[CANCER RESEARCH 52, 1580-1586, March 15, 1992]
Influence of Chromatin Structure on the Induction of DNA Double Strand Breaks
by Ionizing Radiation
Michael C. Elia and Matthews O. Bradley1
Department of Genetic and Cellular Toxicology, Merck Sharp & Dohme Research Laboratories, West Point, Pennsylvania 19486
ABSTRACT
Pulsed field gel electrophoresis was used to examine the influence of
chromatin structure on the induction of DNA double strand breaks by 7irradiation in CHO-WBL cells, nuclei, and a series of protein-depleted
chromatin substrates. We developed a method to isolate intact nuclei in
agarose plugs that avoids DNA shearing and nucleolytic degradation
during sample preparation, and facilitates nuclear protein extraction.
Agarose plug-isolated nuclei are extracted with increasing concentrations
of NaCl to selectively strip off: (a) nonhistone chromosomal proteins
(NHP); (b) NHP and histone HI; (c) NHP, HI, and histone H2A-H2B
dimers; or (d) NHP, HI, and H2A-H2B dimers and histone H3-H4
tetramers. Following treatment with up to 40 Gy of -y-radiation, DNA
from each sample is purified and the relative induction of DNA double
strand breaks is assayed by asymmetric field inversion gel electrophore
sis. At a dose of 20 Gy, removal of nonhistone proteins from nuclei
results in a 3-fold increase in DNA double strand breaks, compared to
intact CHO cells. Additional stripping of histone HI results in an
incremental increase in double strand break induction, whereas further
removal of H2A-H2B dimers yields a greater than 10-fold increase in
DNA double strand breaks compared to intact CHO cells. The doseresponse profile for this latter sample is similar to that observed for
purified DNA. These data indicate that distinct classes of chromosomal
proteins afford the DNA with different levels of protection against 7ray-induced DNA double strand breaks. Thus, chromatin domains that
differ in tertiary structure and protein composition may also differ in
their susceptibility to DNA double strand breaks induced by ionizing
radiation and, perhaps, other clastogens.
sent a continuing challenge for molecular biologists.
The existence of structural heterogeneity within eukaryotic
chromatin suggests that different regions of the genome may
be more or less susceptible to clastogenic damage, depending
upon the relative accessiblity of the clastogen to a particular
genetic domain. Thus, DNA compacted into higher order so
lenoids within a heterochromatic gene would be expected to be
less susceptible to damage compared to a highly decondensed,
transcriptionally active gene. Using high resolution cytogenetic
techniques, Yunis et al. (2) provided evidence for the nonrandom interaction of xenobiotics with eukaryotic DNA. These
authors showed that a diverse array of chemical and physical
mutagens induce a limited and recurrent set of DNA breaks,
termed chromosomal fragile sites. Interestingly, several of the
tested agents, including actinomycin D, bleomycin, diethylnitrosamine, benzo(a)pyrenediolepoxide,
dimethylsulfate, bromoacetaldehyde, and radiation, have been reported to prefer
entially damage DNA within transcriptionally active genes (Ref.
2 and references cited therein).
The DNA DSB2 is a particularly dangerous lesion for the
In mammalian cells, different regions of interphase chromatin exist in a variety of structurally heterogeneous forms, reflect
ing the presence of a system of hierarchical regulatory controls
that can either facilitate or restrict functional access to partic
ular chromosomal domains. The presence of topologically dis
tinct looped domains within the nucleus provides a discrete
control system for modulating the activity of neighboring ge
netic elements by differentially regulating the superhelical den
sity of individual domains. The association of different histone
and nonhistone chromosomal proteins with DNA can dramat
ically affect the topologica! state and accessibility of particular
chromosomal domains. For example, the structure of transcriptionally active (or potentially active) chromatin is generally
considered to be more accessible to soluble factors, compared
to transcriptionally inert, compacted heterochromatin. This
increased accessibility has been correlated with increased nucleosomal histone acetylation, decreased levels of histone HI,
the presence of nucleosome-free control regions, increased lev
els of high mobility group proteins 14 and 17, as well as the
presence of histone variants (reviewed in Ref. l). Unfortunately,
the molecular mechanisms by which these modifications partic
ipate in chromosomal unfolding are not yet known, and repre-
cell, since the resultant physical discontinuity of the target
chromosome can lead to the loss of information contained
within the disrupted gene. Theoretically, one unrepaired or
misrepaired DSB in the functional copy of a required gene can
result in cell death (3). Alternatively, nonlethal misrepair of the
break, for example by error-prone recombination or ligation
systems, may enhance the probability of survival, albeit at the
cost of cellular mutation (4, 5), and possibly, cellular transfor
mation (6, 7). Since DNA DSB occur as a result of exposure to
ionizing radiation and clastogenic chemicals, as well as during
repair of DNA damage from UV and some alkyating agents
(8), it is important to understand the role chromatin structure
plays in DNA damage and repair.
To study the influence of different levels of chromatin struc
ture on the susceptibility of mammalian DNA to clastogenic
damage, we measured the relative frequency of radiation-in
duced DNA DSB in Chinese hamster ovary cells, nuclei, and a
series of protein-depleted chromatin substrates. The protein
extraction procedure used results in substrates with increasingly
decondensed chromatin. AFIGE was used to assay the fre
quency of radiation-induced DNA DSB in the various samples.
The AFIGE DSB assay has several advantages over other
techniques for measuring DNA DSB, such as neutral filter
elution, including greater sensitivity for detecting damage at
low doses, and the ability to obtain DNA fragment size infor
mation under appropriate electrophoretic conditions (9). Like
the neutral filter elution assay, however, the AFIGE assay yields
an indirect measure of DNA DSB, namely the fraction of DNA
that enters the gel (reviewed in Ref. 10). Treatment of mam
malian cells with ionizing radiation leads to a dose-dependent
Received 7/29/91; accepted 12/26/91.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Present address: Genetic Medisyn, 9620 Medical Center Drive, Rockville,
MD 208SO.
2The abbreviations used are: DSB, double strand breaks; AFIGE, asymmetric
field inversion gel electrophoresis; PBS, phosphate-buffered saline; PFG, pulsed
field gel electrophoresis; PACE, programmed autonomously controlled elec
trodes; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis;
CHEF, contour clamped homogeneous electric field; NHP, nonhistone chromo
somal proteins; F.A.R., fraction of radioactivity released.
INTRODUCTION
1580
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
CHROMATIN
STRUCTURE
AND DNA DOUBLE STRAND BREAKS
increase in the fraction of DNA that enters the gel; DNA from
untreated control samples fails to enter the gel and remains at
the origin. When necessary, specific assay conditions can be
calibrated to determine the relationship between the fraction of
DNA entering the gel and the absolute number of DNA DSB
present in a population of cells (11).
MATERIALS
AND METHODS
Cell Culture and DNA Labeling. Chinese hamster ovary fibroblasts,
free of Mycoplasma, were grown in the absence of antibiotics in Dulbecco's moilifiul Eagle medium supplemented with 10% fetal bovine
serum. In this study, we used CHO-WBL cells, a strain characterized
by a relatively stable karyotype that is routinely used in cytogenetic
testing. To radiolabel DNA, exponentially growing cells were incubated
in the presence of 0.02 ^Ci/ml [14C]thymidine (40 Ci/mol; New England
14°C.Asymmetric field inversion gel electrophoresis was carried out
for 36-40 h using repeating cycles of a 900-s pulse of 1.25 V/cm in the
direction of net DNA migration, followed by a 75-s pulse of 5 V/cm in
the reverse direction. For each experimental condition, triplicate sam
ples were analyzed, and the entire dose-response experiment was re
peated 4 times, twice using the electrophoretic conditions described
here and twice using alternative conditions. The gel was stained with
ethidium bromide and photographed under UV illumination, prior to
removal of individual sample wells and lanes for radioactivity measure
ments. The fraction of radioactivity released from the well and migrat
ing into the lane gives an indirect measure of DNA DSB. Since this
work is concerned with the relative induction of DNA DSB in various
cellular and chromatin substrates, an absolute measure of DNA DSB
is not necessary, and no '"I-labeled decay calibration experiments are
reported. See Iliakis et al. (11) for calibration of the AFIGE/DSB assay
under conditions that are similar to those used here. Molecular weight
standards routinely used include Schizosaccharomyces pombe (3.5-5.7
megabases), Saccharomyces cerevisiae (0.2-2.2 megabases), and X-ladders (48.5-1100 kilobases); all were purchased from Bio-Rad.
Analysis of DNA Fragment Size Using PACE/PFG. Megabase size
DNA fragments from irradiated cells, nuclei, and the various chromatin
substrates were analyzed using a CHEF Mapper instrument (Bio-Rad)
that utilizes PACE technology (13) to yield enhanced speed, resolution,
and design flexibility compared to a standard CHEF system (reviewed
in Ref. 10). Agarose plugs containing purified DNA from irradiated
samples were loaded into the wells of a 0.8% agarose gel prepared and
run in 0.5 x Tris-Acetate-EDTA buffer. The gel was run at 14°Cat 2
V/cm for 44.27 h with a switch angle of 106°,with switch times linearly
Nuclear) until they reached the desired density, then fresh medium
without radiolabel was added for at least 12 h prior to harvesting.
Preparation of Nuclei and Protein-depleted Chromatin Substrates.
Radiolabeled CHO-WBL cells were washed 3 times with ice-cold PBS
without calcium or magnesium, and dislodged from the substratum by
treatment with 0.01% trypsin in PBS. The cells were washed twice in
medium containing 10% fetal bovine serum, resuspended at a concen
tration of 5 x IO6cells/ml in 0.8% Insert agarose (FMC Bioproducts)
in PBS at 37°C,poured into plastic molds (Bio-Rad), and allowed to
solidify for 3-4 min on ice. Individual agarose plugs (3-5) were then
transferred to 50-ml polypropylene tubes containing 30 ml of ice-cold
buffer A (10 mM Tris-HCl, pH 7.6, 140 HIMNaCl, 1 mM MgCl2). After
ramped from 20 to 40.4 min, after which the field strength was increased
to 6 V/cm with a switch angle of 120°,and the switch times linearly
10 min on ice, the buffer was replaced with either fresh buffer A (for
cells) or buffer A supplemented with 0.5% Triton X-100 (for nuclei and
ramped from 6.91 s to 1 min and 53.86 s over a period of 4.1 h. After
the various protein-depleted chromatin substrates). After 20-30 min,
staining with ethidium bromide, the gel was photographed under UV
the plugs were washed 3 times with buffer A to remove residual
illumination with Polaroid type 55 positive/negative film. The molec
detergent. To prepare protein-depleted chromatin substrates, separate
ular weight markers used were the same as described above for AFIGE.
sets of plugs were incubated in buffer A containing a total of either
SDS-PAGE. Agarose plugs containing the various chromatin sam
0.35 M NaCl, 0.6 M NaCl, 1.2 M NaCl, or 2.0 M NaCl (see Fig. 2). ples were incubated for 20 min at 37°Cwith 1000 units of DNase I per
After 60 min on ice with intermittent mild agitation, the buffer was plug in a total volume of 0.25 ml. Plugs containing nuclei were treated
replaced and the incubation was continued overnight. Plugs were
instead with 5000 units of DNase I. One-tenth volume of 0.5 M EDTA,
washed once more in the appropriate buffer before incubation in PBS
pH 8.0, was then added to halt the reaction. Next, SDS-PAGE sample
and subsequent -y-irradiation. Phenylmethylsulfonylfluoride was added
buffer was added, and the mixture was boiled for 2 min immediately
freshly to each buffer throughout the procedure to inhibit proteolysis.
prior to electrophoresis on 10-20% gradient minigels (Integrated Sep
To prepare purified DNA, cells embedded in agarose plugs were washed
briefly in buffer A, incubated overnight at 50°Cin lysis buffer (0.5 M aration Systems) at 200 V; electrophoresis was continued until the
tracking dye migrated off the gel (about 2 h).
EDTA, pH 8.0, 1% AMaurylsarcosine, l mg/ml proteinase K; Boehringer Mannheim Biochemicals), and then washed with 10 mM Tris-HCl,
1 mM EDTA, pH 8.0 prior to irradiation or electrophoresis.
RESULTS
Irradiation and Preparation of Samples for Pulsed Field Gel Electro
Preparation of Nuclei and Protein-depleted
Chromatin Sub
phoresis. Agarose plugs containing either CHO-WBL cells, nuclei, or
protein-depleted chromatin substrates were irradiated on ice in PBS at
strates in Agarose Plugs. To facilitate sample preparation and
a radiant flux of 15.83 Gy/min using a ' "Cs-irradiator (J. L. Sheppard).
avoid DNA shearing, intact nuclei were prepared from CHO
Immediately following •¿>
irnuliaiio». the agarose plugs were incubated
cells embedded in agarose plugs by treatment
with isotonic
in lysis buffer at 50°Cfor 12-16 h. Plugs were then washed extensively
buffer containing 0.5% Triton X-100. Light micrographs of 2in 10 mM Tris-HCl, l mM EDTA, pH 8.0, prior to electrophoresis.
nm fixed sections stained with méthylèneblue demonstrate that
The RNase A step that is commonly used during DNA purification for
treatment of CHO cells with 0.5% Triton X-100 results in the
pulsed field gel electrophoresis was omitted since it was found not to
loss of most of the soluble contents of the cytosol, while leaving
affect the results of the DNA double strand break assay.
some
cytoskeletal remnants and intact nuclei (Fig. 1). When
AFIGE/DSB Assay. The technique of Starnato and Denko (9) takes
performed on cells attached to tissue culture plates, this pro
advantage of the fact that fragmentation of chromosomal DNA with
cedure generates a "nuclear monolayer" that remains bound to
radiation leads to a dose-dependent increase in the fraction of DNA
the substratum via the cytoskeletal remnants (14-16). Nuclei
that enters the gel, whereas intact mammalian chromosomes are unable
to enter the gel. Electrophoresis was performed using conventional gel prepared in this manner exhibit no evidence of DNA degrada
boxes (model H4; Bethesda Research Laboratories) connected to a tion as assayed by asymmetric field inversion gel electrophoresis
switching apparatus of local construction (see Ref. 12 for electrical
(see below).
schematic) that regulated both the forward and reverse pulse times and
Nuclei in agarose plugs were extracted with various concen
voltages, from a standard power supply. Agarose gels (0.8%) were cast
trations of NaCl in order to generate chromatin samples de
in 75 mM Tris, 25 mM boric acid, 0.1 mM EDTA, pH 8.9, using
proteins (Fig. 2).
modified trays that can accommodate up to 80 samples each. After pleted of various classes of chromosomal
Denaturing
polyacrylamide
gel electrophoresis
was used to
insertion of the agarose plug into the sample well, molten agarose
assess the efficiency of the extraction procedure (Fig. 3). We
(0.8%) was overlayed to seal the plug in the well. Buffer was recirculated
noticed that nuclei prepared in plugs using Triton X-100 genthroughout the run, and the temperature was maintained at a constant
1581
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
CHROMATIN
STRUCTURE
AND DNA DOUBLE STRAND BREAKS
Fig. 1. Light micrographs of 2->im-thick sections of CHO-WBL cells embedded
in 0.8% agarose before (. I ) and after (B) treatment with buffer containing 0.5%
Triton X-100. The sections were stained with méthylène
blue.
erally had a lower NHP content compared to nuclei prepared
in solution by standard hypotonie cell lysis and Dounce homogenization (data not shown). However, since nuclei prepared
by this latter procedure had an unacceptably high level of DNA
double strand breaks as assayed by AFIGE, we decided to use
nuclei prepared using the Triton X-100 "in plug" procedure.
As seen in Fig. 3, treatment of nuclei with 0.35 M NaCl releases
the vast majority of nonhistone chromosomal proteins. Increas
ing the ionic strength to 0.6 M NaCl removes both NHP and
the bulk of histone HI, whereas 1.2 M NaCl removes NHP,
HI, and H2A-H2B tinners. Extraction of nuclei with 2.0 M
NaCl removes virtually all chromosomal proteins, including
the histone H3-H4 tetramer. The band seen in the nuclei and
ehrt »ma
tin samples at about M, 31,000 is the DNase I that is
added prior to SDS-PAGE (see "Materials and Methods" for
details). The identity of few proteins remaining in the plug
following treatment with 2.0 M NaCl is not known conclusively
at this time, although similar extraction procedures are known
to leave behind nuclear matrix proteins, such as the nuclear
lamins (17).
It should be noted that the salt extraction steps described
here apparently leave intact the underlying nuclear scaffold (18,
19). However, since chromatin exists in the nucleus as a series
of topologically isolated looped domains containing different
types and amounts of bound proteins, the protein depletion
protocol we used will have different effects on structurally
distinct domains, especially at the lower salt concentrations.
For example, a highly decondensed, transcriptionally active
domain that may be relatively depleted of nonhistone proteins
and histone HI is not expected to be as dramatically affected
by a 0.35 or 0.6 M NaCl extraction step as is a highly condensed,
NHP- and Hl-rich domain. Treatment of the latter domain
with 0.6 M NaCl would remove both NHP and HI, leading to
the selective decondensation of that particular domain. At
higher salt concentrations, however, the effects of further pro
tein removal are expected to be more uniform across what had
originally been structurally different domains. For example,
treatment with 1.2 M NaCl removes most NHP, HI, and H2AH2B dimers, leaving behind H3-H4 tetramers bound to the
DNA. Thus, unless a particular domain had originally been
completely protein-free in the untreated, intact cell (or nucleus),
extraction with 1.2 M NaCl leaves all domains with H3-H4
tetramer/DNA. Thus, after any given level of salt extraction,
subsequent analysis of radiation-induced damage will reflect an
averaged response of different chromatin domains; these do
mains may be more or less susceptible to damage, depending
upon the degree to which the salt extraction has perturbed their
original structure.
Due to the extraction strategy used, our SDS-PAGE analysis
reveals only those proteins remaining in the plug-purified chro
matin samples following the various NaCl extractions. Due to
the relatively large volumes of buffer used in the extractions,
we are not able to examine by SDS-PAGE the buffer washes
for the presence and identity of solubili/ed proteins. Thus,
although our SDS-PAGE analysis did not reveal any unex
pected or large losses of protein in the various chromatin
samples, we cannot exclude the possibility that small losses of
protein may have occurred at NaCl concentrations below those
typically expected to solubilize a particular class of chromo
somal protein (see below).
•¿y-Irradiation
of CHO Cells, Nuclei, and Protein-depleted
Chromatin Substrates. Treatment of CHO cells with 7-rays
leads to a dose-dependent increase in DNA double strand
breaks, which can be detected using the AFIGE/DSB assay (9).
The electrophoretic conditions of this assay are such that DNA
fragments below about 6 megabases in size enter the gel and
concentrate in a fairly narrow zone close to the origin. DNA
greater than 6 megabases cannot enter the gel and remains in
the well (i.e., at the origin). It is important to note that the
NaCl extraction scheme used to generate the various chromatin
substrates did not result in any detectable increase in DNA
DSB induction compared to intact CHO cells. When triplicate
unirradiated control samples from cells, nuclei, and the chro
matin substrates were analyzed using the AFIGE/DSB assay,
we found that only 1.95 ±0.26% (SD; n = 18) of the DNA
entered the gel. This background was subtracted from the values
obtained for irradiated samples.
CHO-WBL cells, nuclei, and the various chromatin sub
strates embedded in agarose plugs were treated with 0-40 Gy
of 7-rays, and the relative induction of DNA DSB was measured
by AFIGE. Figs. 4 and 5 show the results obtained from samples
prepared using either exponentially growing or plateau phase
cultures, respectively. The plot of the fraction of DNA entering
the gel, termed F.A.R., versus dose, yields a measure of radia
tion-induced DNA DSB. In Figs. 4 and 5, it is apparent that
increased stripping of chromosomal proteins leads to an in
creased sensitivity to DNA DSB induction. For both exponen
tial and plateau phase cultures, intact cells and nuclei are the
1582
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
CHROMATIN
STRUCTURE
AND DNA DOUBLE STRAND BREAKS
"C-THYMIDINE LABELED CHO
CELLS IN AGAROSE PLUGS
TRITON X-100
LYSIS
Fig. 2. Flow chart of procedure used to
generate protein-depleted chromatin samples.
CHO-WBL cells embedded in agarose plugs
were lysed in buffer containing Triton X-100,
and then washed to remove residual detergent.
Next, the plug-purified nuclei were treated
with buffers containing increasing concentra
tions of NaCI in order to strip off distinct
classes of chromosomal proteins.
0.14 M
NaCI
035 U
NaCI
06 M
NaCI
12 M
NaCI
20 M
NaCI
- TREAT SAMPLES WITH 0-40 GY OF GAMMA RADIATION
•¿
MEASURE DNA DSB BY PULSED FIELD GEL ELECTROPHORESIS
most resistant to radiation-induced DNA DSB; the radiation
sensitivity of salt-stripped samples increases as more and more
nuclear proteins are removed. Interestingly, while the absolute
F.A.R. is higher for irradiated samples prepared from plateau
phase cultures compared to the corresponding samples pre
pared from exponentially growing cultures, the pattern of rel
ative induction of DNA DSB appears similar for a given growth
state. Thus, the substrates can be ranked in order of increasing
radiosensitivity: cells, nuclei, and 0.35,0.6,1.2, and 2.0 M NaCI
stripped nuclei. This relative order is the same for both expo
nential and plateau phase samples, while the absolute F.A.R.
for the plateau phase samples is higher than that of the expo
nential growth phase samples. We note that the decreased doseresponse of irradiated exponential versus plateau phase cells
has been observed by others, and appears to be related to the
decreased mobility of replicating DNA in pulsed field gels (9,
11). A similar phenomenon appears in the analysis of DNA
DSB by neutral filter elution (see Refs. 10, 11, and 20).
Compared to intact CHO cells, plug-purified nuclei exhibit a
small increase in DNA DSB induction (about 1.5- to 2-fold,
when comparing at a dose of 20 Gy), in agreement with the
results of Radford (16), who used neutral filter elution to
measure DNA DSB. Extraction of various classes of chromo
somal proteins leads to significant differences in the frequency
of radiation-induced DNA double strand breaks. Removal of
nonhistone chromosomal proteins with 0.35 M NaCI results in
about a 3-fold increase in sensitivity to -y-rays, compared to
intact cells (compare curves at a dose of 20 Gy). This is notable
since nonhistone proteins account for about half of the total
mass of nuclear protein (1), and yet their loss only leads to a
relatively small increase in the induction of DNA DSB. Re
moval of both NHP and histone HI with 0.6 M NaCI results in
a slight increase in DNA DSB induction compared to NHPstripped chromatin. A greater than 10-fold increase in sensitiv
ity is observed in the 1.2 M NaCI stripped sample that is
depleted of NHP, HI, and histone H2A-H2B dimers. Some
what surprisingly, a nearly identical dose-response curve is
obtained after treatment with 2.0 M NaCI, which removes, in
addition to NHP, HI, and H2A-H2B dimer, histone H3-H4
tetramere from the core of the nucleosome (see Fig. 3). The
dose-response curve for this 2.0 M NaCI stripped sample is
virtually identical to that obtained using cellular DNA purified
using standard protease digestion procedures (data not shown).
For both the 1.2 and 2.0 M NaCI stripped samples, the doseresponse curves appear to plateau when about 90% of the DNA
has entered the gel. A similar plateau has been reported by
IHakis et al. (11) in their AFIGE analysis of cells treated with
high doses of radiation. This plateau could be due to some sort
of saturation phenomenon, or could result from migration
anomalies associated with DNA cross-linking (9).
Thus, it appears that the sensitivity of 1.2 and 2.0 M NaClextracted nuclei to 7-ray-induced DNA DSB is nearly identical,
even though the protein content of the 2 samples appears quite
different: the former sample still has histone H3-H4 tetramers
bound to DNA, whereas the latter sample is virtually proteinfree. As mentioned before, we cannot exclude the possibility
that small losses of protein may have occurred at NaCI concen
trations below those typically expected to solubilize a particular
class of chromosomal protein. For example, although 1.2 M
NaCI is known to extract NHP, HI, and H2A-H2B from nuclei
and leave H3-H4 tetramers bound to the DNA, a small portion
of weakly bound tetramers may be solubili/ed by 1.2 M NaCI,
perhaps as a result of posttranslational modifications that may
weaken the H3-H4 tetramer-DNA interaction compared to
unmodified tetramers (see "Discussion"). While such losses
may lead to an overestimation of the frequency of DNA DSB
induction in a particular sample, the data presented in Fig. 2
argue against this scenario as being a major contributor to the
observed DNA DSB dose-response curve.
Analysis of DNA Fragment Size in the Irradiated Samples. In
the AFIGE/DSB assay system, DNA fragments below about 6
megabases elute from the well and tend to accumulate in a
fairly narrow compression zone close to the well. Thus, in order
to obtain size information on the DNA fragments released from
the well in the irradiated samples, it is necessary to use an
alternate PFG technique. We used a recently developed tech-
1583
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
CHROMATIN
STRUCTURE
AND DNA DOUBLE STRAND BREAKS
below the well; this zone is essentially the exclusion limit of the
gel and represents a collection of DNA fragments greater than
6 megabases in size. Interestingly, in the 2.0 M NaCl stripped
samples, almost no DNA is found in this compression zone,
even at the lowest dose analyzed (10 Gy). In the 1.2 M NaCl
stripped sample, DNA is found in this compression zone only
with the lowest dose analyzed (10 Gy); at higher doses, the
DNA migrates as a broad distribution between 0.2 and 4.6
megabases. Thus, using these particular electrophoretic condi
tions, the most significant change in the size distribution of
200 kD -
97.4kD -
10
21.5kD
20
30
40
GAMMA RAY (Gy)
Fig. 4. Measurement of DNA DSB in i-irradiated CHO cells, nuclei, and
protein-depleted chromatin substrates prepared from exponentially growing cul
tures. Triplicate plugs of each substrate were irradiated with 0-40 Gy of -y-rays
in PBS on ice. Immediately after irradiation, DNA from the plugs was purified
and analyzed using AFIGE. The fraction of DNA entering the gel (i.e., F.A.R.)
provides an indirect measure of radiation-induced DNA DSB (see "Materials and
Methods").
14.3 kD
Fig. 3. Denaturing polyacrylamide gel electrophoresis of plug-purified nuclei
(lane 3), and nuclei treated with either 0.35 M NaCl (lane 4). 0.6 M NaCI (lane
5), 1.2 M NaCI (lane 6). or 2.0 M NaCl (lane 7). Lane 1 contains molecular
weight standards, whereas lane 2 contains total histones purified from calf thymus.
100-
90-
20 M NaCl
1.2 M NaCl
nique termed PACE to develop a protocol to resolve DNA
fragments between 0.1 and 6 Mb on a single run in about 2
days. In contrast, a similar run using a standard CHEF system
would require 7 to 14 days (21) to achieve a similar degree of
resolution.
Fig. 6 shows the results of the PACE/PFG analysis of puri
fied DNA from irradiated cells, nuclei, and the various chromatin substrates. Several points are worth noting, (a) The
ethidium bromide staining pattern clearly shows, as expected,
that for any particular sample, treatment with increasing doses
of 7-radiation leads to an increase in the amount of DNA
released from the well, (b) Across the sample groups, no signif
icant change in the DNA size distribution pattern is observed
until the integrity of the nucleosome core particle is perturbed.
Thus, with cells, nuclei, and 0.35 and 0.6 M NaCl stripped
samples, increasing doses of radiation generate a broad distri
bution of DNA fragments ranging in size from about 1.6
megabases to greater than 6 megabases. In contrast, when 1.2
and 2.0 M NaCl stripped chromatin samples are irradiated, the
size distribution of DNA fragments is shifted down, such that
most fragments are found between 0.2 and 4.6 megabases. For
all but the 2.0 M NaCl stripped samples, a significant portion
of the DNA accumulates in a narrow compression zone, directly
8070600.6 M NaCl
rr
<
50-
0.35 M NaCl
10
20
30
40
Gamma Ray (Gy)
Fig. 5. Measurement of DNA DSB in -, irradiated CHO cells, nuclei, and
protein-depleted chromatin substrates prepared from plateau phase cultures.
Triplicate plugs of each substrate were irradiated with 0-40 Gy of-y-rays in PBS
on ice. Immediately after irradiation, DNA from the plugs was purified and
analyzed using AFIGE. The fraction of DNA entering the gel (i.e., F.A.R.)
provides an indirect measure of radiation-induced DNA DSB (see "Materials and
Methods").
1584
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
CHROMATIN
STRUCTURE
AND DNA DOUBLE STRAND BREAKS
Cells
Nuclei
0.35 M
0.6 M
1.2
M
2.0 M
5.7 Mb
Fig. 6. PACE/PFG analysis of DNA frag
ments from 7-irradiated CHO cells, nuclei,
and protein-depleted chromatin substrates.
DNA from cells, nuclei, and nuclei washed
with either 0.35, 0.6, 1.2, or 2.0 M NaCl were
treated with 0, 10, 20, or 40 Gy of ^-radiation
and then analyzed by PACE pulsed field gel
electrophoresis. as described in "Materials and
Methods." A photograph of the ethidium bro
mide-stained gel is shown, with the arrange
ment of sample groups noted above the wells.
Within each group, samples are arranged in
order of increasing dose of y-radiation (0, 10,
20, or 40 Gy) from left to righi.
4.6
3.5
2.2
1.6
1.1
0.2
(22), a small portion of weakly bound tetramers also may be
solubilized by 1.2 M NaCl. This solubilization may be as a result
of posttranslational modifications that could weaken the H3H4 tetramer-DNA interaction compared to unmodified tetra
mers (23,24). If true, this could give rise to an apparent increase
in DSB in the 1.2 M NaCl-extracted chromatin sample. A
second, and perhaps more likely possibility is that the tetramers
that remained bound to the DNA may slide along the DNA
during the treatment with 1.2 M NaCl and aggregate at various
points within a given looped domain. Nucleosomal sliding is
DISCUSSION
known to occur at elevated NaCl concentrations (1). If this
In this report, we used a pulsed field gel electrophoresis assay aggregation occurs, large patches of protein-free, naked DNA
to study the role of chromatin structure in determining the could be exposed, such that the majority of the DNA would
sensitivity of genomic DNA to radiation-induced DNA double
have a similar sensitivity to that of naked DNA.
strand breaks. A series of protein-depleted chromatin substrates
In contrast, data in support of the possibility that the simi
was used in an attempt to dissect the relative contribution of larity in DNA DSB induction observed with the 1.2- and 2.0 M
various levels of DNA compaction in protecting the DNA from
NaCl extracts is real comes from studies of Barone et al. (25)
DSB damage. Removal of nonhistone chromosomal proteins
concerning the location of DNA DSB within nucleosomal
led to a 3-4-fold increase in DNA DSB induction (compared
DNA. Using isolated trinucleosomes irradiated with X-rays and
to intact cells); we noted a smaller increase in sensitivity follow
analyzing the resultant DNA DSB by neutral sucrose gradient
ing additional removal of histone HI. Samples treated with centrifugation, these authors found that DNA associated with
either 1.2 or 2.0 M NaCl yielded the greatest increase in DSB proteins in trinucleosomes was 3-4-fold more resistant than
induction (compared to intact cells). The dose-response of these
isolated DNA (25). These authors also found that the nucleo
latter samples was similar to that obtained with purified DNA.
somal structure hid only a few of the induced breaks, suggesting
The similarity of the dose-response curves of the 1.2- and 2.0 that DSB occurred preferentially in the linker DNA. Interest
M NaCl-extracted nuclei suggest that under the particular ex
ingly, in our study the 0.6 M NaCl-extracted chromatin sample
perimental conditions used in this study, removal of H2A-H2B
was about 4-fold more resistant to DSB induction compared to
dimers from CHO cell chromatin (in addition to removal of either the 1.2 or 2.0 M NaCl-extracted samples. Taken together,
NHP and histone HI) leaves the nuclear DNA exposed to y- these data suggest that the presence of chromosomal proteins
ray induced DSB to an extent that is similar to that of naked
may protect DNA from hydroxyl radicals generated by the
DNA (Fig. 4). Thus, these data suggest that H3-H4 tetramerincident radiation. Histones may protect the nucleosomal DNA
DNA is as susceptible to 7-ray-induced DNA DSB as is naked
by physically shielding it, scavenging free radicals, or perhaps,
via the exclusion of DNA-bound water. Disruption of histone
DNA. While this may in fact be true, several other points must
be considered that can temper this observation. First, as was H2A-H2B dimer-H3-H4 tetramer contacts is thought to lead
pointed out in "Results," the 1.2 M NaCl extraction procedure
to a partial dissociation of nucleosomal DNA from the remain
may remove more H3-H4 tetramers than is apparent from our ing protein core (26). Such an unraveling of the nucleosome
SDS-PAGE analysis of salt-extracted nuclei. For example, al
may result in an even greater percentage of DNA becoming
though 1.2 M NaCl is known to extract NHP, HI, and H2Asusceptible to radical attack. Thus, rather than invoking the
H2B from nuclei and leave H3-H4 tetramers bound to the DNA
preceding caveats, it is possible that the loss of H2A-H2B
1585
DNA fragments occurs only after treatments that disrupt the
integrity of the nucleosome core particle (i.e., stripping nuclei
with 1.2 or 2.0 M NaCl). These data coincide with the results
of our AFIGE/DSB assay (Fig. 4), which showed that the
greatest increase in DNA DSB induction occurred after strip
ping nuclei with 1.2 or 2.0 M NaCl, treatments that disrupt or
eliminate, respectively, the structure of the nucleosome core
particle.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
CHROMATIN
STRUCTURE
AND DNA DOUBLE STRAND BREAKS
dimers could account for the dramatic increase in radiation
sensitivity seen in the nuclei treated with 1.2 M NaCl.
In any case, the data presented here clearly indicate the
important role of the core histones and nucleosomal structure
in protecting the DNA from radiation-induced DSB. The great
est increase in DNA DSB induction occurred only after the
nucleosomal structure was disrupted by either 1.2 or 2.0 M
NaCl. In comparison, removal of nonhistone chromosomal
proteins and histone HI yielded only a relatively modest in
crease in sensitivity. This is noteworthy, since NHP comprise
about half of the total mass of protein found in the nucleus (1).
Thus, our results suggest that different regions of the genome
may be more or less susceptible to radiation-induced DNA
DSB, depending upon the extent to which they are protected
by their association with nuclear proteins.
In addition to studying DNA DSB within the entire genome,
pulsed field gel electrophoresis offers the potential to analyze
the influence of chromatin structure on the induction of DNA
DSB within specific chromosomal loci. Previous work by Oleinick et al. (27) suggests that ionizing radiation can induce single
strand breaks and cross-links preferentially within the more
solvent accessible, transcriptionally active chromatin (27, 28).
Since it is estimated that about two-thirds of the damage
produced by low linear energy transfer radiation within cells
occurs indirectly as a result of the ionization of nuclear water
(29, 30), it is possible that even a physical clastogen such as yradiation can induce DNA DSB preferentially within discrete
chromatin domains. Our PACE/PFG results showing a shift in
the size distribution of DNA fragments from salt stripped
chromatin (Fig. 6) support the suggestion that more accessible
regions are more susceptible to clastogenic damage. Efforts to
test the possibility that radiation-induced DNA DSB are dis
tributed nonrandomly throughout the genome are currently
under way in our laboratory.
7.
8.
9.
10.
11.
' 2.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
ACKNOWLEDGMENTS
23.
We thank Philip Hertzog for preparing the 2-jim-thick sections,
Gerald Peklak for assistance in constructing the timing/switching con
trol units used in the AFIGE assays, and Rosina Hill and Carole Bradt
for help in the use of their photomicroscopes. We are grateful to Dr.
James Young and Dr. Brian Ledwith for critically reviewing this
manuscript.
24.
25.
26.
REFERENCES
1. van Holde, K. E. Chromatin. New York: Springer Verlag, 1989.
2. Yunis, J. J., Soreng, A. L., and Bowe, A. E. Fragile sites are targets of diverse
mutagens and carcinogens. Oncogene, 1: 59-69, 1987.
3. Frankenberg-Schwager, M., Frankenberg, D., Blocher, D., and Harbich, R.
Evidence for DNA double-strand breaks as the critical lesions in yeast cells
irradiated with sparsely or densely ionizing radiation under oxic or anoxic
conditions. Radiât.Res., 88: 524-532, 1981.
4. Waldren, C. A., Carrell, L., Sognier, M., and Puck, T. T. The measurement
of X-ray mutagenesis in relation to human disease. Proc. Nati. Acad. Sci.
USA, «J:4839-4843, 1986.
5. Graf, L. H., and Chasin, L. A. Direct demonstration of genetic alterations at
the dihydrofolate reducÃ-aselocus after gamma irradiation. Mol. Cell Biol.,
2:93-96, 1982.
6. Borek, C.. and Hall, E. S. Induction and modulation of radiogenic transfor
27.
28.
29.
30.
mation in mammalian cells. In: J. D. Boice and J. F. Fraumeni, Jr. (eds.),
Radiation Carcinogenesis, Epidemiology and Biological Significance, pp.
292-302. New York: Raven Press, 1984.
Bryant, P. E., and Riches, A. C. Transformation of murine C3H10T1/2 cells
resulting from DNA double-strand breaks induced by a restriction endonuclease. Br. J. Cancer, 60: 852-854, 1989.
Bradley, M. (.).. and Taylor, V. DNA double-strand breaks induced in normal
human cells during the repair of ultraviolet light damage. Proc. Nati. Acad.
Sci. USA, 78: 3619-3623, 1981.
Starnato, T. D., and Denko, N. Asymmetric field inversion gel electropho
resis: a new method for detecting DNA double-strand breaks in mammalian
cells. Radiât.Res., 121: 196-205, 1990.
Elia, M. C., DeLuca, J. G., and Bradley, M. O. Significance and measurement
of DNA double strand breaks in mammalian cells. Pharmacol. Ther., 57:
291-327,1991.
Iliakis, G. E., Cicilioni, O., and Metzger, L. Measurement of DNA double
strand breaks in CHO cells at various stages of the cell cycle using pulsed
field gel electrophoresis: calibration by means of '"I decay. Int. J. Radiât.
Biol., 59:343-357, 1991.
Denko, N., Giaccia, A., Peters, B., and Starnato, T. D. An asymmetric field
inversion gel electrophoresis method for the separation of large DNA mole
cules. Anal. Biochem., 178: 172-176, 1989.
Clark, S. M., Lai, E., Birren, B. W., and Hood, L. A novel instrument for
separating large DNA molecules with pulsed homogeneous electric fields.
Science (Washington DC), 241: 1203-1205, 1988.
Bell, D. DNA synthesis in nuclear monolayers from BSC-1 cells infected
with herpes virus. Nature (Lond.), 248: 505-508, 1974.
Osborn, M., and Weber, K. The detergent-resistant cytoskeleton of tissue
culture cells includes the nucleus and the microfilament bundles. Exp. Cell
Res., 106: 339-349, 1977.
Radford, I. R. Use of nuclear monolayers to identify factors influencing DNA
double-strand breakage by X-rays. Int. J. Radiât.Biol., 52: 853-858, 1987.
Berezney, R., and Coffey, D. S. Nuclear matrix: isolation and characterization
of a framework structure from rat liver nuclei. J. Cell Biol., 73: 616-637,
1974.
Cook, P. R. A general method for preparing intact nuclear DNA. EMBO J.,
3: 1837-1842, 1984.
Jackson, D. A., Yuan, J., and Cook, P. R. A gentle method for preparing
cyto- and nucleo-skeletons and associated chromatin. J. Cell. Sci., 90: 365378, 1988.
Okayasu, R., Bloecher, D., and Iliakis, G. Variation through the cell cycle in
the dose-response of DNA neutral filter elution in X irradiated synchronous
CHO cells. Int. J. Radial. Biol., 53: 729-747, 1989.
Ahn, S. Y., Nevaldine, B., and Hahn, P. J. Direct measurement by pulsed
field gel electrophoresis of induction and rejoining of X-ray induced double
strand breaks in cultured mouse cells. Int. J. Radiât.Biol., 59: 661-675,
1991.
Burton, D. E., Butler, M. J., Hyde, J. E., Phillips, D., Skidmore, C. J., and
Walker, I. O. The interaction of core histones with DNA: equilibrium binding
studies. Nucleic Acids Res., 5: 3643-3663, 1978.
Norton, V. G., Imai, B. S., Yau, P., and Bradbury, E. M. Histone acetylation
reduces nucleosome core particle linking number change. Cell, 57: 449-457,
1989.
Norton, V. G., Marvin, K. W., Yau, P., and Bradbury, E. M. Nucleosome
linking number change controlled by acetylation of histones H3 and H4. J.
Biol. Chem., 265: 19848-19852, 1990.
Barone, F., Belli, M., Rongoni, E., Sapora, O., and Tabocchini, M. A. X-ray
induced DNA double-strand breaks in polynucleosomes. ///: F. J. Bums, A.
C. Upton, and G. Silini (eds.). Radiation Carcinogenesis and DNA Altera
tions, pp. 293-296. New York, Plenum Publishing Corp., 1986.
Burlingame, R. W., Love, W. E., Wang, B. C., Hamlin, R., Xuong, N-H.,
and Moudrianakis, E. N. Crystallographic structure of the octameric histone
core of the nucleosome at a resolution of 3.3 A. Science (Washington DC),
22«:546-553, 1985.
Oleinick, N. L., Chiù,S. M., Friedman, L. R., Xue, L. Y., and Ramakrishnan,
N. DNA-protein cross-links: new insights into their formation and repair in
irradiated mammalian cells. In: R. Simic, L. Grossman, and A. Upton (eds.),
DNA Damage and Repair: Implications for Carcinogenesis and Risk Assess
ment, pp. 181-192. New York, Plenum Publishing Corp., 1986.
Chiù, S. M., Oleinick, N. L., Friedman, L. R., and Stambrook, P. J.
Hypersensitivity of DNA in transcriptionally active chromatin to ionizing
radiation. Biodi ini. Biophys. Acta, 699: 15-21, 1982.
Mee, L. K., and Adelstein, S. J. Formation of strand breaks in the DNA of
7 irradiated chromatin. Radiât.Environ. Biophys., 26: 13-22, 1987.
Roots, R., Chatterjee, A., Blakely, E., Chang, P., Smith, K., and Tobias, C.
Radiation responses in air-, nitrous oxide-, and nitrogen-saturated mamma
lian cells. Radiât.Res., 92: 245-254, 1982.
1586
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
Influence of Chromatin Structure on the Induction of DNA
Double Strand Breaks by Ionizing Radiation
Michael C. Elia and Matthews O. Bradley
Cancer Res 1992;52:1580-1586.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/52/6/1580
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.