Cytometric detection of chromatin relaxation, an early reporter of DNA damage response

Abstract

One of the early events of the DNA damage response (DDR), particularly if the damage involves induction of DNA double-strand breaks, is remodeling of chromatin structure characterized by its relaxation (decondensation). The relaxation increases accessibility of the damaged DNA sites to the repair machinery. We present here a simple cytometric approach to detect chromatin relaxation based on the analysis of the proclivity of DNA in situ to undergo denaturation after treatment with acid. DNA denaturation is probed by the metachromatic fluorochrome acridine orange (AO) which differentially stains single-stranded (denatured) DNA by fluorescing red and the double-stranded DNA by emitting green fluorescence. DNA damage was induced in both human leukemic TK6 cells and mitogen-stimulated human peripheral blood lymphocytes by exposure to UV light or by treatment with H₂O₂. Chromatin relaxation was revealed by diminished susceptibility of DNA to denaturation, likely reflecting decreased DNA torsional stress, seen as soon as 10 min after subjecting cells to UV or H₂O₂. While cells in all phases of the cell cycle showed a comparable extent of chromatin relaxation upon UV or H₂O₂ exposure, H2AX was phosphorylated on Ser139 predominantly in S-phase cells. The data are consistent with the notion that chromatin relaxation is global, affects all cells with damaged DNA, and is a prerequisite to the subsequent steps of DDR that can be selective to cells in a particular phase of the cell cycle. The method offers a rapid and simple means of detecting genotoxic insult on cells.

Introduction

The DNA damage response (DDR) is a complex, highly choreographed process involving the induction of a plethora of molecular interactions between numerous molecules. Halting cell cycle progression and division to prevent transfer of DNA damage to progeny cells, engagement of the DNA damage repair machinery, and/or activation of the apoptotic pathway to exterminate cells whose damaged DNA cannot be successfully repaired are the three critical goals of the DDR (reviewed in refs. 1-7).

One of the early events of the DDR, preceding ATM activation, is remodeling of chromatin that involves its relaxation (decondensation).^8-15 In fact, ATM activation appears to be triggered by chromatin changes induced by the altered torsional stress of the DNA double helix that occurs following DNA damage.¹⁶ High mobility group proteins (HMGs) play an important role in organization of chromatin that provides a dynamic response to DNA damage.^15,17,18 HMGs and histone H1 form a network of proteins that is continuously moving along the chromatin fiber and interacting with each other and with internucleosomal DNA. HMGs are among the most extensively modified nuclear proteins, phosphorylated, acetylated, methylated, ribosylated or sumoylated in response to changes in the physiological state of the cell, induction of cellular stress or cell cycle phase.¹⁹ This network provides a continuous highly dynamic interplay between a variety of nuclear structural proteins, modulating their binding to each other and to nucleosomes.^17-21 Of particular importance is the nucleosome binding protein HMGN1, which modulates the architecture of chromatin and affects the levels of post-transcriptional modifications of the tails of nucleosomal histones. HMGN1 binds to nucleosomes and reduces the rate of histone H3 phosphorylation on Ser10,²² a modification that facilitates chromatin condensation like that occurring during mitosis.²³ Thus, DNA damage induced activation of HMGN1 via post-translational modification may be directly associated with chromatin decondensation and, as a result, activation of ATM.

TIP60 histone acetyltransferase has also been shown to play a role in modulation of chromatin dynamics. In response to DNA damage, in addition to acetylation of histone H2AX, which is a prerequisite for its phosphorylation on Ser139, TIP60 regulates the ubiquitination of H2AX via the ubiquitin-conjugating enzyme UBC13.²⁴ Sequential acetylation and ubiquitination of H2AX by TIP60-UBC13 promotes enhanced histone dynamics, which in turn stimulates a DNA damage response. It should be noted that the presence of wt p53 appears to be critical for the induction of chromatin relaxation upon DNA damage^10,25 perhaps through its effect on the tumor suppressor p33ING2.²⁶ Chromatin relaxation enhances the accessibility of sites of DNA damage to the repair machinery.

The methods used to measure chromatin relaxation such as assessment of DNA accessibility to nucleases²⁵ or net phosphorus to nitrogen ratio obtained from the elemental maps with an imaging filter electromagnetic spectrometer²⁷ or energy-filtering transmission electron microscopy²⁸ are complex and cumbersome. To date, chromatin relaxation during DDR has yet to be assessed by flow or image-assisted cytometry.

One of the markers of chromatin condensation/relaxation is the propensity of DNA in situ to undergo denaturation after treatment with heat or acid. The extent of DNA denaturation in chromatin can be probed by the metachromatic fluorochrome acridine orange (AO), which upon binding to double-stranded (ds) DNA (by intercalation) fluoresces green with maximum emission at 530 nm while its interaction with the denatured, single-stranded DNA results in red fluorescence emission at ~640 nm.^29,30 Therefore, the ratio of red to green fluorescence of AO-stained cells provides information on the relative proportion of denatured to ds DNA, respectively. It has been observed that DNA in condensed chromatin of mitotic^31,32 or apoptotic cells^32-34 as well as in quiescent G₀ cells³⁵ was distinctly more denatured when stressed with heat or acid than DNA of interphase cells having less condensed chromatin. This feature provided a convenient marker to distinguish mitotic, apoptotic or G₀ cells by flow cytometry. We hypothesized that relaxation of chromatin during DDR may also be detected by an increase in resistance of DNA in situ to denaturation. The present data indicate that, indeed, exposure of cells to the DNA damaging agents UV light or the oxidant H₂O₂ rapidly enhances the resistance of DNA in situ to denaturation.

Results

Figure 1 presents the analysis of changes in chromatin structure of TK6 cells treated with UV or H₂O₂ as detected by the induction of partial DNA denaturation by acid.²⁰ It is quite evident that treatment with UV or H₂O₂ led to an increase in intensity of green and a decrease of red fluorescence, indicating that DNA in the treated cells was more resistant to acid (0.1 N HCl) driven denaturation. The degree of DNA denaturation is presented as the value α_t, which is the ratio of the mean value of red fluorescence intensity of the cell population (reporting AO-interactions with the denatured, ssDNA) to the mean total (red plus green) intensity of fluorescence (reporting AO interactions with ssDNA and dsDNA, i.e., total DNA content) of the same cells. Thus, the α_t index is an approximate estimate of the fraction of denatured DNA, and theoretically can vary from 0 (all DNA is in ds conformation) to 1.0 (all DNA is ss).³⁰ This ratio decreased in the cells treated with H₂O₂ or UV compared to untreated control cells with the extent of decrease more pronounced after treatment with H₂O₂. The UV- or H₂O₂-induced decrease in α_t occurred rather uniformly in all phases of the cell cycle, including mitotic cells (Fig. 1, bar graphs). It should be noted that in addition to the data shown in Figure 1, more detailed studies were carried out testing UV at 50 J/m² followed by different times of incubation (5–30 min) after exposure to UV, as well as analyzing cells after different times of treatment with H₂O₂ (5–60 min). The distinct decrease in α_t of DNA in TK6 cells was seen as early as 10 min after either their irradiation with 50 J/m² or incubation with 100 μM H₂O₂ (data not shown).

Figure 1.

Relaxation of chromatin of TK6 cells treated with UV or H₂O₂ detected by partial denaturation of DNA and staining with AO. Ctrl, untreated cells; UV, the cells were exposed to 100 J/m² UV light and then cultured for 30 min; H₂O₂, the cells were cultured with 100 μM H₂O₂ for 30 min. Cells were fixed then treated with 0.1 N HCl, stained with AO at pH 2.6 and their fluorescence measured by flow cytometry.³⁰ The red and green fluorescence reports AO binding to ssDNA and dsDNA, respectively.^29-31 The bivariate distributions (scatterplots) show the intensity of red and green fluorescence of individual cells. Mitotic cells (M) are characterized by a high degree of DNA denaturation, much higher than the interphase cells, which is reflected by their high intensity of red and low intensity of green fluorescence. The G₁, S, G₂ and M cell clusters were gated as shown by the dashed lines, and the mean intensity of red fluorescence (ssDNA) to total (red plus green) intensity of fluorescence (reporting ssDNA + dsDNA) was calculated for cells in each cell cycle phase, as described.³⁰ This ratio (α_t) provides an approximate estimate of the fraction of the denatured (ssDNA) with respect to total DNA content per cell. The mean values of α_t (×100) calculated for populations of G₁, S, G₂ and M cells are plotted as bar graphs (right).

Chromatin changes of proliferating human lymphocytes following treatment with UV or H₂O₂, as shown in Figure 2. The exposure of lymphocytes to UV or H₂O₂, as was the case with TK6 cells, led to an increase in resistance of their DNA in situ to denaturation reflected by an increase in intensity of green- and a decrease in red fluorescence after treatment with acid and staining with AO. The effect, expressed as a decrease in α_t, was of similar degree in G₁, S and G₂ cells and was not much different following treatment with UV or H₂O₂.

Figure 2.

Relaxation of chromatin in mitogenically stimulated lymphocytes after their treatment with UV light or H₂O₂. Human peripheral blood lymphocytes induced to proliferation by PHA were untreated (Ctrl), exposed to 50 J/m² UV then cultured for 30 min (UV), or treated in culture with 100 μM H₂O₂ for 20 min (H₂O₂). The cells were then fixed, treated with HCl and stained with AO. The extent of their DNA denaturation was assessed by analysis of the intensity of red and green fluorescence by flow cytometry.³⁰ G₁, S, G₂, M and apoptotic (Ap) cells were identified by their characteristic differences in red and green fluorescence, as described before.^32,33,35 Populations of cells in G₁, S and G₂ were gated as shown by the dashed outlines to obtain mean values of intensity of their red and green fluorescence to estimate the DNA denaturation α_t index³⁰ which is shown in the bar-graph to the right.

The induction of γH2AX in TK6 cells treated with UV or H₂O₂ with respect to the cell cycle phase is shown in Figure 3. It is evident that phosphorylation of H2AX induced by UV or H₂O₂ was cell cycle phase specific, maximal in S phase and minimal in G₁ phase cells. The raw data that show the type of bivariate distributions of γH2AX vs. DNA content of cells treated with UV or H₂O₂ have bas presented previously.^36,38

Figure 3.

Induction of γH2AX in TK6 cells treated with UV or H₂O₂. The cells were untreated (Ctrl), treated with 57 J/m² UV and then cultured for 30 min (UV), or treated with 100 μM H₂O₂ for 60 min (H₂O₂), then fixed. The expression of γH2AX was detected immunocytochemically and measured by cytometry in conjunction with cellular DNA content as described.^36,37 By gating analysis the mean values (+SD) of γH2AX expression were estimated for populations of G₁, S and G₂M cells and plotted as a bar graph.

Discussion

We report here that a very early event of the DDR, relaxation (decondensation) of chromatin can be detected by a method utilizing the metachromatic fluorochrome AO to differentially stain double-stranded versus denatured (ss) DNA. As mentioned, this methodology developed by us several decades ago^29,30 has been widely used in variety of applications to distinguish cells differing in their degree of chromatin condensation, such as mitotic³¹ or quiescent G₀,³⁵ cells from the proliferating interphase cells, apoptotic from non-apoptotic cells^22,23 or normal from abnormal spermatozoa.^34,39 Because optimal staining with AO requires stringent conditions of equilibrium between the dye and DNA (permeabilized cells) the method is more amenable to flow cytometry rather than to microscope imagining. However, with proper control of AO concentration, the method can be applied in microscopy.³²

Chromatin decondensation as measured by the decreased predisposition of DNA in situ to denaturation (DNA denaturation index α_t) could be detected 10 min after exposure of TK6 cells to UV light (50 or 100 J/m²) or 10 min after treatment with 100 μM H₂O₂ (Fig. 1). The extent of α_t change was approximately similar in the TK6 cells treated with 50 J/m² as compared with 100 J/m² of UV light, and lasted for up to 60 min (not shown). Comparable effects were also seen in the case of proliferating normal human lymphocytes (Fig. 2).

The present data demonstrate that chromatin relaxation was global, affecting cells in all phases of the cell cycle, including mitotic cells, i.e., the cells with initially maximally condensed chromatin. In contrast, confirming our earlier findings,^36,38 we observed that the induction of γH2AX either by UV or H₂O₂ was uneven, cell cycle phase dependent and much more pronounced in DNA replicating than in G₁ phase cells (Fig. 3). Other steps of DDR induced by H₂O₂ such as activation of ATM through its phosphorylation on Ser1981, Chk2 activation through Thr68 phosphorylation and p53 activation through Ser15 phosphorylation were also seen to be cell cycle phase dependent and maximal in S-phase cells.^37,38 However, the recruitment of Mre11 upon the treatment with H₂O₂, similar to chromatin relaxation, was shown to be global, affecting all cells regardless of their cell cycle phase.³⁸ Thus, DNA damage triggers chromatin relaxation and Mre11 recruitment with no regard to the cell cycle phase whereas the DDR steps subsequent to chromatin relaxation and Mre11 recruitment appear to be cell cycle-phase dependent, apparently regulated by cell cycle specific factors.

It is not entirely clear why DNA in condensed chromatin shows a higher propensity to denature when stressed by heat or acid compared to DNA in relaxed chromatin. It has been proposed that due to chromatin supercoiling, which is more intense in condensed chromatin, the torsional stress on the double helical structure of DNA is greater which makes the structure less stable and prone to denature (DNA melting) upon exposure to heat or acid.^29,31,40 In fact, DNA in the condensed chromatin of metaphase chromosomes, likely due to the increased torsional stress, shows the presence of ss sections sensitive to the ss-specific S1 or mung bean nucleases.^41,42 This global torsional DNA stress in nuclear chromatin dependent on the degree of chromatin condensation should not be confused with the local torsional stress that propagates through the chromatin fiber and appears to be essential in transcription regulation.^43-45

It should be noted that AO has been used to measure DNA damage and repair in cytometric assays based on the alkaline unwinding of DNA.^46-48 Following DNA damage or during nucleotide excision repair (NER) the damaged DNA sections become denatured under alkali conditions and AO is then used to measure the extent of DNA denaturation. Unlike in the present studies, in which we observed a decrease in the extent of DNA denaturation induced by acid following DNA damage, in the alkaline unwinding cytometric assay the increase in extent of DNA denaturation is a marker of DNA damage.

Chromatin dynamics during DDR is currently the focus of wide interest in many disciplines of biology and medicine.^6-11,49-52 Remodelling of chromatin is considered to be the prerequisite for further events associated with DNA repair. The cytometric method to detect chromatin relaxation we present here, by virtue of its simplicity, offers an attractive alternative to more complex assays such as those involving assessment of DNA accessibility to nucleases,²⁵ electron spectroscopic analysis of the elemental maps requiring an imaging filter electromagnetic spectrometer²⁷ or energy-filtering transmission electron microscopy.²⁸

Materials and Methods

Cells, cells treatment

Human B cell lymphoblastoid TK6 cells were kindly provided by Dr. Howard Liber of Colorado State University, Fort Collins, CO.⁵³ The cells were grown in 25 ml FALCON flasks (Becton Dickinson Co., Franklin Lakes, NJ) in RPMI 1640 medium supplemented with 10% fetal calf serum, 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM L-glutamine (all from GIBCO/BRL) at 37.5°C in an atmosphere of 5% CO₂ in air. At the onset of the experiments, there were fewer than 5 × 10⁵ cells per ml in culture such that the cells were in an exponential and asynchronous phase of growth. Human peripheral blood lymphocytes, obtained from healthy volunteers by venipuncture were isolated by density gradient centrifugation as described.⁵⁴ The cells were washed twice with phosphate-buffered saline (PBS) and resuspended in RPMI-1640 supplemented with 10% fetal bovine serum, 100 units/ml of penicillin, 100 μg/ml streptomycin and 2 mM L-glutamine (all from GIBCO/BRL Life Technologies, Inc., Grand Island, NY) at a density of ~5 × 10⁵ cells/ml. The cells were then treated with 10 μg/ml of phytohemagglutinin (PHA-P, Sigma Chemical Co., St. Louis, MO) and incubated in 25-ml (12.5 cm²) polystyrene flasks (Becton Dickinson, Franklin Lakes, NJ) in a mixture of 95% air and 5% carbon dioxide at 37.5°C for 72 h. The cultures of TK6 cells as well as of nonstimulated and PHA-stimulated lymphocytes were exposed to UV-B light (~300 nm) by placing them onto a UV-light gel illuminator (Fotodyne; West Berlin, WI), as described before³⁶ with the doses as shown in the Figure legends. The cultures were then transferred into the incubator at 37.5°C for up to 60 min. Other cultures were treated with 100 μM H₂O₂ (Sigma Chemical Co., St. Louis, MO) for different time intervals, as described in the legends to figures.

Cell staining, fluorescence measurements

The cells were fixed in suspension in 70% ethanol for over 12 h then suspended in phosphate buffered saline (PBS) and incubated in the presence of 100 μg/ml RNase A (Sigma) at 37°C for 20 min. Subsequently, the cells were treated with 0.1 M HCl followed by a solution phosphate-citrate buffer (pH 2.6) containing of 6 μg/ml of AO, at room temperature. The details of this methodology are presented elsewhere.³⁰ Green (Fl 1) and red (Fl 3) fluorescence were measured using a FACScan flow cytometer (Becton Dickinson, San Jose, CA) with the standard emission filters. Gating analysis was carried out to obtain the mean values of integrated green and red fluorescence of G₁, S and G₂M cells as shown in the Figures. The extent of DNA denaturation for cells in each phase of the cell cycle is presented as α_t value which represents the ratio of the mean red fluorescence intensity of cells at that phase of the cell cycle divided by the total (red plus green) mean fluorescence intensity of the same cells, as defined previously.^29,30 Other details are presented in Figure legends. Cells from the cultures parallel to these analyzed with respect to DNA denaturation as presented above were also subjected to the analysis of histone H2AX phosphorylation (Fig. 3) by the method described by us before.^38,55,56