Sage Journals: Discover world-class research

Abstract

Deuterium oxide (D₂O) increases both the fluorescence lifetime and the fluorescence intensity of the intercalating dyes propidium iodide (PI) and ethidium bromide (EB) when bound to nucleic acid structures. We have used spectroscopic analysis coupled with conventional and phase-sensitive flow cytometry to compare the alterations in intensity and lifetime of various DNA-binding fluorochromes bound to DNA and Chinese hamster ovary (CHO) cells in the presence of D₂O vs phosphate-buffered saline (PBS). Spectroscopic and flow cytometric studies showed a differential enhancement of intensity and lifetime based on the mode of fluorochrome-DNA interaction. The fluorescence properties of intercalating probes, such as 7-aminoactinomycin D (7-AAD) and ethidium homodimer II (EthD II) were enhanced to the greatest degree, followed by the probes TOTO and YOYO, and the non-intercalating probes Hoechst 33342 (HO) and 4',6-diamidino-2-phenylindole (DAPI). The non-intercalating probe mithramycin (MI) gave unexpected results, showing a great enhancement of fluorescence intensity and lifetime in D₂O, indicating that when staining is performed in PBS, much of the MI fluorescence is quenched by the solvent environment. Apoptotic subpopulations of HL-60 cells had a shorter lifetime compared to nonapoptotic subpopulations when stained with EthD II. These results indicate that accessibility of the dye molecules to the solvent environment, once bound to DNA, leads to the differential enhancement effects of D₂O on fluorescence intensity and lifetime of these probes.

Keywords

fluorescence lifetime fluorescence intensity flow cytometry DNA-binding fluorochromes deuterium oxide apoptosis

Dna-binding fluorochromes have been used extensively in many applications, including flow cytometry and gel and capillary electrophoresis. Fluorochromes utilized in these applications must satisfy several criteria. They must bind specifically to nucleic acids, there must be a low quantum yield as free dye in solution, and there must be enhancement of quantum yield on binding to nucleic acid structures. Several classes of DNA-specific fluorochromes with different modes of base pair (bp) binding are available: (a) the DNA intercalators propidium iodide (PI) and ethidium bromide (EB), which lack appreciable bp specificity; (b) DNA intercalators with A-T bp preference, such as ethidium homodimer II (EthD II) or with G-C preference, such as 7-amino actinomycin D (7-AAD); and (c) non-intercalating probes with either A-T bp specificity, such as Hoechst 33342 (HO) or 4',6-diamidino-2-phenylindole (DAPI) or with G-C specificity, such as mithramycin (MI). Two high quantum yield cyanine fluorochromes, TOTO and YOYO, are dimers of the dyes thiazole orange and oxazole yellow, respectively (Lee et al. 1986). These fluorochromes are believed to bind to DNA by intercalation, without base specificity (Haugland 1992) but with some base sequence preference (Netzel et al. 1995).

Enhancement of the fluorescence intensity of DNA-binding probes improves the signal-to-noise ratio, allowing the use of lower dye concentrations and thereby reducing potential self-quenching between dye molecules bound in close proximity. Deuterium oxide (D₂O; heavy water) increases EB and PI fluorescence intensity (Mazzini et al. 1983; Mazzini and Giordano 1980; Olmsted and Kearns 1977) as well as fluorescence lifetime (Sailer et al. 1996) in flow cytometry. Szabo et al. (1986) also demonstrated spectrophoto-metrically that the lifetime of DAPI bound to DNA is increased in D₂O. Both the enhancement of fluorescence intensity and the increase in fluorescence lifetime have been attributed to the stabilization in the proton transfer rate between dye bound to DNA and the solvent system (Olmsted and Kearns 1977). Increased time in the excited state leads to increased fluorescence emission and extended fluorescence lifetimes.

Measurement of fluorescence lifetime using phasesensitive electronics has recently provided an additional parameter for flow cytometry analysis (Deka et al. 1994; Pinsky et al. 1993; Steinkamp and Crissman 1993; Steinkamp et al. 1993). Phase-sensitive flow cytometry provides the capability to quantify fluorescence lifetimes (Steinkamp et al. 1993) and to resolve the emissions of multiple fluorochrome labels that have different lifetime values (Steinkamp and Crissman 1993). Quantitation of the lifetime of DNA-bound fluorochromes provides information on the molecular interactions of a fluorochrome within the target molecule (Sailer et al. 1996; Heller and Green-stock 1994). The lifetimes of DNA-bound fluorochromes are altered by physical and chemical factors such as solvents, cations, pH, energy transfer, and quenching. Measurement of the lifetime of the DNA-bound fluorochrome by phase-sensitive flow cytometry coupled with simultaneous DNA content analysis allows discrimination of subpopulations of cells with discrete lifetime alterations. Application of this new flow cytometric technology showed that subpopulations of apoptotic HL-60 cells with altered chromatin structure had a shorter lifetime compared to the non-apoptotic portion of the population after staining with PI or EB (Sailer et al. 1996).

Alterations in chromatin structure change the accessibility of the DNA-bound fluorochromes and influence both the intensity (Evenson et al. 1986; Darzyn-kiewicz et al. 1984) and the lifetime of the bound dye (Sailer et al. 1996; Heller and Greenstock 1994; Lakowicz 1992; Dewey 1991; Jameson and Reinhart 1989). Differences in the structure of fluorochrome molecules and in their mode of DNA binding lead to differences in D₂O-induced enhancement in the intensity and lifetime. Prior studies in our laboratory showed that D₂O had a more pronounced effect on the fluorescence intensity and lifetime of PI bound to chromatin of fixed CHO cells compared to EB (Sailer et al. 1996). Enhancement of the fluorescence intensity of DNA-bound PI in D₂O significantly improved the G₀/G₁ peak CV values for PI-stained CHO cells when these were analyzed on a commercial flow cytometer with limited laser power (Sailer et al. 1996).

In this study we used conventional and phase-sensitive flow cytometry to examine the effects of D₂O on dyes with different modes of DNA binding. D₂O enhanced the fluorescence intensity and fluorescence lifetime of all the dyes, but to an extent dependent on the mode of interaction of the dye with the DNA. Fluorescence intensity of non-intercalating, A-T base-specific dyes, such as HO and DAPI, was enhanced only slightly by D₂O, whereas the intensity of intercalating dyes with base specificity, such as 7-AAD (G-C-specific) and EthD II (A-T-specific) was enhanced to a greater degree, although not to as great an extent as previously noted for PI and EB. Fluorescence lifetimes of stained cells were increased by D₂O to an extent some-what proportional to the enhanced fluorescence intensity. Variation in either dye concentrations or the relative proportion of D₂O in the solvent provides the potential to preselect a lifetime value for a probe in studies utilizing phase-resolved measurements to separate the fluorescence of dyes on the basis of their respective lifetimes rather than of their spectral emission patterns. The interaction of EthD II with subpopulations of apoptotic cells also significantly reduced fluorescence lifetime compared to the non-apoptotic cells.

The results of this study indicate that the different modes of fluorochrome-DNA interaction, as well as differences in solvent accessibility of the fluorochromes, lead to a differential enhancement effect of D₂O on intensity and lifetime. The use of D₂O to study intensity and lifetime provides additional information on the microenvironment of the fluorochrome-DNA interaction. An example in this study was the unexpected D₂O-induced enhancement noted for the non-intercalating G-C-specific MI. On the basis of the non-intercalating mode of MI-DNA binding, a minimal enhancement of fluorescence intensity and lifetime comparable to results noted for HO and DAPI would have been expected. The great enhancement of MI intensity and the lifetime values in D₂O infer that when PBS is used as the solvent, a large portion of the MI fluorescence is potentially being lost due to solvent quenching. The results of this study will lead to further insight into the mechanisms and environmental influences on fluorochrome-DNA interactions.

Materials and Methods

Cells and Culture Conditions

Chinese hamster ovary (CHO) cells were grown in suspension culture at 37C in spinner flasks containing Ham's F10 medium (Gibco BRL; Grand Island, NY) supplemented with 15% heat-inactivated bovine calf serum (Hyclone Laboratories; Logan, UT), 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco). Cell densities were maintained at 0.25-4.5 × 10⁵ cells/ml. The population doubling time was 13-14 hr.

Human promyelocytic (line HL-60) cells were maintained in suspension culture at 37C in T-150 flasks containing RPMI-1640 (Gibco) supplemented with 10% fetal bovine serum (Hyclone Laboratories), 100 U/ml penicillin, 100 μg/ml streptomycin (Gibco), and 1.25 mM L-glutamine (Gibco). The population doubling time was 15-18 hr when cell densities were maintained at 0.2-1 × 10⁶ cells/ml.

Induction of Apoptosis in HL-60 Cells and Extraction of Low Molecular Weight DNA

HL-60 cells in culture were treated with 0.15 μM campto-thecin (Sigma; St Louis, MO) for 3 hr before fixation. Cells were harvested from suspension culture by centrifugation at room temperature (RT) for 5 min at 200 X g and thoroughly resuspended in one volume of cold Puck's saline A, without Ca²⁺ and Mg²⁺ and containing 1.0 mM EDTA (Gibco). Three volumes of cold 95% ethanol were then added to yield a final cell concentration of 1.0 × 10⁶/ml in 70% ethanol. After at least 24 hr of fixation, the cells were centrifuged at RT for 5 min at 200 X g and the fixative aspirated. Cell pellets were resuspended in 50 μl of phosphate-citrate buffer (192 parts 0.2 M Na₂HPO₄, 8 parts 0.1 M citric acid, pH 7.8) and incubated at RT for 15 min to extract low molecular weight DNA before staining (Gong et al. 1994).

Ethanol Fixation and Cell Staining

CHO cells were harvested from suspension culture by centrifugation at RT for 5 min at 200 X g and ethanol-fixed as described above to yield a final cell concentration of 1.0 × 10⁶/ml in 70% ethanol. Cells were fixed for at least 24 h, before staining. For staining, cells were centrifuged at RT for 5 min at 200 X g and the fixative aspirated. Stock solutions of DAPI (Molecular Probes; Eugene, OR) were prepared at 1 mg/ml in distilled H₂O or D₂O, HO (Calbiochem; La Jolla, CA) or MI (Pfizer Laboratories; Groton, CT) was mixed at 1 mg/ml in either PBS, pH 7.0, or D₂O with 0.15 M NaC1 (D₂O-saline, pH 7.1). Stock solutions of 7-AAD (Molecular Probes) and EthD II (Molecular Probes) were mixed at 1 mg/ml in DMSO and further diluted in PBS or D₂O-saline before staining. TOTO (Molecular Probes) or YOYO (Molecular Probes) was obtained at 1 mM in DMSO and was diluted in PBS or D₂O-saline before staining. Cells were resuspended in either PBS or D₂O-saline at 1 × 10⁶ cells/ml containing RNASE (Sigma) at a concentration of 50 μg/ml. For staining with MI, MgCl₂ was added to 5 mM. For the fluorochromes used, the modes of fluorochrome-DNA interaction and bp binding preference are listed in Table 1.

Table 1

DNA fluorochromes: modes of binding and DNA base sequence specificity a

Dye	Binding mode	Base pair specificity
HO	Non-intercalation	A-T
DAPI	Non-intercalation	A-T
7-AAD	Intercalation	G-C
EthD II	Intercalation	A-T
MI	Outside of DNA helix	G-C
TOTO	Intercalation	None
YOYO	Intercalation	None

^aDye binding modes and base pair specificity for all dyes except MI (Ward et al. 1965), referenced from Haugland (1992).

Spectrofluorometric Analysis

Spectrofluorometric scans of the fluorochromes were performed using a computer-driven SPEX (Edison, NJ) Fluorolog 1680 0.22 m double spectrometer with a 500-W xenon lamp source. All entrance and exit slits were set at 0.5 mm during sample runs. Free dye in solution, calf thymus DNA (Sigma) with fluorochromes in solution (50 μg/ml), or fixed and stained CHO cells were analyzed. Dye concentrations used for spectrophotometry were 0.5 μg/ml HO, 1.0 μg/ml DAPI, 20 μg/ml EthD II, 20 μg/ml 7-AAD, 5 μg/ml MI, 1 μg/ml TOTO, and 1 μg/ml YOYO.

Conventional Flow Cytometry

Flow cytometric measurements were performed on a Los Alamos flow cytometer (Steinkamp et al. 1982, 1991). An Innova 90-5 argon ion laser (Coherent; Palo Alto, CA) operating at 488 nm (300 mW) was used as the excitation source for TOTO and YOYO, 457 nm (200 mW) for MI, and 514 nm (300 mW) for 7-AAD and EthD II. A Innova Krypton laser (Coherent) was used for uv excitation (300 mW) of HO and DAPI. Dye concentrations for cell staining before flow cytometric measurements were 0.5 μg/ml HO, 1.0 μg/ml DAPI, 20 μg/ml EthD II, 20 μg/ml 7-AAD, 5 μg/ml MI, 1 μg/ml TOTO, and 1 μg/ml YOYO with fixed CHO cells. For one set of experiments, the mixture ratios of PBS to D₂O-saline were varied for cell staining.

Lifetime Flow Cytometry

Fluorescence lifetime measurements were performed on the Los Alamos phase-sensitive flow cytometer which has been described in detail (Steinkamp and Crissman 1993; Steinkamp et al. 1993). Stained cells were excited by a cw laser beam from an Argon ion laser (Model 2025; Spectra Physics, Palo Alto, CA), with the intensity being sinusoidally modulated at 29 MHz. Excitation wavelengths for the various fluorochromes were: uv excitation 351 nm for HO and DAPI, 457 nm for MI, 488 nm for TOTO and YOYO, and 514 nm for 7-AAD and EthD II. The intensity modulation was accomplished using an electro-optic modulator Model 350 (Conoptics; Danbury, CT). The fluorescence from the cells was detected by a photomultiplier tube (PMT Model 4526; Burle Industries, Lancaster, PA). The modulated fluorescence signal was processed by analogue phase-sensitive detection electronics to obtain the phase shift and lifetime at 29 MHz. Nonfluorescent beads (visible excitation; Duke Scientific, Palo Alto, CA) or TiO₂ particles (uv excitation; Fisher Scientific, Pittsburgh, PA) were analyzed at the same PMT gains as fluorescent samples by removing the laser barrier filter and replacing it with neutral density filters to maintain the orthogonal light scatter and fluorescence signals at the same values for nulling the phase-sensitive output signals for the fluorochrome-labeled sample. Coulter DNA Check alignment fluorospheres (Coulter; Hialeah, FL) (7 nsec lifetime) (Deka et al. 1994) were utilized to calibrate the flow cytometer. Samples were loaded into 3-ml syringes and delivered to the flow cell by a syringe pump for phase-sensitive lifetime analysis. Dye concentrations for phase-sensitive measurements on apoptotic HL-60 cells were 0.5 μg/ml HO, 1.0 μg/ml DAPI, 20 μg/ml EthD II, 20 μg/ml 7-AAD, 5 μg/ml MI, 1 μg/ml TOTO, and 1 μg/ml YOYO. For other phase-sensitive experiments using fixed CHO cells, the dye concentrations were varied over the following ranges: 0.001-10 μg/ml HO, TOTO or YOYO, 0.01-100 μg/ml DAPI, 7-AAD, EthD II, or MI.

Table 2

Percent increase in fluorescence intensity of different fluorochromes bound to different substrates in D₂O compared to PBS a

Dye	Free dye	Calf thymus DNA	Fixed CHO cells
DAPI	20	58	16
7-AAD	177	125	160
HO	61	41	66
EthD II	115	129	186
MI	43	118	137
TOTO	Neg b	14	47
YOYO	Neg b	6	24

^aValues represent relative comparisons of dyes bound to different substrates in D₂O-saline compared to PBS.

^bNegligible values for free dyes as obtained by Haugland (1992).

Statistical Analysis

Descriptive statistics and t-tests were performed in Microsoft Excel (Microsoft; Redmond, WA). Results are expressed as the mean ± SEM of at least three experiments.

Results

Comparison of Fluorescence Intensity and Spectral Alterations of Fluorochromes in PBS or D₂O-Saline Bound to Calf Thymus DNA and Stained CHO Cells

The use of D₂O-saline in place of PBS as the solvent in spectral studies resulted in increased fluorescence emission intensities for all of the free, unbound dyes in solution (except for TOTO and YOYO, which have no appreciable fluorescence except when DNA-bound), ranging from 20% to 177%. In order of increasing enhancement by D₂O were DAPI<MI<HO<EthD II<7-AAD (Table 2). When bound to calf thymus DNA, all of the dyes exhibited a 6-129% increase in fluorescence intensity in D₂O-saline compared to PBS, with the dyes ordered by increasing enhancement: YOYO<TOTO<HO<DAPI<MI<7-AAD<EthD II (Table 2). When bound to fixed, RNASE-treated CHO cells the fluorescence enhancement in D₂O-saline compared to PBS was 16-186%, with the dyes ranked as: DAPI<YOYO<TOTO<HO<MI<7-AAD<EthD II (Table 2).

Changes in the excitation peaks shown in Table 3 were observed among the different substrates utilized (free dye, calf thymus DNA, fixed CHO cells, respectively) for DAPI (all p<0.05). HO, EthD II, and MI showed changes in excitation peak only between the free dye in solution and bound to DNA, either calf thymus DNA or within fixed CHO cells (all p<0.05). 7-AAD showed changes in excitation peak between either the free dye in solution or within fixed CHO cells and bound to calf thymus DNA (p<0.05). Although changes in the excitation spectra were noted for the different substrates, only minor changes were observed when the solvent used was either D₂O-saline or PBS (all p>0.05). Changes were noted in the emission peak wavelength for 7-AAD (free dye or within fixed CHO cells vs calf thymus DNA; p<0.05), HO (all p<0.05), and EthD II (all p<0.05) among the different substrates. Only minor changes were noted in the emission spectra when the stained substrates in PBS or D₂O-saline were compared, except for the change in emission peak for free DAPI in solution between PBS and D₂O-saline (440 to 460 nm, respectively; p<0.01).

Table 3

Excitation and emission peak comparisons of fluorochromes in D₂O and PBS [expressed in nm (Ex/Em)] a

Dye-solvent	Free dye	Calf thymus DNA	Fixed CHO cells
DAPI-PBS	343/440	363/457	356/444
DAPI-D₂O	344/460	363/455	358/449
7-AAD-PBS	520/652	543/642	521/650
7-AAD-D₂O	516/650	539/639	516/649
HO-PBS	342/498	352/450	353/444
HO-D₂O	344/500	358/453	352/444
EthD II-PBS	504/629	535/618	534/621
EthD II-D₂O	499/630	535/617	533/620
MI-PBS	447/549	402/550	402/548
MI-D₂O	447/552	401/554	401/555
TOTO-PBS	Neg b	515/530	515/532
TOTO-D₂O	Neg b	515/530	515/532
YOYO-PBS	Neg b	490/508	490/509
YOYO-D₂O	Neg b	490/509	490/510

^aValues represent means of three independent measurements.

^bNegligible values for free dyes as obtained by Haugland (1992).

Figure 1

DNA content histograms illustrating the enhancement of fluorescence intensity for fixed CHO cells stained in D₂O-saline vs PBS. The mean of the G₀/G₁ peak in D₂O-saline compared to PBS was increased 150.5 ± 37.5% for MI (p<0.01), 34.5 ± 2.1% for 7-AAD (p<0.01), and only 9.2 ± 8.4% for DAPI (p>0.05). MI had the largest increase in fluorescence intensity, followed by the intercalating probes and the non-intercalating probes.

Conventional Flow Cytometric Fluorescence Intensity Measurements Using Various Solvent Ratios of PBS to D₂O-Saline

The effects of PBS on the fluorescence enhancement provided by D₂O was determined using conventional flow cytometric analysis of fixed CHO cells stained in mixture ratios of PBS to D₂O-saline with 10% volume increments from 100% PBS to 100% D₂O-saline. Figure 1 illustrates the enhancement of fluorescence intensity when samples were stained in 100% D₂O-saline vs 100% PBS for MI, DAPI, and 7-AAD. Fluorochrome intensity enhancement in D₂O varied, with the increase of the mean G₀/G₁ peak value from 100% PBS to 100% D₂O-saline ranging from 7.7 to 150.5% (Table 4). Figure 2 shows examples of the fluorochrome responses to various mixture ratios of PBS to D₂O-saline. The mean value of the G₀/G₁ peak was linearly correlated with the percent PBS for EthD II (r ²=0.952), MI (r ²=0.986), and for TOTO (data not shown; r ²=0.941) (all p<0.05). The probes 7-AAD, HO, and YOYO were not affected linearly by the addition of D₂O-saline (data not shown). 7-AAD showed a constant mean G₀/G₁ peak value from 100% PBS until a ratio of 40% PBS:60% D₂O-saline, at which point the mean increased (p<0.05). For HO, this shift occurred at 40% PBS:60% D₂O-saline (p<0.05), and for YOYO occurred at 70% PBS:30% D₂O-saline (p<0.05). G₀/G₁ peak mean values for DAPI (data not shown) were only slightly decreased over the range of mixture ratios (p>0.05). For dyes that changed linearly over the range of mixture ratios, the slope of the regression line is another indicator of how D₂O-saline affects the fluorescence intensity. The slope for MI (-0.87 ±0.15) is larger than for TOTO (-0.03±0.04) or EthD II (-0.34±0.09) (both p<0.01), indicating that PBS has a greater quenching effect on MI fluorescence.

Fluorescence Lifetime of DNA Fluorochromes

For phase-sensitive flow cytometric lifetime measurements, each of the fluorochromes was utilized to stain fixed CHO cells in PBS, and Figure 3 illustrates the fluorescence lifetime histograms obtained. Lifetime values for the probes were: 2.7 ± 0.05 nsec for DAPI, 2.3± 0.06 nsec for HO, 1.5 ± 0.02 nsec for MI, 0.8 ± 0.002 nsec for 7-AAD, 10.0 ± 0.45 nsec for EthD II, 2.2 ± 0.14 nsec for TOTO, and 2.1 ± 0.09 nsec for YOYO. PI and EB had lifetime values of 15 ± 0.5 and 19.5 ± 0.4 nsec, respectively.

Table 4

Percent increase in mean G₀/G₁ peak value for samples stained in D₂O-saline vs PBS a

Dye	% increase in D₂O-saline
DAPI	9.2 ± 8.4
7-AAD	34.5 ± 2.1 ∗∗
HO	19.4 ± 2.6 ∗
EthD II	63.0 ± 7.1 ∗∗
MI	150.5 ± 37.5 ∗∗
TOTO	7.7 ± 2.3
YOYO	9.1 ± 0.3 ∗

^aValues represent mean ± SEM of three measurements.

^∗ p<0.05.

^∗∗ p<0.01.

Figure 2

Enhancement of fluorescence intensity for fixed CHO cells stained in various mixture ratios of PBS to D₂O-saline. Various responses were noted for the fluorochromes: (a) a linear response from 100% PBS to 0% PBS (100% D₂O-saline), as shown for EthD II and MI, and (b) a variable response, shown for 7-AAD, in which the fluorescence intensity dropped suddenly with increasing proportions of PBS and then remained constant. Points represent the means of three measurements; error bars represent the SEM.

Figure 3

Fluorescence lifetime histograms for fixed CHO cells stained in PBS with various DNA fluorochromes. Fluorescence lifetime values for the dyes were: 2.7 ± 0.05 nsec for DAPI, 2.3 ± 0.06 nsec for HO, 1.5 ± 0.02 nsec for MI, 0.8 ± 0.002 nsec for 7-AAD, 10.0 ± 0.45 nsec for EthD II, 2.2 ± 0.14 nsec for TOTO, 2.1 ± 0.09 nsec for YOYO, 15 ± 0.5 nsec for PI, and 19.5 ± 0.4 nsec for EB.

Fluorescence Lifetime Alterations in D₂O

Figure 4 illustrates the effects of D₂O on fluorescence lifetime for DAPI-, MI-, and EthD II-stained CHO cells. At dye concentrations somewhat routinely utilized for DNA content analysis by flow cytometry, all of the fluorochromes showed an increase in fluorescence lifetime when staining was performed in D₂O-saline vs PBS (Table 5). Significant increases in fluorescence lifetime in D₂O-saline were seen for all of the fluorochromes. The increase of fluorescence lifetime in D₂O-saline for the fluorochromes parallels the spectral and conventional flow cytometric intensity observations, in which the intercalating fluorochromes 7-AAD and EthD II and the non-intercalating MI exhibit the largest increases.

Fluorescence Lifetime Alterations with Increasing Dye Concentration

Figure 5 illustrates the change in fluorescence lifetime in D₂O-saline vs PBS with increasing dye concentrations for MI, 7-AAD and TOTO. MI showed an increasing lifetime up to staining concentrations of 1 μg/ml, but lifetime values decreased at higher dye concentrations. All of the other probes showed the highest fluorescence lifetime at the lowest dye concentrations, with the lifetimes decreasing as the dye concentration increased. One exception was the sample stained with 100 μg/ml 7-AAD in D₂O-saline, in which the lifetime increased compared to the sample stained with 10 μg/ml (p<0.05). In most cases, samples stained in D₂O-saline had a higher lifetime than the corresponding sample in PBS. The only exception was for 7-AAD (Figure 5) at 0.01 μg/ml dye concentration, for which the lifetime for the sample in PBS was 30.8% higher (p<.05) than the sample in D₂O-saline (1.7 ± 0.17 and 1.3 ± 0.56 nsec, respectively).

Figure 4

Fluorescence lifetime histograms illustrating the increase in fluorescence lifetime for fixed CHO cells stained with DAPI, MI, or EthD II in D₂O-saline vs PBS. Fluorescence lifetime for DAPI-stained CHO cells in D₂O-saline compared to PBS increased 14.8% (1.0 μg/ml, 2.7 ± 0.05 to 3.1 ± 0.04 nsec; p<0.01), 85.7% for MI (10 μg/ml, 1.5 ± 0.02 to 2.9 ± 0.18 nsec; p<0.01), and 12.0% for EthD II (10 μg/ml; 10.0 ± 0.45 to 11.2 ± 0.72 nsec; p<0.05).

Differential Stainability of Apoptotic Subpopulations in D₂O-Saline Compared to PBS

Apoptotic cell subpopulations are apparent in a DNA content histogram as a sub-G₀/G₁ peak after treatment of HL-60 cells with CAM for 3 hr, ethanol fixation, and PC buffer extraction of low molecular weight DNA. Depending on the dye used for staining, the apoptotic subpopulations are shifted disproportionately from the non-apoptotic G₀/G₁ peak in D₂O-saline compared to PBS (Table 6).

Fluorescence Lifetime of Apoptotic Subpopulations

Figure 6 illustrates the change in fluorescence lifetime for apoptotic cells stained with EthD II in PBS (Figures 6A and 6E) or D₂O-saline (Figures 6B and 6F). In PBS, the lifetime of the apoptotic subpopulation was 26% lower than that of the non-apoptotic cells (7.2 ± 0.18 and 9.8 ± 1.2 nsec, respectively; p<0.01), and in D₂O-saline the apoptotic subpopulation was 25% lower (7.9 ± 0.3 and 10.5 ± 1.4 nsec, respectively; p<0.01). When stained with the other fluorochromes utilized in this study, such as MI (Figures 6C and 6G, PBS; Figures 6D and 6H, D₂O-saline), the apoptotic subpopulation did not have a different lifetime from that of the non-apoptotic cells (all p>0.05). In Figure 6B, a shoulder appears next to the peak representing the non-apoptotic cells in the sample stained with EthD II in D₂O-saline, and is apparently a portion of the G₀/G₁ sub-population (Figure 6F). It is possible that we are measuring a subset of the G₀/G₁ population, but further study will be required before this determination can be made.

Table 5

Increase in fluorescence lifetime for samples stained in D₂O-saline vs PBS a

Dye	Lifetime PBS (nsec)	Lifetime D₂O-saline (nsec)	% increase
DAPI	2.7 ± 0.05	3.1 ± 0.04	14.8 ∗∗
7-AAD	0.8 ± 0.002	0.9 ± 0.04	12.5 ∗∗
HO	2.3 ± 0.06	2.5 ± 0.03	8.7 ∗∗
EthD II	10.0 ± 0.45	11.2 ± 0.72	12.0 ∗
MI	1.5 ± 0.02	2.9 ± 0.18	85.7 ∗∗
TOTO	2.2 ± 0.14	2.6 ± 0.13	18.2 ∗
YOYO	2.1 ± 0.09	2.4 ± 0.18	14.2 ∗

^aValues represent mean ± SEM of three measurements.

^∗ p<0.05.

^∗∗ p<0.01.

Table 6

Percent increase in fluorescence intensity of apoptotic and non-apoptotic subpopulations for samples stained in D₂O-saline vs PBS a

Dye	Apoptotic subpopulation	Non-apoptotic G₀/G₁ peak
DAPI	27.0 ± 4.6	9.3 ± 3.2
7-AAD	54.8 ± 8.9	33.3 ± 6.1
HO	14.4 ± 3.6	20.7 ± 1.3
EthD II	143.5 ± 42.0	57.2 ± 19.4
MI	182.5 ± 66.4	125.5 ± 8.7
TOTO	-2.7 ± 1.8	5.0 ± 0.5
YOYO	21.7 ± 6.2	9.8 ± 4.7

^aValues represent mean ± SEM of three measurements.

Figure 5

Change in fluorescence lifetime with increasing dye concentration for MI-, 7-AAD-, and TOTO-stained CHO cells. At high dye concentrations, fluorescence lifetimes decreased, indicating that dye-dye self-quenching was occurring. Samples stained in D₂O-saline had greater lifetime values than the corresponding sample stained in PBS, the only exception being samples stained with 0.01 μg/ml 7-AAD. Points represent the means of three measurements; error bars represent the SEM.

Figure 6

Fluorescence lifetime of camptothecin treated HL-60 cells after extraction of low molecular weight DNA. Fluorescence lifetime histograms for cells stained with EthD II in PBS (A) or D₂O-saline (B), or with MI in PBS (C) or D₂O-saline (D). (E, F) Bivariate dot-plots of DNA content versus fluorescence lifetime corresponding to histograms A-D. When stained with EthD II in PBS, apoptotic subpopulations, denoted by their lower DNA content, had a fluorescence lifetime of 7.2 ± 0.18 nsec, whereas the non-apoptotic cells had a lifetime of 9.8 ± 1.2 nsec (26% lower; p<0.01). When stained in D₂O-saline, the apoptotic subpopulation had a lifetime of 7.9 ± 0.3 nsec, vs 10.5 ± 1.4 nsec for the non-apoptotic cells (25% lower; p<0.01). Apoptotic subpopulations stained with MI or the other fluorochromes utilized in this study did not have a different fluorescence lifetime from the non-apoptotic cells, in either PBS or D₂O-saline.

Discussion

The fluorochromes analyzed in this study have been used previously for the spectroscopic quantification of DNA (Barcellona and Gratton 1990; Van Lancker and Gheyssens 1986; Kapuscinski and Szer 1979), by flow cytometry (Crissman and Hirons 1994; Hirons et al. 1994; Schmid et al. 1992; Evenson et al. 1986; Darzynkiewicz et al. 1984) and for analysis of DNA after gel electrophoresis (Glazer et al. 1990; Nairn et al. 1982). In this study we compared the effects of D₂O and PBS on the intensity and lifetime of different DNA fluorochromes by flow cytometry. Spectroscopically, all of the dyes examined (see Table 1) showed an increase in fluorescence intensity when staining was performed in D₂O-saline vs PBS (Table 2), but, with the exception of the emission peak of unbound DAPI in solution, only minor changes in the excitation or emission peaks were noted in the DNA-bound fluorochromes between PBS and D₂O-saline. The absence of change in the excitation and emission peak wavelengths among identically prepared PBS- and D₂O-saline-stained samples indicates that D₂O does not spectrally alter the binding of the probe to the substrate, affecting only the fluorescence properties of the probes. Our study did demonstrate changes in the peak excitation wavelength when the dyes were bound to the different substrates. These alterations are postulated to indicate different binding modes of the dyes and changes in the accessibility of the dye molecules to the solvent environment (Heller and Greenstock 1994).

Phase-sensitive flow cytometric analysis of the DNA-binding fluorochromes provided fluorescence lifetime values in accord with previously published results (Netzel et al. 1995; Hochstrasser and Millar 1992; Wadkins and Jovin 1991; Barcellona and Gratton 1990; Barcellona et al. 1990; Olmsted and Kearns 1977). These results indicate that our flow cytometric measurements of fluorescence lifetimes are accurate, and provide a valuable new parameter for single-cell analyses.

Flow cytometric analysis showed that the various fluorochromes are affected by D₂O in a similar manner, but to a degree that is dependent on their mode of DNA binding. The fluorescence intensities of DAPI, HO, TOTO, and YOYO were not enhanced dramatically when they were bound to fixed CHO cells, as was the case for 7-AAD, EthD II, and MI in this study, or for PI and EB in our previous study (Sailer et al. 1996). These results indicate that differences in the DNA-fluorochrome interaction, or in solvent accessibility of the fluorochrome once bound to DNA, affect the intensity enhancement effects of D₂O. Intercalating probes that bind without base specificity, such as PI and EB (Haugland 1992), are enhanced in intensity and lifetime to the greatest degree, followed by the intercalating probes 7-AAD and EthD II and the minor groove, non-intercalative binding MI. Intercalating probes, such as TOTO and YOYO, and the non-intercalating A-T probes, DAPI and HO, are enhanced the least. Because TOTO and YOYO are much larger than the other intercalating probes, they may be more accessible to the solvent environment when bound, and therefore may be quenched to a greater extent by the solvent environment than are the smaller intercalating probes. The non-intercalating DAPI and HO also appear to be more accessible to the solvent environment. However, MI, another non-intercalating fluorochrome, may be shielded from the solvent environment once bound to the DNA, and therefore demonstrates a large enhancement of fluorescence intensity in D₂O. On the basis of its mode of DNA binding, MI would have been expected to show results similar to HO and DAPI. When the structures of the various dye molecules are compared, MI has a large number of sites that can be protonated by D₂O (Haugland 1992). Protonation by D₂O has been hypothesized to lead to a decrease in the proton transfer rate between fluorochrome and solvent (Olmsted and Kearns 1977), leading to a longer time in the excited state, enhancing the fluorescence intensity, and leading to longer fluorescence lifetime values.

In our studies utilizing a range of dye concentrations, fluorescence lifetime decreased concomitant with increasing dye concentrations. This decrease in fluorescence lifetime may be due to fluorescence quenching, a process that reduces the quantum yield of a probe without changing the fluorescence spectrum (Haugland 1992; Lakowicz 1983). One source of quenching involves energy transfer between dye molecules that are bound in close proximity on the substrate, and this dye-dye self-quenching has been noted previously (Haugland 1992; Dörr 1983; Lakowicz 1983). At low dye concentrations, D₂O enhances the fluorescence lifetime by limiting the quenching effect of solvent on the dye molecules. At high dye concentrations, however, D₂O-induced enhancement in lifetime is usually diminished, indicating that self-quenching between dye molecules supersedes the effects of the D₂O on the individual dye molecules.

Apoptosis, or programmed cell death, is characterized by internucleosomal cleavage of nuclear DNA which can be extracted from the cells after permeabilization of the cell membrane (Gong et al. 1994). The remaining nuclear DNA represents nucleosomes and fragments of DNA loops attached to the nuclear matrix (Darzynkiewicz et al. 1992), and apoptotic cells with reduced DNA content after extraction of low molecular weight DNA are regarded as a hallmark of the apoptotic process (Gong et al. 1994; Darzynkiewicz et al. 1992). In addition, our study shows further altered fluorochrome binding to the DNA within apoptotic cells. Specifically, D₂O enhances the fluorescence intensity of fluorochrome bound within apoptotic subpopulations, although not to the same extent as within non-apoptotic cells, indicating that solvent access to the bound fluorochrome within the apoptotic subpopulation may be different owing to the structure of the remaining chromatin, or that fluorochrome-DNA interactions may be changed on binding to the damaged chromatin within apoptotic cells. Moreover, we have shown in this study that subpopulations of apoptotic cells stained with EthD II, and with PI and EB in our previous study (Sailer et al. 1996), have a lower fluorescence lifetime than non-apoptotic cells, suggesting an altered interaction between DNA and the fluorochromes within the apoptotic subpopulation. However, not all the fluorochromes utilized in this study yielded detectable or decreased lifetime values for the apoptotic subpopulation. This appears to indicate that the DNA-binding properties of some fluorochromes are not altered in apoptotic subpopulations compared to the non-apoptotic cells.

It is interesting to note that smaller molecular weight intercalating dyes, such as PI and EB, have relatively longer lifetime values than the larger molecular weight fluorochromes. Intercalation stabilizes the fluorochrome molecule and shields it from the solvent environment, increasing its lifetime value relative to the unbound dye. In the presence of D₂O, the fluorescence lifetimes are further increased, indicating that interaction of the fluorochromes with the D₂O solvent environment leads to a further stabilization of the bound fluorochrome. The larger molecular weight fluorochromes may not be shielded from the solvent environment to the same extent once intercalated, leading to lower lifetime values and decreased enhancement by D₂O. Also of note is the fact that the lifetime for EB bound to fixed cells (19.5) is almost twice as large as the lifetime of EthD II (10.53) when the cells are stained in PBS. These results appear to indicate that even though these two dyes have the same fluorochrome group, the size of the molecule and the interaction of the probe with DNA and the solvent alter the fluorescence properties.

As was noted in experiments with changing dye concentrations, the fluorescence lifetime of the DNA fluorochromes is dependent on the dye concentration as well as on the solvent. This effect leads to potential preselection of specific lifetime value for a probe through a combination of dye concentration and solvent, either PBS or D₂O-saline. The use of multiple DNA fluorochromes has been used previously to elucidate changes throughout the cell cycle (Steinkamp and Crissman 1993) and for the determination of apoptotic populations (Darzynkiewicz et al. 1992). Because many of these fluorochromes have overlapping emission spectra, their use in combination for conventional flow cytometry is limited. However, because dyes with overlapping emission spectra, such as PI, EB, and EthD II, have different lifetime values, the potential exists for use of these dyes in combination for phase-resolved measurements of fluorescence (Steinkamp and Crissman 1993). Information on the dye-dye self-quenching of the various fluorochromes will allow the determination of optimal dye concentrations to use for studies on the dyes in combination.

In summary, we have utilized D₂O to illustrate differences in the binding of various DNA-specific fluorochromes measured by spectrophotometry, conventional flow cytometry, and phase-sensitive flow cytometry. We have shown that differences in chromatin structure, dye structure, dye-DNA interaction, and solvent accessibility lead to differences in the fluorescence intensity and fluorescence lifetime of the bound fluorochrome. These results, combined with future studies, will lead to further insight into the microenvironment of DNA-fluorochrome interaction and the effects of environmental influences on these interactions.

Footnotes

Acknowledgements

Supported by the United States Department of Energy and the Los Alamos National Laboratory Flow Cytometry Resource (NIH grants p41-RR01315, RR06758, and R01 RR07855).

References

Barcellona

Cardiel

Gratton

(1990) Time-resolved fluorescence of DAPI in solution and bound to polydeoxynucleotides. Biochem Biophys Res Commun 170:270–280

Barcellona

Gratton

(1990) The fluorescence properties of a DNA probe: 4′,6-diamidino-2-phenylindole (DAPI). Eur Biophys J 17:315–323

Crissman

Hirons

(1994) Staining of DNA in live and fixed cells. In Darzynkiewicz

Robinson

Crissman

, eds. Methods in Cell Biology. Vol 41. San Diego, Academic Press, 195–209

Darzynkiewicz

Bruno

Del Bino

Gorczyca

Hotz

Lassota

Traganos

(1992) Features of apoptotic cells measured by flow cytometry. Cytometry 13:795–808

Darzynkiewicz

Traganos

Kapuscinski

Staiano-Coico

Melamed

(1984) Accessibility of DNA in situ to various fluorochromes: relationship to chromatin changes during erythroid differentiation of Friend leukemia cells. Cytometry 5:355–363

Deka

Sklar

Steinkamp

(1994) Fluorescence lifetime measurements in a flow cytometer by amplitude demodulation using digital data acquisition technique. Cytometry 17:94–101

Dewey

(1991) Biophysical and Biochemical Aspects of Fluorescence Spectroscopy. New York, Plenum Press

Dörr

(1983) Mechanisms of energy transfer In Hoppe

Lohmann

Markl

Ziegler

, eds. Biophysics. Berlin, Springer-Verlag, 265–288

Evenson

Darzynkiewicz

Jost

Janca

Ballachey

(1986) Changes in accessibility of DNA to various fluorochromes during spermatogenesis. Cytometry 7:45–53

10.

Glazer

Peck

Mathies

(1990) A stable double-stranded DNA-ethidium homodimer complex: application to picogram fluorescence detection of DNA in agarose gels. Proc Natl Acad Sci USA 87:3851–3855

11.

Gong

Traganos

Darzynkiewicz

(1994) A selective procedure for DNA extraction from apoptotic cells applicable for gel electrophoresis and flow cytometry. Anal Biochem 218:314–319

12.

Haugland

(1992) Molecular Probes Handbook of Fluorescent Probes and Research Chemicals. 5th ed. Eugene, OR, Molecular Probes, Inc.

13.

Heller

Greenstock

(1994) Fluorescence lifetime analysis of DNA intercalated ethidium bromide and quenching by free dye. Biophys Chem 50:305–312

14.

Hirons

Fawcett

Crissman

(1994) TOTO and YOYO: new very bright fluorochromes for DNA content analyses by flow cytometry. Cytometry 15:129–140

15.

Hochstrasser

Millar

(1992) Fluorescence self-quenching of ethidium bromide intercalated in DNA. SPIE 1640:599–605

16.

Jameson

Reinhart

(1989) Fluorescent Biomolecules: Methodologies and Applications. New York, Plenum Press

17.

Kapuscinski

Szer

(1979) Interactions of 4′,6-diamidine-2-phenylindole with synthetic polynucleotides. Nucleic Acids Res 6: 3519–3534

18.

Lakowicz

(1992) Time-resolved laser spectroscopy in biochemistry III. Proc SPIE 1640:807

19.

Lakowicz

(1983) Principles of Fluorescence Spectroscopy. New York, Plenum Press

20.

Lee

Chen

C-H

Chiu

(1986) Thiazole orange: a new dye for reticulocyte analysis. Cytometry 7:508–517

21.

Mazzini

Giordano

(1980) Effects of some solvents on the fluorescence intensity of phenantridinic derivatives-DNA complexes: flow cytofluorometric applications. In Laerum

Lindmo

Thorud

, eds. Flow Cytometry IV. Oslo, Universitetsforlaget, 74–76

22.

Mazzini

Giordano

Riccardi

Montecucco

(1983) A flow cytometric study of the propidium iodide staining kinetics of human leukocytes and its relationship with chromatin structure. Cytometry 3:443–448

23.

Nairn

Dodson

Humphrey

(1982) Comparison of ethidium bromide and 4′,6-diamidino-2-phenylindole as quantitative fluorescent stains for DNA in agarose gels. J Biochem Biophys Methods 6:95–103

24.

Netzel

Nafisi

Zhao

Lenhard

Johnson

(1995) Base-content dependence of emission enhancements, quantum yields, and lifetimes for cyanine dyes bound to double-strand DNA: photophysical properties of monomeric and bichromophoric DNA stains. J Phys Chem 99:17936–17947

25.

Olmsted

Kearns

(1977) Mechanism of ethidium bromide fluorescence enhancement on binding to nucleic acids. Biochemistry 16:3647–3654

26.

Pinsky

Ladasky

Lakowicz

Berndt

Hoffman

(1993) Phase-resolved fluorescence lifetime measurements for flow cytometry. Cytometry 14:123–135

27.

Sailer

Nastasi

Valdez

Steinkamp

Crissman

(1996) Interactions of intercalating fluorochromes with DNA analyzed by conventional and fluorescence lifetime flow cytometry utilizing deuterium oxide. Cytometry 25:164–172

28.

Schmid

Krall

Uittenbogaart

Braun

Giorgi

(1992) Dead cell discrimination with 7-amino-actinomycin D in combination with dual color immunofluorescence in single laser flow cytometry. Cytometry 13:204–208

29.

Steinkamp

Crissman

(1993) Resolution of fluorescence signals from cells labeled with fluorochromes having different lifetimes by phase-sensitive flow cytometry. Cytometry 14:210–216

30.

Steinkamp

Habbersett

Hiebert

(1991) Improved multilaser/multiparameter flow cytometer for analysis and sorting of cells and particles. Rev Sci Instrum 62:2751–2764

31.

Steinkamp

Stewart

Crissman

(1982) Three-color fluorescence measurements on single cells excited at three laser wavelengths. Cytometry 2:226–231

32.

Steinkamp

Yoshida

Martin

(1993) Flow cytometer for resolving signals from heterogeneous fluorescence emissions and quantifying lifetime in fluorochrome-labeled cells/particles by phase-sensitive detection. Rev Sci Instrum 64:3440–3450

33.

Szabo

Krajcarski

Cavatorta

Masotti

Barcellona

(1986) Excited state pKa behaviour of DAPI. A rationalization of the fluorescence enhancement of DAPI in DAPI-nucleic acid complexes. Photochem Photobiol 44:143–150

34.

Van Lancker

Gheyssens

(1986) A comparison of four frequently used assays for quantitative determination of DNA. Anal Lett 19:615–623

35.

Wadkins

Jovin

(1991) Actinomycin D and 7-aminoactino-mycin D binding to single-stranded DNA. Biochemistry 30: 9469–9478

36.

Ward

Reich

Goldberg

(1965) Base specificity in the interaction of polynucleotides with antibiotic drugs. Science 149: 1259–1263

Differential Effects of Deuterium Oxide on the Fluorescence Lifetimes and Intensities of Dyes with Different Modes of Binding to DNA

Abstract

Keywords

Materials and Methods

Cells and Culture Conditions

Induction of Apoptosis in HL-60 Cells and Extraction of Low Molecular Weight DNA

Ethanol Fixation and Cell Staining

Spectrofluorometric Analysis

Conventional Flow Cytometry

Lifetime Flow Cytometry

Statistical Analysis

Results

Comparison of Fluorescence Intensity and Spectral Alterations of Fluorochromes in PBS or D2O-Saline Bound to Calf Thymus DNA and Stained CHO Cells

Conventional Flow Cytometric Fluorescence Intensity Measurements Using Various Solvent Ratios of PBS to D2O-Saline

Fluorescence Lifetime of DNA Fluorochromes

Fluorescence Lifetime Alterations in D2O

Fluorescence Lifetime Alterations with Increasing Dye Concentration

Differential Stainability of Apoptotic Subpopulations in D2O-Saline Compared to PBS

Fluorescence Lifetime of Apoptotic Subpopulations

Discussion

Footnotes

Acknowledgements

References

Comparison of Fluorescence Intensity and Spectral Alterations of Fluorochromes in PBS or D₂O-Saline Bound to Calf Thymus DNA and Stained CHO Cells

Conventional Flow Cytometric Fluorescence Intensity Measurements Using Various Solvent Ratios of PBS to D₂O-Saline

Fluorescence Lifetime Alterations in D₂O

Differential Stainability of Apoptotic Subpopulations in D₂O-Saline Compared to PBS