HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 24, 21683-21690, June 14, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/24/21683    most recent M201027200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Widlak, P. Articles by Garrard, W. T. Search for Related Content PubMed PubMed Citation Articles by Widlak, P. Articles by Garrard, W. T. Social Bookmarking                      What's this?

Modeling Apoptotic Chromatin Condensation in Normal Cell Nuclei

REQUIREMENT FOR INTRANUCLEAR MOBILITY AND ACTIN INVOLVEMENT^*

Piotr Widlak, Olena Palyvoda, Slawomir Kumala, and William T. Garrard^§¶

From the  Department of Experimental and Clinical Radiobiology, Center of Oncology, 44-100 Gliwice, Poland and the ^§ Department of Molecular Biology, University of Texas Southwestern Medical Center, Dallas, Texas 75390

Received for publication, January 30, 2002, and in revised form, March 25, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Hallmarks of the terminal stages of apoptosis are genomic DNA fragmentation and chromatin condensation. Here, we have studied the mechanism of condensation both in vitro and in vivo. We found that DNA fragmentation per se of isolated nuclei from non-apoptotic cells induced chromatin condensation that closely resembles the morphology seen in apoptotic cells, independent of ATP utilization, at physiological ionic strengths. Interestingly, chromatin condensation was accompanied by release of nuclear actin, and both condensation and actin release could be blocked by reversibly pretreating nuclei with Ca²⁺, Cu²⁺, diamide, or low pH, procedures shown to stabilize internal nuclear components. Moreover, specific inhibition of nuclear F-actin depolymerization or promotion of its formation also reduced chromatin condensation. Chromatin condensation could also be inhibited by exposing nuclei to reagents that bind to the DNA minor groove, disrupting native nucleosomal DNA wrapping. In addition, in cultured cells undergoing apoptosis, drugs that inhibit depolymerization of actin or bind to the minor groove also reduced chromatin condensation, but not DNA fragmentation. Therefore, the ability of chromatin fragments with intact nucleosomes to form large clumps of condensed chromatin during apoptosis requires the apparent disassembly of internal nuclear structures that may normally constrain chromosome subdomains in non-apoptotic cells.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Characteristics of the terminal stages of apoptosis are genomic DNA fragmentation and chromatin condensation (reviewed in Ref. 1). Although internucleosomal DNA breakdown is often temporally correlated with such chromatin condensation, it is not absolutely required to trigger this event (2-4). Three pathways have been identified that mediate apoptotic chromatin condensation: (i) a caspase-3-independent pathway triggered by mitochondrial apoptosis-inducing factor, which leads to an accompanying large-scale DNA fragmentation without internucleosomal DNA cleavage (5); (ii) a caspase-3-dependent pathway triggered by the protein acinus, which occurs without inducing any DNA fragmentation (6); and (iii) a caspase-3-dependent pathway that leads to internucleosomal DNA cleavage mediated by activated DFF,¹ also termed caspase-activated deoxyribonuclease or caspase-activated nuclease (7-11).

Although little is known regarding the mechanisms that lead to chromatin condensation in apoptotic cells, considerable information has appeared in the literature on the components that are linked to chromatin condensation in non-apoptotic cells. In interphase nuclei, heterochromatic regions are often associated with specific chromosomal proteins, post-translational modifications, and methylated DNA (reviewed in Refs. 12-14). In mitotic cells, chromosome condensation requires post-translational modifications and the action of an ATP-dependent complex called "condensin" to introduce positive DNA supercoils into DNA substrates in the presence of topoisomerases (Ref. 15; reviewed in Ref. 16). In summary, many parameters and protein factors have been associated with the formation of heterochromatin and condensation of mitotic chromosomes, but which, if any, of these apply to chromatin condensation during apoptosis remains to be determined.

Here, we have investigated the mechanism of chromatin condensation mediated by the internucleosomal DNA cleavage pathway in apoptosis. Previously, we demonstrated that purified activated recombinant DFF triggers chromatin condensation when added to nuclei isolated from non-apoptotic cells (8). In the present study we have exploited this observation to dissect the mechanism of condensation triggered by internucleosomal DNA fragmentation using the well established model of isolated HeLa cell nuclei for in vitro reconstitution experiments of apoptotic events (7, 17). Where possible, we have complemented our in vitro findings by utilizing cultured human leukemic HL-60 cells triggered to undergo apoptosis with etoposide or cisplatin (18). Our results reveal that chromatin condensation at physiological ionic strength requires intranuclear mobility, which is allowed by the apparent disruption of components that may compartmentalize chromosome subdomains in normal cells. In addition, such condensation requires intact nucleosomes.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture and Treatment-- Human lymphoblastoid HL-60 cells, cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, were exposed to etoposide (Bristol-Myers Squibb Co.) or cisplatin (Ebewe) at final concentrations of 34 and 27 µM for 6 and 12 h, respectively. In some experiments, DNA minor groove- or actin-binding drugs (at concentrations indicated in below) were added to the medium 30 min prior to etoposide. Cells were then washed with PBS, fixed with 10% formaldehyde for 1 h at 4 °C, washed with PBS, transferred onto microscope slides, dried, and stained with 10 µM 4,6-diamidino-2-phenylindole. DNA from portions of PBS-washed cells was purified and analyzed by agarose gel electrophoresis as described below. For labeling replicating DNA, 200 µM each BrdUrd and deoxycytidine were added to the above medium; and after 10 h, cells were pelleted, resuspended in fresh medium, and treated with drugs as described above for 6 h. Samples that were fixed with 10% formaldehyde for 10 min were used for detection of BrdUrd-labeled DNA as described (19). Briefly, after washing fixed cells with PBS, treatment with methanol/acetic acid (3:1), and transfer onto slides, DNA was denatured, reacted with mouse anti-BrdUrd monoclonal antibody (Sigma), and visualized with Texas Red-conjugated goat anti-mouse Ig antibody (Jackson ImmunoResearch Laboratories, Inc.). For nuclease treatment in situ, HL-60 cells were permeabilized with lysolecithin as described (20). Briefly, 2 × 10⁶ PBS-washed cells were incubated for 1 min at 37 °C with 0.05% lysolecithin in incubation buffer (0.1 M KCl, 0.15 M sucrose, 4 mM MgCl₂, 1 mM CaCl₂, and 20 mM Hepes, pH 7.5). Cells were then washed once with incubation buffer by gentle centrifugation and suspended in 0.2 ml of isolation buffer (10 mM KCl, 0.25 M sucrose, 4 mM MgCl₂, 1 mM dithiothreitol, 20 mM Hepes, pH 7.5, and Complete^TM protease inhibitor mixture (Roche Molecular Biochemicals)) supplemented with 20 units of MNase (Worthington). After 20 min of incubation at 34 °C, cells were fixed with 10% formaldehyde.

Isolation and Treatment of Nuclei-- Nuclei were purified from HeLa S3 or HL-60 cells. Cells were washed by centrifugation with PBS at 4 °C and suspended in cold isolation buffer. Nonidet P-40 was added to 0.5%, and cells were incubated for 10 min on ice. After washing twice with isolation buffer by centrifugation, nuclei were suspended in isolation buffer with 50% glycerol and stored at -80 °C or used fresh. The purified nuclei (~30 µg of protein) were incubated with activated DFF (100 units/ml), MNase (50 units/ml), or DNase I (10 units/ml; Roche Molecular Biochemicals) for 30 min at 34 °C in 50 µl of buffer consisting of 100 mM KCl, 4 mM MgCl₂, 1 mM CaCl₂, 0.5 mM ATP, 1 mM dithiothreitol, and 20 mM Hepes, pH 7.5. Human recombinant DFF was purified from an Escherichia coli expression system and activated by incubation with caspase-3 as described previously (21), and then caspase was inhibited by 10 µM acetyl-Asp-Glu-Val-Asp-aldehyde (Ac-DEVD-Cho). In some experiments, nuclei were incubated with AluI enzyme (1000 units/ml; Roche Molecular Biochemicals) or bleomycin (1 mg/ml; Nippon Kayaku). When indicated, nuclei were pretreated with apyrase (10 units/ml) for 15 min at 34 °C; or ATP was replaced with ATPS or AMP-PNP; or nuclei were treated with nucleases in the presence of 1 mM ZnCl₂ or 10 mM KCl. In some experiments, immediately before treatment with nucleases, nuclei were incubated for 30 min at 34 °C with the following compounds (at concentrations indicated in the figure legends): distamycin A, chromomycin A₃, bisbenzimide (Hoechst 33258), methyl green, ethidium bromide, etoposide, and phalloidin (all from Sigma) and latrunculin A and jasplakinolide (Molecular Probes, Inc.).

Extraction of Nuclei and Stabilization of Intranuclear Components-- HeLa nuclei were extracted with 0.2 or 0.35 M NaCl in a buffer consisting of 5 mM MgCl₂, 1 mM EDTA, 1 mM dithiothreitol, and 20 mM Hepes, pH 7.5. Nuclei were incubated for 30 min on ice and then centrifuged for 15 min at 400 × g (control nuclei were incubated in the same buffer, but containing 50 mM NaCl). HeLa nuclei were treated with acidic buffer according to the procedure of Lawson and Cole (22). Nuclei were suspended in 68 mM citric acid/sodium phosphate buffer, pH 4.0, containing 25 mM KCl, 0.25 M sucrose, and 5 mM MgCl₂ and then incubated on ice for 30 min. Nuclei were centrifuged for 15 min at 400 × g and then washed twice with neutralization buffer (25 mM KCl, 0.25 M sucrose, 5 mM MgCl₂, and 20 mM Hepes, pH 7.5). HeLa nuclei were incubated on ice with 5 mM CuCl₂ (pretreatment of nuclei with Cu²⁺ diluted to 1 mM was equally effective in blocking condensation), or 10 mM diamide in neutralization buffer, or 25 mM CaCl₂ in neutralization buffer at 37 °C for 15 min and then washed twice with cold neutralization buffer.

Analysis of DNA Laddering and Chromatin Condensation-- After incubation of nuclei with nucleases, one-fourth of the reaction mixture was subjected to DNA analyses by agarose gel electrophoresis as described previously (21). The rest of the reaction mixture was fixed with 10% formaldehyde for 1 h on ice. Nuclei were than centrifuged and suspended in 10 µM 4,6-diamidino-2-phenylindole dissolved in the reaction mixture buffer. Samples were visualized by epifluorescence microscopy, and images were acquired with a color CCD camera. Numbers below panels on the figures show the percentage of nuclei exhibiting chromatin condensation (average of two to three independent experiments, with 200-300 purified nuclei or 500-1000 cells being scored per experiment).

Protein Analyses-- Proteins from nuclei or nuclear extracts were separated on SDS-15% polyacrylamide gels and then stained with Coomassie Brilliant Blue or electrophoretically transferred onto nitrocellulose membranes (Protran, Schleicher & Schüll). Membrane-immobilized proteins were probed with mouse anti-human actin monoclonal antibody and visualized with horseradish peroxidase-conjugated goat anti-mouse Ig antibody (Oncogene Research Products).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Internucleosomal DNA Fragmentation of Nuclei Isolated from Non-apoptotic Cells Induces Chromatin Condensation Resembling That Seen in Apoptotic Cells-- We first directly compared the morphology of chromatin condensed in cultured cells undergoing apoptosis with that triggered by DNA fragmentation in vitro in nuclei isolated from the corresponding non-apoptotic cells. For this purpose, we treated cultured human HL-60 cells with either etoposide or cisplatin to induce apoptosis and compared pairwise the extent of nucleosomal DNA laddering with chromatin condensation in these samples and likewise in nuclei isolated from untreated HL-60 cells (Fig. 1). In agreement with our previous results (8), the morphology of nuclei in cells undergoing apoptosis was similar to that induced in nuclei isolated from normal cells after DFF treatment, as visualized by fluorescent microscopy after DNA staining with 4,6-diamidino-2-phenylindole (Fig. 1B). In the context of the work reported here, we will refer to this morphology as that representing "chromatin condensation," which we define as the coalescence of chromatin into distinct clumps localized mostly at the nuclear periphery. However, such condensation was not DFF-specific because DNA fragmentation by MNase, DNase I, AluI, or even the radiomimetic antibiotic bleomycin (reviewed in Ref. 23) also led to chromatin condensation (Fig. 1B). We have also found in this in vitro system that the extent of condensation correlated with the degree of DNA fragmentation and that condensation could be observed within 5 min if high levels of nuclease were used; endonuclease G, another protein involved in apoptotic DNA breakdown (24), was also effective in inducing chromatin condensation (data not shown). Chromatin condensation could also be induced in nuclei of cells treated with nuclease in situ after cytoplasmic membrane perforation (Fig. 1C). Differences in the morphology of condensation between apoptotic cells (Fig. 1B, panels 2 and 3) and nuclease-treated nuclei or cells (Fig. 1, B, panels 5-9; and C, panel 2) were evident, including a lower background of DNA staining between condensed chromatin clusters and more irregularly shaped nuclear envelopes in the apoptotic cells. Nevertheless, in response to DNA breakdown per se in this in vitro system, it is significant that the morphology of chromatin condensation closely resembles peripheral chromatin condensation visible in apoptotic cells (late stage I or early stage II as defined in Refs. 25 and 26).

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Comparison of DNA fragmentation and chromatin condensation patterns between HL-60 cells undergoing apoptosis and nuclei purified from non-apoptotic cells. A, DNA from cells exposed in culture to etoposide or cisplatin or from isolated nuclei incubated with activated DFF, MNase, DNase I, AluI, or bleomycin (BLM) was analyzed by 1.5% agarose gel electrophoresis. B, shown is a micrograph illustrating the morphology of nuclei from apoptotic cells or purified nuclei treated with nucleases or bleomycin in vitro. C, shown are the results from the analysis of DNA fragmentation and morphology of nuclei of HL-60 cells treated with MNase after permeabilization with lysolecithin. Scale bar = 5 µm. The percentage of nuclei exhibiting chromatin condensation is indicated below each panel (also for Figs. 2-5). Except for Fig. 4, S.D. values were below ±5% of the mean values and are not shown.

Chromatin Condensation Is Energy-independent and Requires Physiological Concentrations of Mono- and Divalent Cations-- Mitotic chromosome condensation requires ATP hydrolysis (15), and an ATP requirement has been found for certain apoptotic cell extracts to trigger chromatin condensation (27). We therefore investigated the energy requirements for condensation in nuclei isolated from non-apoptotic HeLa cells. Nucleosomal DNA laddering mediated by either DFF or MNase was independent of the presence of ATP, ATPS, or apyrase (Fig. 2A, lanes 3-6 and 8-11), but was blocked by 1 mM ZnCl₂ (lanes 2 and 7), a known potent inhibitor of nucleases (11, 21). In addition, varied KCl concentrations slightly modulated such laddering (low [KCl] stimulated MNase, but decreased DFF activity) (Fig. 2C). Chromatin condensation in these same samples was manifested only when DNA laddering had occurred and was not inhibited by ATPS or apyrase (Fig. 2, A and B) or AMP-PNP (data not shown). Furthermore, condensation mediated either by DFF or MNase was less evident at lower KCl concentrations (Fig. 2D, compare panels 1 and 2 and panels 3 and 4) and could be largely reversed by addition of 10 mM EDTA to digested nuclei (Fig. 2D, lower panels). Condensation could also be induced upon shifting MNase-digested nuclei from 10 to 100 mM KCl (data not shown). The observed reversibility of condensation to changes in the concentrations of mono- and divalent ions is not unexpected because previous studies have shown that even isolated mononucleosomes bearing histone H1 can be quantitatively precipitated by exposure to either 100 mM KCl or 5 mM MgCl₂ and subsequently reversibly quantitatively solubilized by exposure to low ionic strengths and EDTA (28, 29). Nevertheless, we conclude that coalescing of fragmented chromatin into condensed clumps is energy-independent and optimal at physiological ionic strengths.

View larger version (50K): [in this window] [in a new window]   Fig. 2.   Energy and ionic strength requirements for chromatin condensation. A and B, DNA fragmentation and chromatin condensation, respectively, in nuclei isolated from normal HeLa cells treated with activated DFF or MNase in the presence or absence of Zn2+, ATP, or ATPS or pretreated with apyrase; C and D, DNA fragmentation and chromatin condensation, respectively, in nuclei treated with activated DFF or MNase in the presence of either 100 or 10 mM KCl as indicated. To remove Mg2+, nuclease-treated nuclei (C and D, lanes/panels 1 and 3) were incubated for 15 min on ice with 10 mM EDTA before fixing with formaldehyde. Scale bar = 5 µm.

Chromatin Condensation Is Blocked by Stabilization of Internal Nuclear Components-- In considering the mechanism of chromatin condensation in our in vitro system, it seems clear that digested chromatin fragments need to be able to diffuse within isolated nuclei to coalesce into clumps of condensed chromatin. However, nuclei within living cells are known to be organized into subcompartments, which have established territories for individual chromosomes and limit chromatin segment movement to nuclear subregions (Ref. 30; reviewed in Refs. 12 and 31-33). Possibly, this internal nuclear organization becomes altered during the processes of cell lysis and nuclei isolation such that intranuclear diffusion of chromatin fragments becomes possible. Indeed, during apoptosis in cultured cells, the accompanying proteolysis of nuclear lamins and other nuclear proteins leads to disorganization of nuclear subcompartments (25, 34, 35).

To test the idea that dissolution of internal nuclear structure is required for chromatin condensation in our in vitro system, we took advantage of the observations made by Laemmli and co-workers (36-38) that reversible treatment with Ca²⁺ or Cu²⁺ stabilizes DNA attachment sites both in HeLa cell mitotic chromosomes and interphase nuclei (reviewed in Refs. 39-41). Indeed, it has recently been demonstrated that cellular chromatin in vivo possesses unexpectedly high concentrations of Ca²⁺ (4-9 mM for interphase nuclei and 20-32 mM for mitotic chromosomes), which are at 3-fold higher levels than those of Mg²⁺ ions; furthermore, such Ca²⁺ is enriched in the chromosomal axis and co-localizes with the scaffolding proteins topoisomerase II and ScII (42). We also used two other techniques to stabilize internal nuclear components: exposure of nuclei to mild oxidation to generate disulfide bonds within and between protein species (reviewed in Refs. 39-41) and exposure of nuclei to pH 4.0 followed by neutralization to pH 7.5 (our empirical observation). As shown in Fig. 3A, these pretreatments did not significantly disrupt chromatin structure as revealed by the patterns of DNA laddering generated by subsequent digestion with MNase (Fig. 3A, lanes 2-5) or DNase I (data not shown). However, such pretreatments severely inhibited subsequent chromatin condensation (Fig. 3B, panels 2-5). (For these and additional experiments, for convenience, we primarily employed MNase as the nuclease used to trigger condensation.) In addition, chromatin condensation in control nuclei could not be reversed if any of these treatments occurred after nuclease digestion (data not shown).

View larger version (56K): [in this window] [in a new window]   Fig. 3.   Nuclease-induced chromatin condensation is inhibited by stabilizing nuclear components. A and B, DNA fragmentation and chromatin condensation, respectively, in nuclei isolated from normal HeLa cells treated with MNase after reversible exposure to CaCl2, CuCl2, diamide, or pH 4. Scale bar = 5 µm. C, SDS-PAGE analysis of proteins extracted from nuclei after reversible exposure to stabilizing agents. Lane labels are as follows: T, total nuclear proteins; 1, 0.2 M NaCl extracts; 2, 0.35 M NaCl extracts. The percentage of extracted proteins is indicated below the stained gel.

Significantly, all pretreatments that blocked chromatin condensation in isolated nuclei affected the extractability of nuclear proteins by 0.2 and 0.35 M NaCl. Reversible treatment with CaCl₂, diamide, or low pH significantly reduced the amount of released protein, whereas CuCl₂ pretreatment almost totally prevented nuclear protein release (Fig. 3C). In conclusion, enhancement of metalloprotein interactions, induction of disulfide bond formation, and apparent protein denaturation caused by reversible low pH treatment each apparently stabilize internal nuclear components and thereby disallow digested chromatin fragments the intranuclear mobility required for coalescence into clumps of condensed chromatin.

Evidence for Nuclear Actin Involvement in Chromatin Condensation-- Actin is among the candidate filament proteins that may participate in nuclear subcompartmentalization (reviewed in Ref. 41). To test directly for actin involvement in chromatin condensation, we employed the drugs latrunculin A and phalloidin, which depolymerize and polymerize actin filaments (F-actin), respectively (43, 44). As shown in Fig. 4A, treatment of nuclei with these reagents did not significantly disrupt chromatin structure as revealed by the patterns of DNA laddering. Actin is a known inhibitor of DNase I, which depolymerizes F-actin (45). Significantly, treatment of nuclei with phalloidin markedly increased the sensitivity of chromatin to DNase I digestion (Fig. 4A); reversible exposure of nuclei to pH 4 (but not pretreatment with diamide, CaCl₂, or CuCl₂) also sensitized nuclei to DNase I digestion (data not shown). Furthermore, phalloidin treatment partially inhibited chromatin condensation (Fig. 4B, panel 2). It is known that phalloidin will cause the formation of F-actin from G-actin (44). However, the partial inhibition caused by phalloidin was apparently not due to causing this conversion because a similar extent of inhibition was observed when latrunculin A was added at the same time to block F-actin formation (Fig. 4B, panel 4). Although such inhibition was detected in only ~30% of the nuclei, it proved to be highly reproducible and statistically significant (p < 0.05). In addition, use of jasplakinolide, another reagent known to stabilize F-actin (46), gave the same degree of partial inhibition of condensation (data not shown). Furthermore, chromatin condensation in control nuclei could not be reversed if the phalloidin treatment occurred after nuclease digestion (data not shown). Interestingly, if the nuclei were first extracted with 0.35 M salt, a condition leading to ~60% loss of nuclear actin, chromatin condensation could no longer be partially inhibited by phalloidin (data not shown). We also analyzed nuclear proteins released from MNase-digested nuclei by SDS-PAGE and actin by Western blotting (Fig. 4, C and D). Concomitant with the condensation process, we found that >80% of the nuclear actin was lost from control nuclei or nuclei pretreated with latrunculin A (Fig. 4, C and D, lanes 1, 2, 5, and 6). Significantly, such a loss was blocked by reversible exposure to pH 4 (Fig. 4D, lanes 9 and 10) as well as by pretreatment with CaCl₂ (lanes 7 and 8), CuCl₂ (lanes 11 and 12), or diamide (lanes 13 and 14). Interestingly, phalloidin treatment only partially blocked actin release (Fig. 4, C and D, lanes 3 and 4). This correlates with the observation that phalloidin treatment only fractionally inhibited condensation, unlike pH 4, CuCl₂, or diamide. Taken together, these results suggest that F-actin partially restricts chromatin fragment mobility for condensation.

View larger version (82K): [in this window] [in a new window]   Fig. 4.   Partial inhibition of chromatin condensation by blocking F-actin depolymerization. Shown is the DNA fragmentation (A) and chromatin condensation (B) in isolated nuclei incubated with 10 µM phalloidin and/or latrunculin A and then digested with MNase. Nuclei treated with phalloidin and latrunculin A or reversibly exposed to low pH, CaCl2, CuCl2, or diamide were digested with MNase and then pelleted by low speed centrifugation. Pelleted nuclei (P) and corresponding supernatants (S) were analyzed by either SDS-PAGE (C) or Western blotting with anti-actin antibody (D). Shown is the DNA fragmentation (E) and chromatin condensation (F) in HL-60 cells treated with etoposide in the presence of 0.3 µM jasplakinolide and latrunculin A. Cells were labeled with BrdUrd prior to etoposide/jasplakinolide treatment, and then nuclei were stained with both 4,6-diamidino-2-phenylindole (DAPI) and anti-BrdUrd antibody (Texas Red-conjugated secondary antibody) (G). Scale bars = 5 µm (B) an 20 µm (F and G). Numbers below panels are the means ± S.D. for three to five independent experiments.

We also assayed actin-targeting reagents for their ability to block chromatin condensation in cultured HL-60 cells triggered to undergo apoptosis by etoposide. We selected jasplakinolide to stabilize F-actin because it is much more permeable to cells than phalloidin (46) and latrunculin A, which targets G-actin (43). Because we found that the drugs were quite toxic to cells, we had to perform experiments with reagents diluted to 0.3 µM. Under these conditions, we found that jasplakinolide (with or without latrunculin A) treatment did not affect DNA laddering induced by etoposide (Fig. 4E, lanes 5 and 6), whereas these drugs reduced chromatin condensation by ~20% compared with the etoposide-treated control (70 and 69% versus 89%) (Fig. 4F, panels 4-6). Jasplakinolide by itself also induced minor DNA breakdown (Fig. 4E, lane 2), which is in agreement with a previous report that the reagent enhances apoptosis (47). Nevertheless, we have evidence both in vitro and in vivo that F-actin apparently partially restricts chromatin fragment mobility for condensation.

The observed partial inhibition of chromatin condensation by drugs that stabilize F-actin may be related to cell cycle events that modulate nuclear actin (levels, post-translational modifications, organization, and/or associated proteins). One might expect that nuclei of S phase cells may be less restrictive to chromatin mobility because this period corresponds to the time when chromosome domains must become repositioned to immobile replication centers for their subsequent DNA replication (48, 49). Therefore, we analyzed the extent of chromatin condensation in nuclei of cells containing newly replicated DNA. HL-60 cells growing at log phase with a ~25-h doubling time were pulse-labeled with BrdUrd for 10 h immediately before treatment with etoposide and jasplakinolide during the chase. Under these conditions, ~35% of the cells had incorporated BrdUrd. Significantly, in this cell population, ~50% of the nuclei with condensed chromatin contained newly synthesized DNA, whereas only ~5% of the cells that lacked condensed chromatin were labeled with BrdUrd (Fig. 4G). Therefore, stabilization of F-actin is primarily inhibitory to apoptotic chromatin condensation in non-S phase cells.

Agents That Target the DNA Minor Groove Block Chromatin Condensation-- To dissect further the mechanism of chromatin condensation in our in vitro system, we searched empirically for other reagents that could uncouple condensation from DNA fragmentation at physiological ionic strengths. We selected for study agents that bind to or intercalate into either the minor or major groove of DNA, including distamycin A and bisbenzimide (Hoechst 33258), which preferentially bind to the minor groove of AT-rich DNA (50, 51); chromomycin A₃, which preferentially binds to the minor groove of GC-rich DNA (52, 53); ethidium bromide, which intercalates into the DNA minor groove (54); and methyl green, which binds to the DNA major groove (55).

As shown in Fig. 5, several of these reagents blocked condensation without inhibiting DNA fragmentation. With the exception of ethidium bromide (Fig. 5A, lane 7), incubation of nuclei with any of the DNA-binding compounds did not markedly disrupt chromatin structure as revealed by the patterns of DNA laddering on a low resolution agarose gel (Fig. 5A). Because ethidium bromide is known to perturb chromatin structure (56), this result was not unexpected. However, both distamycin A and bisbenzimide, but not chromomycin A₃ or ethidium bromide, affected the rate of digestion (Fig. 5A). Significantly, distamycin A, chromomycin A₃, bisbenzimide, and ethidium bromide, all minor groove-targeting drugs, each severely blocked chromatin condensation (Fig. 5B, panels 3-5 and 7). In marked contrast, incubation with the major groove binder methyl green or with the topoisomerase II inhibitors etoposide, VM-26 (teniposide), and 4'-(9-acridinylamino)methanesulfon-m-anisidide did not inhibit subsequent condensation (Fig. 5B, panels 6 and 8; data not shown). Furthermore, addition of minor groove reagents after DNA fragmentation did not reverse chromatin condensation (Fig. 5B, panels 9-11), whereas chromatin condensation was still inhibited by DNA-binding drugs if nuclei were first extracted with 0.35 M NaCl (data not shown). We conclude that interactions that involve the DNA minor groove are fundamentally required to trigger chromatin condensation in our in vitro system, regardless of DNA sequence preferences.

View larger version (62K): [in this window] [in a new window]   Fig. 5.   Effect of DNA groove-specific binding or intercalating agents on chromatin condensation. Shown is the DNA fragmentation (A) and chromatin condensation (B) in nuclei isolated from normal HeLa cells treated with MNase in the presence of 80 µM distamycin A, chromomycin A3, bisbenzimide (Hoechst 33258), methyl green, ethidium bromide, or etoposide. Alternatively, drugs were added for 15 min after incubation with the nuclease (post-treatment). Scale bar = 5 µm. Shown is the inhibition of chromatin condensation in MNase-digested nuclei cotreated with different concentrations of DNA-binding drugs (C). DNA from nuclei incubated with 50 µM ethidium bromide, chromomycin A3, or distamycin A or 100 µM methyl green and digested with MNase (D) or DNase I (E) was analyzed on native 6% acrylamide or denaturing 7 M urea and 7.5% acrylamide (10:1 acrylamide/bisacrylamide ratio) gels, respectively.

Analysis of the dose response of the inhibitory reagents revealed that 50% inhibition of condensation occurred between 20 and 50 µM drug (Fig. 5C). We have quantitated spectrophotometrically the amount of these reagents that bound to nuclei at 50% inhibition and estimate that if all reagent that is bound to nuclei is associated with DNA, then 1 molecule of drug would exist for every 10-20 bp. One would expect that DNA wrapping about and within nucleosomes might be disrupted at these levels of drug, although we did not see any disruption of the nucleosomal repeat on a low resolution gel, with the exception of ethidium bromide treatment (Fig. 5A). However, analysis of digestion products on a higher resolution acrylamide gel revealed that the 146-bp barrier, which is a characteristic of the nucleosome core particle (57), was lost at drug concentrations that block condensation, and there was a higher internucleosomal band background (Fig. 5D), indicating that the drugs increased nuclease cleavage within nucleosomes. To examine alterations in DNA wrapping caused by these reagents in more detail, we assayed for the intactness of the ~10-base subnucleosomal DNA repeat that is generated by DNase I digestion of chromatin (57, 58). As shown in Fig. 5E, this subnucleosomal DNA repeat was also disrupted by the inhibitory drugs, as evidenced by a higher interband background. By contrast, no detectable disruption in nucleosomal DNA wrapping was observed at even 100 µM methyl green, a concentration that we estimate from what bound to nuclei would represent roughly 1 molecule of drug for every 20 bp (Fig. 5E). Although similar concentrations of minor groove drugs were toxic to HL-60 cells, at lower doses, they were partially inhibitory to etoposide-induced apoptotic condensation (data not shown). We conclude that intact nucleosomes are required for chromatin condensation in our system.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Conflicting reports exist in the literature on whether DNA cleavage of nuclei isolated from non-apoptotic cells by any nuclease can lead to chromatin condensation nearly mimicking that seen in apoptotic cells (2, 5, 33, 59, 60). If this indeed were true, then a model in vitro system could be established to dissect the requirements for this process. Here, we have successfully developed such a system. We provide also an explanation for discrepancies in the literature because we have demonstrated that such condensation requires mono- and divalent metal ions at physiological ionic strengths, conditions not always used by other investigators. It should also be noted that the nature of DNA fragment ends is not important, as condensation can occur with components possessing blunt ends, with 5'- or 3'-overhangs, and with either 5'- or 3'-phosphate groups.

Condensation Apparently Requires Disruption of Nuclear Subcompartmentalization-- We have evidence that chromatin condensation in our in vitro system requires intranuclear mobility mediated by the disruption of components that may participate in the subcompartmentalization of chromosome subdomains in living cells. We found that enhancement of metalloprotein interactions, induction of disulfide bond formation, and apparent protein denaturation caused by reversible low pH treatment each were effective in blocking chromatin condensation. These treatments inhibit condensation by a mechanism different from that of the drugs that target the DNA minor groove because we have found that the subnucleosomal DNase I-generated ladder remained intact in such treated nuclei (data not shown). In only one of these four conditions (low pH), pre-extraction of nuclei with 0.35 M NaCl eliminated the block in apparent intranuclear mobility (data not shown).

Although it can be argued that these stabilization treatments are quite harsh and may lead to artifacts, stabilization of nuclear F-actin led to partial inhibition of chromatin condensation both in vitro and in vivo, strongly supporting the notion that disruption of intranuclear organization is indeed required for chromatin condensation. Furthermore, leakage of nuclear actin was markedly reduced by pH 4, Ca²⁺, Cu²⁺, or diamide pretreatments, conditions that each severely inhibited chromatin condensation. Interestingly, previous microdissection experiments with Xenopus oocytes have revealed that even after mechanical removal of the nuclear envelope, chromosome organization is maintained by a "nucleoplasmic gel" that contains both F-actin and G-actin as major components (61, 62). Indeed, it is now appreciated that nuclear actin plays roles in chromatin remodeling and RNA transcript trafficking (63, 64). Furthermore, certain actin-binding proteins interact with a fibrous network in the nucleoplasm (64, 65). We propose that nuclear actin also participates in establishing chromosomal territories that restrict chromatin fragment mobility. However, stabilization of F-actin was ineffective in blocking chromatin condensation in cells replicating their genomes. This suggests that the apparent intranuclear mobility of chromatin is less restricted by F-actin during S phase of the cell cycle. It is of further interest that actin is a known substrate for caspase-3 during apoptosis (66).

Condensation Likely Requires Preservation of Native Nucleosomal DNA Wrapping through Core Histone-DNA Minor Groove Contacts-- It is striking that agents that target the minor groove of DNA inhibit chromatin condensation both in vitro and in vivo. These agents inhibit condensation by a pathway not linked to actin retention (data not shown). Distamycin A and bisbenzimide preferentially bind to the minor groove of AT-rich DNA, eliminate the known preferential association of histone H1 with AT-rich DNA (67, 68), and cause decondensation of heterochromatin (69, 70). However, chromomycin A₃, which preferentially binds to the minor groove of GC-rich DNA, was also effective in blocking condensation. We have searched by SDS-PAGE for proteins that might be released from nuclei by ethidium, distamycin A, or chromomycin A₃ treatment, but found no evidence for release or weakened interactions with DNA of any protein species under conditions that blocked condensation (data not shown). Nuclease protection assays have revealed, however, that these inhibitory drugs alter nucleosomal DNA wrapping. The globular domain of histone H1 binds to the major groove of DNA (71, 72), whereas core histones bind to the minor groove (73). Native nucleosomal DNA wrapping through core histone interactions therefore seems to be required for nucleosomes to assemble into higher order structures that lead to chromatin condensation. This wrapping is known to neutralize positively charged amino acids by a close association with negatively charged DNA phosphate groups. However, simple electrostatic repulsion between chromatin fragments cannot be the mechanism behind the inhibition of condensation mediated by these drugs because neither distamycin A nor ethidium bromide (at up to 200 µM concentrations) prevented precipitation of isolated oligonucleosomes (1-8-mers) induced by 4 mM MgCl₂ and 100 mM KCl (data not shown). Thus, it appears that these drugs affect the ability of chromatin fragments to coalesce, possibly by promoting interactions between the drug-induced exposed positively charged amino acids and acidic non-histone proteins, which may interfere with internucleosome interactions and/or fragment diffusion. It is interesting to reflect from an evolutionary view that nucleosome structure not only fulfills crucial regulatory and packaging roles in living cells, but also prepares apoptotic cells for the efficient clearance of DNA by phagocytosis during cell death. Although not known at present, we favor the notion that the condensation process that we observe represents a specific form(s) of stacking of nucleosomes into supramolecular structures, rather than nonspecific aggregation phenomena. Indeed, it has been recently shown that at a high concentration of nucleosome core particles (350 mg/ml), which is similar to their intranuclear level, nucleosomes stack into aligned columns that form bilayers constituting a liquid crystalline state (74).

Extensive earlier investigations have demonstrated that histone H1 facilitates higher order chromatin packing (Ref. 75; reviewed in Ref. 76) and confers insolubility even to mononucleosomes at physiological ionic strengths (28, 29). The observed ionic strength dependence on chromatin condensation is consistent with a role for histone H1 in this process. Furthermore, it is known that histone H1-depleted nucleosomes leak out of MNase-digested nuclei at physiological ionic strengths (77). To determine whether condensation could be inhibited by depleting nuclei of histone H1 prior to DNA fragmentation, we extracted nuclei at low pH (22), with polyglutamic acid (68), or with 0.6 M NaCl. However, in our hands, we found that these procedures were either inefficient or disruptive to nuclear integrity (data not shown).

Neither Topoisomerase II Activity nor ATP Is Required for Condensation-- Distamycin A is also known to inhibit the preferential interaction of topoisomerase II with AT-rich DNA (78). Topoisomerase II has been associated with mitotic chromosome condensation in normal cells (79-81), with heterochromatin formation in interphase nuclei (82), with mediation of DNA aggregation in vitro (78, 83), with chromatin condensation triggered by apoptotic cell extracts (84), and with higher order DNA fragmentation during apoptosis (85). In addition, previous studies have shown that ATP is required for chromatin condensation, but not for internucleosomal DNA fragmentation, when apoptotic cell extracts are added to isolated nuclei from non-apoptotic cells (27). However, inhibition of topoisomerase II activity or depletion of ATP was ineffective in blocking chromatin condensation in our experiments with isolated nuclei. Therefore, we suggest that in these earlier studies, the topoisomerase II and ATP requirements may play a role in eliminating inhibitors present in apoptotic cell extracts.

	ACKNOWLEDGEMENTS

We thank Sveta Earnst, Michael Hale, Katherine Meyers, and Ying Zou for help in this work.

	FOOTNOTES

^* This work was supported in part by Grant 6P04A01317 from the Polish Committee for Scientific Research KBN (to P. W.), Grant GMRO1-59809 from the National Institutes of Health, and Grant I-0823 from the Robert A. Welch Foundation (to W. T. G.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence and reprint requests should be addressed: Dept. of Molecular Biology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75390-9148. Tel.: 214-648-1924; Fax: 214-648-1915; E-mail: william.garrard@utsouthwestern.edu.

Published, JBC Papers in Press, April 1, 2002, DOI 10.1074/jbc.M201027200

	ABBREVIATIONS

The abbreviations used are: DFF, DNA fragmentation factor; PBS, phosphate-buffered saline; BrdUrd, bromodeoxyuridine; MNase, micrococcal nuclease; ATPS, adenosine 5'-O-(3-thiotriphosphate); AMP-PNP, 5'-adenylylimidodiphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES