Apoptosis induced kinetic changes in autofluorescence of cultured HL60 cells-possible application for single cell analysis on chip

Introduction: This paper presents a new method using natural cellular fluorescence (autofluorescence, AF) to study apoptosis. Measurement of AF reduces sample preparation time and avoids cellular toxicity due to the fact that no labelling is required.

Methods: Human promyelocytic leukemic HL60 cells were incubated with camptothecin (CPT), tumour necro- sis factor (TNF)-α in combination with cycloheximide (CHX), or irradiated with 6 or 10 Gray, during varying time periods, to initiate apoptosis. AF was measured at the flow cytometer.

Results: Induction of apoptosis results in the shrink- age of the cell and the fragmentation into apoptotic bod- ies. With flow cytometry, 4 subpopulations, viable, early apoptotic, late apoptotic and the necrotic cells, can be dis- tinguished. Induction of apoptosis results in a decrease in AF intensity compared to untreated HL60 cells, espe- cially seen in the late apoptotic subpopulation. The AF intensity is found to decrease significantly in time (be- tween 2 h and 24 h) for all the four apoptotic inducers used.

Conclusions: Our results show that it is possible to specifically measure the apoptotic-induced kinetic changes in AF in HL60 cells. A decrease in AF intensity is seen from 2 h till 24 h. These results open a door for future developments in single-cell analysis.

Keywords: apoptosis; autofluorescence (AF); flow cytometry; HL60 cells.

Introduction

There are a number of techniques present to detect cell death. However, these tools are in most cases not spe- cific or lack quantitative value.1 In fact the very nature of apoptotic cell death promotes the underrecognition of this phenomenon for various reasons. First, apoptosis involves single cells scattered around. Second, the early stages of the apoptotic process evanesce and the apoptotic bodies are small and undergo rapid phagocytosis, an inflamma- tory reaction remains absent. Moreover, the duration of the whole process takes no more than a few hours.2 Nowadays, we are still seeking for a simple technique, which offers us to measure apoptotic cell death without manipulation of cells (e.g., staining which kills cells) and monitor the apoptotic cascade in real time.1

In the past decade, chip technology has shown its great value for chemical analysis in so-called Lab-on- a-Chip systems.3–5 Recently, the use of microtechnolo- gies for cell biology applications receives rapidly growing attention.6,7 Main reason for this is the manipulation of single cells in microfluidic structures and the possibilities for electrical characterisation and detection using micro- fabricated devices.8–10

Here a new method is presented which uses autofluo- rescence (AF) to study the process of apoptosis. Natural cellular fluorescence can be a useful tool to unravel in- tracellular pathophysiological processes and distinguish normal from diseased tissue.11 Many cellular metabolites exhibit autofluorescence,12 all having their specific emis- sion and excitation wavelengths. Collagen and elastin in a cell’s connective tissue and lipofuscin exhibit strong aut- ofluorescence, in the blue-green and yellow spectral re- gions, respectively.13 Nicotin-amide adenine dinucleotide (NAD(P)H) is the main fluorochrome excited by UVA, whereas NAD(P)+ is not fluorescent. NAD(P)H shows intracellular fluorescence in the blue spectral region and can be used as an indicator of redox state. The flavins, when oxidised, have strong autofluorescence in the yellow- green spectral region, and therefore can be considered as the main fluorochromes emitting above 500 nm.13–15 Both components are actively involved in a number of metabolic processes within the cell and play an important role in the energy household of the cell. Autofluorescence colocalizes strongly within the mitochondria and in some extent within the lysosomes, while the nucleus remains dark.16 The flavins and NAD(P)H are mainly responsible for this mitochondrial autofluorescence and therefore this autofluorescence can be used to give an indication of the metabolic activity of the cell.11,13,14,17,18 In most cases autofluorescence is seen as a nuisance which can obscure the fluorescence of various fluorescent probes used in op- tical techniques.12–14 However, measurement of autoflu- orescence can be very advantageous because no labelling is required which reduces sample preparation time and avoids cellular toxicity. In this study, human promyelo- cytic leukemic HL60 cells were incubated with various apoptotic inducers and autofluorescence was measured at the flow cytometer.

Materials and methods

HL60 cells

Human promyelocytic leukemic HL60 cells were ob- tained from the German Collection of Microorgan- isms (Braunschweig, Germany). Tissue culture equip- ment was supplied by Corning (Badhoeverdorp, The Netherlands). HL60 cells were cultured in RPMI-1640 medium supplemented with 10% heat-inactivated and filter-sterilised Foetal Calf Serum, 100 IU/ml peni- cillin, 100 mg/mL streptomycin, 2 mM L-glutamine and 250 µg/mL fungizone (RPMI+ medium). RPMI-1640 medium was obtained from BioWhittaker (Verviers, Bel- gium). Supplements and antibiotics were all obtained from Life Technologies (Grand Island, NY, USA). Cell cultures were maintained in a 5% CO2 humidified at- mosphere at 37◦ C. The medium was refreshed every 3–4 days. Exponentially growing cells were used in the experiments.

Modulation of HL60 cells

Induction of apoptosis. Irradiation. Cells at a concentration of 0.5 106 cells/ml in 24-wells plate were irradiated at a mean distance to target of 100 cm and a dose rate of 4.0 Gray/min with use of a linear accelerator (Varian Clinac 2100.C, Varian, Palo, Alto, CA) with an energy of 6 MV, to a dose of 6 and 10 Gray.

Incubation with camptothecin. Apoptosis was induced by incubation of cells under culturing conditions with 0.15 µM camptothecin (CPT, Sigma, St. Louis, MO, USA) during increasing time periods, varying from 0 to 48 hours.

Incubation with tumour necrosis factor-α and cycloheximide. Apoptosis was induced by incubation of cells under culturing conditions with 3 nM tumour necrosis factor (TNF)- α and 50 µM cycloheximide (CHX) (Sigma, St. Louis, MO, USA) during increasing time periods, varying from 0 to 24 hours.

Induction of necrosis. HL60 cells were heated for 2 hours at 57◦ C to induce necrosis. After heating, the cells were cultured in a 24-wells plate during increasing time periods, varying from 0 to 24 hours. Glucose. HL60 cells were incubated with 5 mM and 15 mM D(+)-Glucose (Merck, Darmstadt, Germany) during increasing time periods, varying from 0 to 48 hours.After incubation/irradiation, HL60 cells were washed twice and resuspended in 0.5 mL PBS. Samples were kept on ice until flow cytometry.

Flow cytometry

Autofluorescence of individual cells was measured with a Coulter Epics XL flow cytometer, using System IITM software with the XL-2 or DOS configuration. Excitation was elicited at 488 nm with the argon laser and measured using the FL-1 (peak 525 nm) and FL-2 (peak 575 nm) channels. In each sample 10,000 events were measured. Flow cytometry data were analysed with the computer program Expo II and gates/markers were set with un- treated HL60 cells. No difference in AF was seen between the FL-1 and FL-2 channel, thus only the results obtained in the FL-1 channel are shown.

Figure 1. (A) Light microscopy (LM) of untreated HL60 cells (con- trol) and HL60 cells 24 and 48 hours after irradiation with 6 Gray. HL60 cells were stained with May Gru¨ nwald staining. Magnifi- cance of the LM pictures is 50 . (B) Scatterdiagrams and fluo- rescence intensity histograms of untreated HL60 cells (control) and HL60 cells 24 and 48 hours after irradiation with 6 Gray. Four cell subpopulations (viable, 1, 2, and 3) can be identified on these scatterdiagrams, differing in their AF intensity, as shown in the fluorescence intensity histograms. Scatterdiagrams show results from a representative experiment.

Results

HL60 cells were incubated with various cell death in- ducers. Irradiation with 6 Gray activates the apoptotic cascade in HL60 cells, which results in cellular de- fragmentation and the formation of apoptotic bodies (Fig- ure 1A). Figure 1B shows the scatterdiagrams of untreated HL60 cells (control) and HL60 cells 24 and 48 hours after irradiation with 6 Gray. Four subpopulations are shown, respectively viable, early apoptotic (region 1), late apop- totic (region 2) and the necrotic (region 3) subpopulation, corresponding to the different stages of the apoptotic cas- cade in vitro. Further, for each subpopulation the fluores- cence intensity histograms are shown. The results from the scatter diagrams are plotted as function of time (Fig- ure 2). Untreated HL60 cells show only minor fluctuations in time (Figure 3), and can be seen as a stable control popu- lation. Autofluorescence intensity was measured using the markers shown in the histograms of Figure 1B (marker v, m-1, m-2 and m-3). Induction of apoptosis results the first 2 hours in a slight increase followed by a decrease in AF intensity till 24 hours. This decrease in AF inten- sity is mainly seen in the late apoptotic subpopulation (Figure 4).Between 24 and 48 hours the AF intensity is increasing again, especially in the early apoptotic region. The other apoptotic inducers (10 Gray, TNF-α/CHX and CPT) give similar results (data not shown). Further, the AF intensity is changing in time, with a maximum at t = 2h and a minimum at t = 24 h. The AF24/2 factor of each region is measured for all the four apoptotic inducers used. Figure 5 shows a decrease in AF24/2 factor in each re- gion compared to untreated HL60 cells (AF24/2 factor for untreated HL60 cells is set at 1), except for the measured AF24/2 factor of HL60 cells incubated with TNF-α/CHX or irradiated with 6 Gray in the early apoptotic region. Here a slight increase in AF24/2 factor is seen compared to untreated HL60 cells. Further, the specificity of this assay was obtained. Incubation with 5 mM glucose shows no change in AF intensity in time, however 15 mM glucose shows a slightly decreased AF24/2 factor. However, when HL60 cells are heated for 2 hours at 57◦ C to induce necrosis, the AF intensity is increasing with a factor 1.7 compared to untreated HL60 cells within 2 hours after heating. Thereafter, the AF intensity stays at this level for 24 hours, which results in an AF24/2 factor of nearly 1. The results of the AF24/2 factor are summarised in Table 1.

Discussion

Here a new method is presented in which AF intensity can be used to discriminate viable from apoptotic cells. HL60 cells were incubated with four different inducers of apop- tosis, all having their specific point of action in the apop- totic cascade. Tumour necrosis factor α activates the cell death receptor at the surface of the plasma membrane,19 while camptothecin arrest the cell cycle by inhibiting DNA topo-isomerase.20 Irradiation induces apoptosis by acting on the mitochondria.21 AF intensity is measured at the flow cytometer, thus only the autofluorescence emerging from the flavins is analysed.13–15 The progress of the AF intensity in time is the same for all these apop- totic inducers, probably because all the apoptotic stimuli integrate at the mitochondria, which serves as a receiv- ing platform.22 The AF intensity is increasing the first 2 hours after incubation followed by a decrease till 24 hours. Between 24 and 48 hours, the AF intensity is increasing again (Figure 4). An important cellular factor driving the cells to apoptotic cell death is the availability of cellu- lar ATP.23 Necrosis on the other hand is characterised by ATP depletion. In the first two hours after incuba- tion with apoptotic stimuli, the oxidised NADH and flavins are responsible for this ATP. The oxidised form of flavins is measured with flow cytometry and is proba- bly responsible for the increase in AF intensity. However, later in the apoptotic process, the cell’s ability to main- tain cellular ATP levels is compromised, which results in a decrease in AF intensity. After 24 hours the AF in- tensity is increasing again. A possible explanation might be that the cell is becoming necrotic, confirmed by light microscopy (Figure 1A 48 h). When cells are heated for 2 hours at 57◦ C to induce necrosis, the AF intensity is increasing, which strengthens this hypothesis. However, during the process of necrosis, ATP is depleted, so one would expect a direct drop in AF intensity. Incubation of HL60 cells with a low glucose concentration (5 mM) shows no change in AF intensity in time. The nutrients in the medium prevents the cell from taking up the extra nutrient, e.g., glucose, thus incubation with 5 mM glu- cose can be seen as an untreated sample. However, when HL60 cells are incubated with the higher glucose con- centration (15 mM), a decrease in AFintensity becomes prominent. Many studies have already shown that high concentrations of glucose induce apoptosis in different cell types.24–26 Further, incubation of cells with glucose prevents necrotic killing, due to the fact that ATP de- pletion is diminished. To translate the effects seen with flow cytometry to a microfluidic chip to be able to per- form single cell analysis, the AF24/2 factor is introduced.

Figure 2. The relative amount of viable, early apoptotic, late apoptotic and necrotic HL60 cells during apoptosis. The different cell populations of the quadrants shown in Figure 1B were plotted as a function of time after irradiation with 6 Gray. The plot represents the averaged cell number of one experiment performed in triple.

Figure 3. The relative amount of viable, early apoptotic, late apoptotic and necrotic HL60 cells. The different cell populations of the quadrants were plotted as function of time for untreated HL60 cells (control population). The plot represents the averaged number of one experiment.

Figure 4. Autofluorescence (AF) intensity of HL60 cells irradiated with 6 Gray. AF intensity is defined as the ratio of the mean fluorescence of the early apoptotic (region 1), late apoptotic (region 2) or necrotic (region 3) subpopulation as compared to the mean fluorescence of the viable population. The plot represents the averaged number of one experiment performed in triple.

Figure 5. AF24/2 factor of HL60 cultures treated with 0.15 µM CPT or 3 nM TNF-α in combination with 50 µM CHX, or irradiated with 6 Gray or 10 Gray. The AF24/2 factor is defined as the ratio of the minimal AF intensity (t = 24 h) compared to the maximal AF intensity (t = 2 h). The AF24/2 factor of untreated HL60 cells is set at 1.

For all four apoptotic inducers, the AF24/2 factor is de- creased (Table 1). This decrease is most prominent in the late apoptotic subpopulation. Although, HL60 cells in the necrotic stage of apoptosis also show a large decrease in AF24/2 factor compared to untreated HL60 cells, this effect is less powerful because the AF intensity of these cells is very low (Figure 4). Moreover, the decrease in AF24/2 factor is specific for apoptotic cells, while necrotic HL60 cells have an AF24/2 factor comparable to untreated HL60 cells. Though, the AF intensity is increasing com- pared to untreated HL60 cells, this increase limits to the first two hours after heating, and then stays at this level, which results in an AF24/2 factor near 1. Recently, many efforts are made in using microfluidic devices to perform single-cell analysis.7 Recent results have shown that it is possible to perform AF measurements on single cells in a microfluidic device.27 Hence, no labelling with fluores- cent probes is required, which limits sample preparation times and potential cell toxicity. Our goal is to trans- late the effects seen with flow cytometry to a microflu- idic chip. Therefore a new microfluidic chip has been de- veloped enabling the capture of viable cells. Once cells go into apoptosis their mechanical properties, e.g., size, will change, and these apoptotic cells will be able to pass the capture position, which enable measurements of the apoptotic cell death kinetics on chip. Optical detection of a decrease in AF intensity will confirm this hypothe- sis. In future developments the optical detection will be transferred to an electrical on-chip cell counter specific for apoptosis.

Conclusion

Our results have demonstrated the usefulness of analysing AF intensity in time. Measuring the AF intensity is a rapid and simple technique to study the process of apoptosis in a specific manner. Further, it offers new possibilities of performing single-cell analysis on chip.