and FlowJo software (Treestar Inc. CA). Paired Student t tests were performed in Microsoft OfficeExcel with P = >0.05 not significant (NS), P = <0.05*, and P = <0.01**.

Cell Sorting

Cells were treated with STS, heat shock 42°C, 1% sodium azide, or 0.01% Triton X-100 for 20 h. Necrotic cells from such cultures were sorted into two populations by gating on DAPI+ve/Annexin V-FITC+ve events, which were either bis-oxonol low or high intensity. Single color controls were used to set compensations as described above using a BD FACSAria I instrument.

Fluorescent Imaging

Treated cells (30 min) or sorted bis-oxonol low and high intensity necrotic cells (20 h treatments) were pelleted and 10 μl placed on glass microscope slide, mounted with a coverslip, and sealed with nail varnish. Cells were imaged using a Zeiss 510 confocal microscope (Jena, Germany) fitted with a meta-head detection system and argon laser (488 nm), violet diode (405 nm), green HeNe (543 nm), and red HeNe (633 nm). Annexin V-FITC was excited with the 488 nm laser and imaged in the 530/30 nm channel; bis-oxonol was excited with the 543 nm laser and imaged in the 560LP channel; DAPI was excited with the 405 nm laser and imaged in the 440/40 nm channel; DilC1(5) was excited by the red HeNe (633 nm) laser and imaged in the 660LP channel.


Real-Time Kinetic Flow Cytometric Analysis of Oncosis and Apoptosis

Induction of apoptosis or oncosis was measured flow cytometrically by a live cell rapid real-time kinetic analysis of changes in fluorescent signals of plasma and mitochondrial membrane potential dyes, bis-oxonol and DiIC1(5) respectively. Live cells (annexin V−ve/DAPI−ve) were gated according to the gating strategy shown in Supporting Information Figure 1. Live cells treated with STS showed a relatively slow mmp hyperpolarization (increase in fluorescence) and pmp hyperpolarization (reduction in fluorescence) respectively over a 25 min period (Fig. 1A). In contrast the oncotic agents sodium azide and Triton X-100, induced in live cells a depolarization of the pmp with an increase in bis-oxonol fluorescence and depolarization of mmp, resulting in reduced fluorescence of DiIC1(5) (Figs. 1C and 1G). Interestingly sodium azide initially induced a hyperpolarization of the mitochondrial membrane before abrogation of mitochondrial function (Fig. 1C). Heat shock treatment (42°C) of cells showed little change in pmp and a gradual reduction in DiIC1(5) fluorescence (Fig. 1E). This was in contrast to that found in cells undergoing heat shock at 56°C were there was an abrogation of mitochondrial function (data not shown). Thus heat shock at 42°C seems to be a relatively mild treatment of cells compared to the other treatments employed in this study.

Figure 1.

Real-time flow cytometric measurement of mmp and pmp with confocal microscopy. Jurkat cells were labeled with bis-oxonol, DiIC1(5), annexin V-FITC, and DAPI. A live cell gate was applied and a 30 s baseline recorded for bis-oxonol and DiIC1(5) on the 610/10 nm and 660/20 nm channels on a BD LSR II shown as overlaid line graphs (see Supporting Information Fig. 1 for gating strategy). Cells were then treated with 1 μM STS (A), 1% sodium azide (C), heat shocked at 42°C (E), and 0.01% Triton X-100 (G). The time course was run from 0 to 25 min, agonists were added after 30 sec as indicated by the arrow. Time course plots are representative experiments, n = 2. Confocal microscopy of treated samples after 25 min show bis-oxonol (red) and DiIC1(5) (purple) staining in cells treated with STS (B), 1% sodium azide (D), 42°C heat shock (F) and 0.01% Triton X-100 (H). Live cells were imaged on a Zeiss 510 confocal Meta microscope using 543 and 633 nm lasers to excite bis-oxonol and DiIC1(5) and 600LP and 650LP emissions collected respectively. Scale bars equal 10 μm. (n = 2).

The relative rapid reduction in mitochondrial function (reduction of fluorescence) observed with oncosis treatments was expected as cells undergoing oncosis have been reported to show a rapid depletion of ATP (10). In contrast apoptosis is known to be an ATP dependent process and thus mitochondrial function does not rapidly reduce, these differences being detected by real-time flow cytometric measurement of the mmp dye, DiIC1(5).

Confocal Microscopy

The effects on membrane potentials within live cells were also observed by confocal microscopy after the various treatments for 30 min (Fig. 1). Cells undergoing apoptosis with STS treatment for 0.5 h showed no obvious observable change compared to controls (data not shown) in DiIC1(5) fluorescence or bis-oxonol fluorescence as detected by kinetic flow cytometric analysis (Figs. 1A and 1B). Induction of oncosis by 1% sodium azide showed a complete abrogation of mmp indicated by the lack of DiIC1(5) fluorescence (Fig. 1D) and matches that observed by kinetic flow cytometry (Fig. 1C). The induction of oncosis by heat shock temperature of 42°C showed no such change in fluorescent staining patterns revealing only a slight visual change in bis-oxonol and DiIC1(5) fluorescence's mirroring that shown by kinetic flow cytometric analysis (Fig. 1E and 1F). Induction of oncosis by Triton X-100 showed an increase in bis-oxonol fluorescence and a decrease in DiIC1(5) fluorescence by confocal microscopy, again mirroring that shown by kinetic flow cytometry (Figs. 1G and 1H). The image and kinetic flow cytometric analysis not only showed the clear differences between the induction processes in apoptosis and oncosis but the different mechanisms of action of the oncotic agents used in this study, that is, physical and chemical treatment.

Mitochondrial Function

Live cell gating (see Supporting Information Fig. 1) of the time dependent data points (0–30 h) and analysis of DiIC1(5) fluorescence allows the determination of mitochondrial function at each time point. The live resting cell population (CTRL) showed a steady >90% of cells with mitochondrial function (Fig. 2). Live cells undergoing apoptosis (STS) showed a gradual fall in mitochondrial function over the 30 h time course, from 0–5 h there was a steady fall mitochondrial function from 85% at 1 h down to 60% after 5 h, with only 15% functional mitochondria present after 20 h (Fig. 2). In contrast all methods of induction of oncosis show a rapid fall in mitochondrial function in a remarkably similar degree over the 30 h time course (Fig. 2). After 1 h of treatment the mitochondrial function of live cells had fallen to 15–30%, rapidly falling to 5% after 5 h (Fig. 2). Interestingly the level of mitochondrial function of live cells which had undergone heat shock treatment showed a rapid increase in mitochondrial function, rising from 10 to 70% from 20–30 h. This apparent increase in mitochondrial function maybe due to the fact that the cells that have survived the heat shock maintain their mitochondrial function and proliferate with other cells going on to bind annexin V and undergo necrosis.

Figure 2.

Time course study of mitochondrial function. Jurkat cells were loaded with DiIC1(5) and bis-oxonol and labeled with annexin V-FITC and DAPI for each time point from 0–30 h. One hundred thousand events were collected. Live cells (annexin V-FITCneg/DAPIneg events) were analyzed for DiIC1(5) fluorescence and the level of mitochondrial function determined for each time point. Jurkat cells were untreated for controls (CTRL), treated with STS, 42°C heat shock (HS 42 C), 1% sodium azide (1% SA), and 0.01% Triton X-100 (T-X 0.01%), n = 3, error bars indicate SEM.

The time dependent profiles of mitochondrial dysfunction indicates that the real-time kinetic analysis of mitochondrial function in live cells undergoing oncosis has measured an actual fall in mitochondrial function within the first minutes of treatment in a reproducible manner given the relatively small SEMs of the data points, with the exception of those observed in cells undergoing heat treatment (Fig. 2).

Time Course Study of Oncosis and Apoptosis

Standard flow cytometric analysis of untreated cells, apoptotic and oncotic cells by the annexin V assay showed that annexin V was detectable during oncosis as previously reported in the literature with sodium azide, heat shock and Triton X-100 after 24 h of treatment, see Supporting Information Figures 2A–2D (7, 11–14, 22).

However, time course studies revealed that although annexin V rapidly binds to cells undergoing apoptosis in a significant manner after 3 h of treatment the same cannot be said of cells undergoing oncosis (Figs. 3A–3D). Chemical induction of oncosis by sodium azide showed a similar binding of annexin V to that observed in cells undergoing 42°C heat shock but was approximately 50% lower than that observed in cells undergoing apoptosis (Figs. 3B and 3C). Detergent induction of oncosis showed few cells with annexin V (<10%) binding capacity and a rapid increase in the dead cell population, this was very different from that observed by treatment with sodium azide and 42°C heat shock (Fig. 3D). In contrast, apoptosis showed a significant amount of death only at the latter stages of the time course (Fig. 3A). The various oncotic reagents induced a marked difference in the appearance of a rising dead cell population. Triton X-100 (0.01%) showed significant cell death at 2 h, whereas sodium azide and 42°C heat shock showed significant cell death at the latter stages of the time course respectively (Figs. 3B–3D).

Figure 3.

Time course study of annexin V binding. Jurkat T cells treated with 1 μM staurosporine (STS) to induce apoptosis (A), 1% sodium azide (B), heat shock 42°C (C), 0.01% Triton X-100 (D) for 30 h. At time points 0, 1, 2, 3, 4, 5, 20, 24, and 30 h cells were loaded with DiIC1(5) and bis-oxonol and labeled with annexin V-FITC and DAPI. 100,000 events were collected for each time point. Annexin V versus DAPI dot-plots were analyzed for each time point to determine percent live cells (annexin V−ve/DAPI−ve), apoptotic or annexin V+ve/DAPI−ve) and dead cells (annexin V+ve/DAPI+ve and annexin V−ve/DAPI−ve), n = 3, error bars indicate SEM.

Sub G1 Analysis of Oncotic and Apoptotic Treated Cells

Cells undergoing apoptosis showed a significantly higher level of DNA fragmentation than controls, whereas heat treated cells showed no significant increase in subG1 levels (Supporting Information Fig. 3, and Table 1). However, cells undergoing oncosis with sodium azide or Triton X-100 showed a higher level of DNA fragmentation than that observed with apoptosis (Supporting Information Fig. 3, and Table 1).

Table 1. Sub G1 analysis was performed after 24 h of treatment for controls, apoptosis induced by STS, heat shock 42°C, 1% sodium azide, and 0.01% Triton X-100 respectively
TreatmentMean Percentage SubG1
  • n = 3, Mean +/− SEM.

  • Paired Student t-test were performed,

  • a

    P = <0.05.

  • b

    P = <0.01.

  • NS = not significant.

Control12.6+/− 3
Apoptosis37.2+/− 6.8a
Heat Shock 42°C25.4+/− 7NS
1% Sodium Azide49.5+/− 4a
0.01% Triton X-10055.5+/− 7b

Time Course Study of Cell Plasma Membrane Potential Depolarization During Oncosis and Apoptosis

The time course study of pmp depolarization, (Fig. 4) showed a steady incidence of cells with pmp depolarization whether living (annexin V−ve/DAPI−ve), apoptotic (annexin V+ve/DAPI−ve) or dead (annexin V+ve/DAPI+ve, DP and annexin V−ve/DAPI+ve, SP) until the 20 h time point for all treatments compared to controls (Figs. 4A–4E). Plasma membrane depolarization increased after this time during apoptosis, sodium azide and Triton X-100 treatments (Figs. 4B, 4D, and 4E, Supporting Information Figs. 4D and 4E). In contrast heat shock treatment showed no change in levels of pmp depolarization even after 20 h and was similar to that found in controls, (Figs. 4A and 4C, Supporting Information Figs. 4A and 4C). Live cells undergoing apoptosis, showed a slight rise in pmp depolarization compared to control cells at 24 h (Figs. 4A and 4B and Supporting Information Figs. 4A and 4B).

Figure 4.

Time course study of plasma membrane depolarization. Jurkat T cells were untreated [control (A)] or treated with 1 μM staurosporine, apoptosis (B), heat shock at 42°C (C), 1% sodium azide (D), and 0.01% Triton X-100 (E) for 30 h. At time points 0, 1, 2, 3, 4, 5, 20, 24, and 30 h cells were loaded with DiIC1(5) and bis-oxonol and labeled with annexin V-FITC and DAPI. One hundred thousand events were collected for each time point and data analyzed to show changes in bis-oxonol fluorescence. Live (annexinV−ve/DAPI−ve), annexin V+ve, and both dead cell populations annexinV+ve/DAPI+ve (DP) and annexinV−ve/ DAPI+ve (SP) were analyzed for bis-oxonol depolarization, n = 3, errors bars indicate SEM.

Annexin V binding populations in all treatments at 24 h showed a higher incidence of pmp depolarization compared to that displayed by untreated live cells (Figs. 4A–4E). The two necrotic cell populations i.e. annexin V+ve/DAPI+ve (DP) and annexin V−ve/DAPI+ve (SP), showed different levels of pmp depolarization with a higher incidence in the double positive population compared to the single positive necrotic cell and annexin V+ve populations respectively (Figs. 4A–4D). This was with the exception of Triton X-100 treatment were the depolarization levels in all these populations was of a similar order (Fig. 4E).

The ‘live cells’ still present with all treatments at the latter time points had little mitochondrial function, see Figure 2 (with the exception of heat shock) and a significant number had depolarized plasma membranes (above control levels). This partially fulfills the criteria for cell death although these cells maintained membrane integrity. The mitochondrial function time course study showed clear differences in the mechanism of action of apoptotic and oncotic agents.

Phenotyping of Late and Early Necrosis

To determine the significance of high and low levels of bis-oxonol in necrotic cells, DNA content, sorting and image analysis of such cells in apoptotic and oncotic cultures was performed. The side scatter of such cells also varied significantly in the different oncotic cell cultures but not in cells undergoing apoptosis (Figs. 5A, 5E, 5I, and 5M). DNA content analysis of necrotic cells which were bis-oxonolhi+ve and bis-oxonollow+ve clearly showed that bis-oxonollow+ve had a higher percentage of cells located in the Sub G1 zone with little DNA, compared to the bis-oxonolhi+ve which had a higher DNA content (Figs. 5B, 5F, 5J, and 5N).

Figure 5.

Characterization of early and late necrotic cells. Jurkat T cells were treated with 1 μM staurosporine (apoptosis), 1% sodium azide, heat shock at 42°C and 0.01% Triton X-100 for 20 h, respectively. Cells were labeled with bis-oxonol, DiIC1(5), annexin V-FITC, and DAPI. Apoptosis and necrosis was determined by annexin V versus DAPI analysis. Necrotic cells with bis-oxonol low and high intensities were then analyzed for side scatter and DNA content (7-AAD 670LP) from apoptotic cell cultures (A), (B), 1% sodium azide (E), (F), heat shock 42°C (I), (J), and 0.01% Triton X-100 (M), (N). The low and high intensity bis-oxonol cells from the dead populations for each treatment were then sorted and imaged by confocal microscopy. DAPI staining indicated by blue, annexin V labeling indicated by green, bis-oxonol (pm) staining indicated by red, DiIC1(5) staining of mitochondria (mito) indicated by purple, these were then merged into a final image. Sorted dead cells undergoing apoptosis with bis-oxonol high (C) and low (D) were imaged followed by dead cells after treatments with 1% sodium azide with bis-oxonol high (G) and low intensities (H), dead cells from heat shock 42°C bis-oxonol high (K) and low (L), and 0.01% Triton X-100 bis-oxonol high (O) and low (P), respectively. Scale bars as indicated in μm. Histograms and confocal imaging representative experiments, n = 2.

Confocal microscopy of these two types of necrotic cells showed cells of approximately 2–20 μm in size depending on the treatment. Oncotic treatments generated necrotic cells of varying size which correlated generally to the level of bis-oxonol fluorescence, except in the case of Triton X-100 treatment were the reverse was true (Figs. 5O and 5P). Cells were otherwise generally larger if they were bis-oxonolhi+ve and smaller if they were bis-oxonollow+ve (Figs. 5G, 5H, 5K, and 5L); whereas, cells undergoing apoptosis were of a similar size, in both low and high intensity bis-oxonol events (Figs. 5C and 5D). Thus dead cells derived from oncosis, with bis-oxonolhi+ve and bis-oxonollow+ve staining appears to be indicative of early and late necrosis respectively as supported by the differences in side scatter, DNA content and the variation in size of bis-oxonollow+ve and bis-oxonolhi+ve stained cells, and generally verified by image analysis of bis-oxonolhi+ve and bis-oxonollow+ve cells.


Previous studies employing flow cytometry to characterize cells undergoing oncosis has focused on the use of annexin V binding to externalized PS to such cells (5, 11, 12, 14). DNA content has also been used to show a difference between apoptotic and heat shock treatment at 56°C (11). Both of these approaches have proved to be nonspecific for characterizing differences between cells undergoing apoptosis and oncosis (22, 13). More recently the use of SYTO16 and viability dyes has allowed the discrimination of apoptosis and oncosis in that no reduction in SYTO16 was observed in cells undergoing oncosis compared to that observed in cells undergoing apoptosis (7). In another advance, propidium iodide has been used in a real-time image analysis study of apoptosis from the start of the induction process and continuously over a 24 h period (27–29). In this approach adherent cells were preloaded with Hoechst and grown in culture with a low concentration of PI (0.25 μg/ml) and then exposed to STS for 24 h with images taken every 15 min in a real-time manner (27). The use of annexin V in real-time image analysis has also allowed the study of apoptosis by real-time live cell microscopy but only covering the 5–42 h period after the induction of apoptosis (30). The method described here, of a rapid real-time analysis of membrane potentials by standard flow cytometry has shown to be able to not only differentiate between apoptosis and oncosis but also show differences in the mechanism of action of different types of oncotic inducing reagents, including drug, chemical, heat, and detergent treatments. Longer time course studies has allowed further elucidation of differences occurring during apoptosis and oncosis, using annexin V, cell viability dyes, mitochondrial inner membrane, and plasma membrane potentials.

The changes in membrane potentials during the first 25 min of induction showed clearly that mitochondrial function was disrupted very early in oncosis over the range of types of reagents tested. It has been previously reported that oncosis is characterized by the rapid depletion of ATP (10). The kinetic changes in DiIC1(5) fluorescence indicates a rapid reduction in mitochondrial function and thus a depletion of ATP in cells undergoing oncosis. These changes in mitochondrial function were not observed in cells undergoing apoptosis as measured by real-time kinetic measurements of mmp. The time course study of mitochondrial function during apoptosis showed a 40% fall after 5 h whilst all oncotic agents showed a rapid reduction to <10% functionality over the same time period. The apoptotic reagent induced no significant changes in pmp, which was contrary to that observed with most cases of oncosis in which cells showed plasma membrane depolarization. Thus the measurement of mitochondrial membrane and plasma membrane potentials in real-time affords a rapid easy method to distinguish oncosis and apoptosis at the induction stage of these two different necrobiological processes. This approach to study oncosis could used in conjunction with that employed to study apoptosis in real-time and SYTO16 and thus enable differentiation between the induction of oncosis and apoptosis in vitro and potentially ex vivo (27–29).

Time course studies of annexin V binding, mitochondrial function, DNA content and plasma membrane depolarization indicate that PS externalization is part of the process of apoptosis, but is not an integral part of oncosis because of the observed low incidence of cells binding annexin V. Oncotic cells thus appear to rapidly lose membrane integrity and undergo necrosis, this was also observed when combining SYTO16 and PI and there was no decrease in SYTO16 signal to indicate an apoptotic response to sodium azide and heat shock (7). The different modes of action of the reagents employed in this study were also reflected in the varying degrees of DNA fragmentation. The degree of DNA fragmentation increased from heat shock treatments (although not significant), to significance with STS, then sodium azide and Triton X-100 treatments. The use of DNA fragmentation estimations can thus be useful in revealing the mechanism of action of the reagent employed in terms of the latter stages of the cell death process under investigation.

The varying degrees of plasma membrane depolarization revealed within the four population's studied, that is live cells, annexin V binding cells, as well dead cells that bind or do not bind annexin V is an interesting observation in that it further subdivides these cell populations. Further studies into the significance of these observations are currently underway. Untreated cells showed few live cells with depolarized membranes. There was a general increase in the proportion of cells showing depolarization in the order of, live cells, dead cells not binding annexin V, then annexin V binding cells, then lastly dead cells that bind annexin V showed a very high degree of plasma membrane depolarization. The proportion of cells displaying depolarization remained constant and only increased at the latter stages of treatments. It is interesting that a subpopulation of live cells undergoing oncosis had no mitochondrial function and had a depolarized plasma membrane, parameters indicative of dead cells and yet maintained plasma membrane integrity as measured by vital dye exclusion.

Necrotic cells were shown via bis-oxonol intensity to be defined as in an early or late phase of necrosis as defined by side scatter and DNA content. The use of bis-oxonol in this manner allows easy discrimination of necrosis into early and late phases and at the same time as measuring levels of annexin V binding and mitochondrial activity. This type of approach has also been employed by using SYTO16 to discriminate live, apoptotic, and necrotic and showed reductions in forward scatter, increases in side scatter and an increase in Sub G1, as the cells move through the apoptotic process (7). However here we show that the necrotic cell population can be further subdivided into early and late necrosis by use of the bis-oxonol signal intensity. This potentially adds another level of complexity to the necrobiological process.

The measurement of plasma membrane and mitochondrial inner membrane potentials can thus be used to investigate the mechanism of action of different oncotic inducing agents in real-time and show, which cells are in an early or late stage of necrosis. Previously SYTO16 has been used to differentiate oncosis and apoptosis in a time dependent manner (7). However, the real-time imaging of cells loaded with Hoechst in the presence of PI allowed the detection of apoptosis an hour after induction (27). This major advance in the study of apoptosis could be used to study the oncotic process too by the employment fluorescent mitochondrial dyes as tested in this study (27). This new approach of rapid real-time flow cytometric analysis of changes in mitochondrial function and plasma membrane depolarization allows the investigation of oncosis in an immediate manner by standard flow cytometry. This new assay will hopefully prove useful in the study of oncosis and drug treatment of tumors.


I would like to thank Dr Paul Allen and Prof Marion Macey for their assistance with the preparation of the manuscript.