Mutagenesis vol.12 no.6 pp.437^W2, 1997
Chromosomes with high gene density are preferentially repaired
in human cells
J.SurraUfe 1 ' 3 , S.Sebastian 1 ' 4 and A.T.Natarajan 1 - 2
'Department of Radiation Genetics and Chemical Mutagenesis, Leiden
University, Leiden and 2J.A.Cohen Institute, Interuniversity Research
Institute for Radiopathology and Radiation Protection, Leiden, The
Netherlands
It is known that DNA repair is heterogeneous in human cells
since open chromatin, active genes and their transcribed
strands are preferentially repaired. It is thus expected that
DNA repair is clustered in chromosomes with high gene
density. We have employed a DNA repair inhibitor, cytosine
arabinoside (Ara-C), to convert ethyl methane sulfonate
(EMS)-induced excision repairable lesions to chromosomal
breaks, to check for the existence of heterogeneity of
repair at the chromosome level. Chromosome staining by
fluorescence in situ hybridization (FISH) was used to
analyze breakage in chromosomes with diverse gene densities. These chromosomes were identified by means of the
CpG island distribution after FISH with a CpG islandrich probe isolated from total human genomic DNA. Thus,
three chromosomes with very high gene density (numbers
1, 19 and 20) were compared with two chromosomes with
very low gene density (numbers 4 and 18) for clastogenicity
and sensitivity to co-treatment with Ara-C and EMS. Our
data indicate that those chromosome with higher gene
density are more sensitive to a combination treatment with
Ara-C and EMS, indicating that the level of excision repair
synthesis is higher in those chromosome. It is therefore
concluded that DNA excision repair is preferentially directed to chromosomes with high gene density. The implications of this finding in human biomonitoring using FISH
techniques are discussed.
Introduction
It is generally accepted that chromosome mutations are causal
events in the development of many human neoplasias. It is
known that increased levels of chromosome breakage are
predictive of future cancer risk in humans (Hagmar et al,
1994). Chromosome staining by fluorescence in situ hybridization (FISH) with whole chromosome-specific libraries has
become a powerful methodology for detecting chromosome
abnormalities (Pinkel et al, 1988; Natarajan et al, 1991,
1992a). FISH has important applications in clinical and tumour
cytogenetics as well as in biological dosimetry of populations
exposed to mutagenic agents and in in vitro studies on
chromosomal aberration induction (Tucker et al, 1995a;
Natarajan et al., 1996a,b).
A general assumption of the use of painting to quantify
chromosome damage in the whole genome by extrapolation
from a few stained chromosomes is that the occurrence of
break points is randomly distributed within and between
chromosomes. However, many studies using banding or FISH
methodologies have suggested that a random distribution is
not actually the case (see Tucker and Senft, 1994; Natarajan
et al., 1996b). It has previously been reported that chromosome
1 breaks are found more often than expected based on a
random distribution of breaks points when lymphocytes are
evaluated in individuals with past radiation exposure (Lucas
et al, 1992; Natarajan et al, 1992a; Sachs et al, 1993).
However, these observations do not reflect actual interchromosomal differences in radiosensitivity, since it is known that the
frequency of unstable chromosome aberrations declines with
age and that selection can act differently depending on the
chromosome involved in the exchange aberration. Recent
studies using several chromosome-specific probes have demonstrated that there is heterogeneity in response to ionizing
radiation between chromosomes in human (Knehr et al., 1996;
Boei et al, 1996) and in Chinese hamster (Domfnguez et al,
1996; Grigorova et al, 1997). These findings could have
important implications in biomonitoring and biological dosimetry, since the assumption of random breakage in quantifying
chromosome damage would have to be re-examined.
Slijepcevic and Natarajan (1994a,b) suggested that the initial
lesions induced by X-rays in hamster cells are not evenly
distributed in the chromosomes. According to Tucker and
Senft (1994), the observed non-random distribution of break
points is most likely explained by non-random breakage or
repair. Indeed, DNA repair is influenced by many factors, such
as chromatin structure and transcriptional activity. It is widely
accepted that open chromatin, active genes and their transcribed
strands are preferentially repaired in human cells (Mullenders
et al, 1991). These observations are explained not only by an
enhanced accessibility of chromatin to repair factors at active
loci but also by transcription-coupled repair mechanisms
(reviewed by Drapkin et al, 1994; Friedberg, 1996).
Human chromosomes are also highly heterogeneous. Genes
are unevenly distributed both within a single chromosome and
between chromosomes. Transcriptional activity is clustered in
some chromosomal sites and other regions, such as heterochromatic bands, are genetically silent. Chromatin structure in
active regions is usually open and accessible to transcription
factors. These active regions are early replicating and rich in
hypomethylated DNA. They can be easily visualized with
diverse chromosome staining methodologies, such as G or
R banding (Verma and Babu, 1995). On the other hand,
transcriptionally inactive chromatin is highly methylated, late
replicating and is distinguished by its lack of acetylated
histones (Jeppesen and Turner, 1993; Surralles et al, 1996;
Surrall^s and Natarajan, 1997).
Interestingly, most genes are clustered in early replicating
^ o whom correspondence should be addressed at present address: Grup de Mutagenesi, Departament de Genetica i de Microbiologia, Edicifi Cn, Universitat
Autdnoma de Barcelona, Campus de Bellaterra s/n, 08193 Bellaterra (Cerdanyola del Valles), Spain. Tel: +34 3 581 2597; Fax: +34 3 581 2387;
Email: [email protected]
4
Present address: Biochemistry Department, Michigan State University, East Lansing, MI, USA
© UK Environmental Mutagen Society/Oxford University Press 1997
437
J-Surralles, S-Sebastian and A.TJSatarajan
R bands of human chromosomes, as seen by FISH with CpG
island-enriched DNA segments (Craig and Bickmore, 1994).
CpG islands are CG-rich regions found at the 5'-end of most
mammalians genes, including all housekeeping genes. They
have an important role in gene regulation through methylationbased mechanisms (Bird, 1986).
As gene density and, therefore, chromatin structure and
transcriptional activity is highly variable between human
chromosomes, we hypothesized that DNA repair could be
preferentially directed to those chromosomes with high gene
density. To test this hypothesis, chromosome breakage was
induced by inhibiting repair synthesis of ethyl methane sulfonate (EMS)-induced DNA lesions with cytosine arabinoside
(Ara-C). Heterogenous breakage after inhibition of repair
synthesis was then examined in chromosomes with high and
low gene density which were identified by FISH with CpG
islands isolated from human genomic DNA. Chromosome
breakage was analysed by double and triple colour FISH with
whole chromosome-specific probes.
Materials and methods
Isolation of CpG islands
CpG island isolation was performed essentially as previously described (Craig
and Bickmore, 1994) with several modifications. Briefly, total human genomic
DNA was isolated from 10 ml heparinized whole blood using a Wizard™
genomic purification kit (Promega). The extracted DNA was quantified and
fully digested with the restriction enzyme HpaM (restriction site CCGG; 5 U/
|ig DNA). Aliquots of the digested DNA were end-labelled with Klenow and
[ P]dCTP and run in a 1% low melting point agarose gel. Unlabelled aliquots
of digested DNA were run in parallel tracks. The part of the gel with
radioactive DNA was authoradiographed to visualize the band with the small
CpG islands. An equidistant band with unlabelled CpG islands was excised
from the gel and the DNA recovered with AgarACE™, an agarose digesting
enzyme (Promega). The recovered DNA containing the CpG islands was
concentrated on a Centricon 10 column. The concentrated DNA was finally
random prime-labelled with biotin and alcohol precipitated along with salmon
sperm DNA and tRNA as a carrier. The biotinilated CpG island-enriched
DNA was then used as a probe for FISH with normal chromosomes.
Cell culture and chemical treatments
Cell culturing was essentially performed as described elsewhere (Natarajan
et al., 1992b). Peripheral blood was drawn into heparinized vacutainers and
whole blood cultures were set up in Ham's F10 medium supplemented with
20% fetal calf serum (Gibco) (heat inactivated for 30 min at 56°C), 300 \lgl
ml phytohaemagglutinin (PHA; Wellcome), 200 mM L-glutamine, heparin
and antibiotics.
Immediately after set-up, cells were left untreated or were treated with 0.5
mM EMS and the DNA synthesis inhibitor Ara-C (1 Hg/ml). Parallel cultures
received either EMS alone or Ara-C alone, but they were excluded from the
FISH analysis due to their low clastogenicity when given separately. Both
treatments were washed out after 16 h incubation in order to avoid any
interference with replicative synthesis. Thus Ara-C and EMS were removed
just before S phase by washing the cells twice in 5 ml prewarmed phosphatebuffered saline (PBS) at 37°C, followed by addition of fresh medium
without PHA but supplemented with 10 Hg/ml 2'-deoxycytidine (Sigma).
Deoxycytidine was added to block residual effects of the remaining Ara-C
bound to the polymerases and allow DNA replication to proceed normally.
Metaphases were arrested with colchicine for 2 h and harvested 48 h after
culture initiation, when cells were treated with a prewarmed hypotonic solution
(0.075 M KG, 37°C) and fixed in acetic acid:methanol. Some untreated
cultures were incubated for 72 h and treated with colchicine for 30 min before
harvesting. Air dried preparations were made and stored at -20°C until FISH
with either biotinilated CpG islands (72 h cultures) or two combinations of
chromosome-specific DNA libraries (chromosomes 1 and 4 and 18, 19 and
20) in a multicolour fashion (48 h cultures).
Labelling of chromosome-specific DNA libraries
DNA libraries specific for chromosomes 1, 4, 18, 19 and 20 were labelled
with either biotin-16-dUTP (Boehringer, Mannheim), digoxigenin-11-dUTP
(Boehringer, Mannheim) or both dUTP conjugates by the polymerase chain
reaction (PCR). For PCR the following mixture was used: 4-10 ng template
(whole chromosome DNA library); buffer consisting of 10 mM Tris-HCl, pH
438
8.3, 1.5 mM MgCl2, 50 mM KC1, 0.1% gelatin; 10 nmol dATP; 10 nmol
dCTP; 10 nmol dGTP; 5-6.5 nmol dTTP, either 5 nmol biotin-16-dUTP or
3.5 nmol digoxigenin-11-d(J IV or a combination of 2 nmol biotin-16-dUTP
and 2 nmol digoxigenin-11-dUTP, 25 pmol each M13 forward and reverse
sequencing primers; 2.5 U AmpliTaq DNA polymerase (Perkin Elmer). The
amplification and labelling were carried out by running 30 cycles of 1 min at
93°C (denaturation), 1 min at 55°C (annealing) and 3 min at 72°C (extension)
for each cycle. Following this procedure, chromosome 1 was labelled
with biotin, chromosome 4 with digoxigenin, chromosome 18 with biotin or
digoxigenin or both (50% each), chromosome 19 with biotin and chromosome
20 with biotin or digoxigenin or both (50% each).
In situ hybridization procedure
The standard protocol as described by Pinkel et al. (1988) with some
modifications was used for in situ hybridization. For each slide, 50-100 ng
biotinilated CpG islands or 3 |ll chromosome-specific PCR product were
mixed with 3 ng human Cot-1 competitor DNA and dried using a speed-vac
machine. The dried DNA was dissolved in hybridization buffer (50% formamide in 2X SSC, 10% dextran sulfate and 40 mM phosphate buffer, pH 7.0),
denatured for 5 min at 70°C and chilled on ice. The repetitive sequences were
competed out for 1-1.5 h at 37°C.
Before in situ hybridization, thawed preparations were pretreated with
RNase A (Boehringer) (100 ng/ml in 2X SSC) for 1 h at 37°C and with
pepsin (50 |ig/ml in 10 mM Hcl; Serva) for 15 min at 37°C. Denaturation of
in situ DNA was achieved by adding a mixture consisting of 70% deionized
formamide, 2X SSC and 10 mM phosphate buffer, pH 7.0, to the preparations
under coverslips. The preparations were placed on an 80°C hotplate for 4 min
and chilled with ice-cold 70% ethanol following further dehydration with
sequential incubations in 90% and absolute ethanol at room temperature. The
denatured probes were mixed in appropriate combinations and were dropped
onto the slides, which were then covered with a coverslip and sealed with
glue. In situ hybridization with the target DNA was allowed to occur overnight
at 37°C in a moist chamber.
Immunological procedures for single and multicolour FISH
After hybridization the preparations were washed in 50% formamide, 2X
SSC, pH 7.0 (4X5 minat42°C), followed by washes in 0.1X SSC (3X5 min
at 60°C) and 4X SSC, 0.05% Tween 20 for 5 min at room temperature. The
preparations were subsequently incubated with 5% natural non-fat dry milk
(Lucerne Ltd) for 15 min at room temperature.
For single colour FISH (CpG island chromosomal distribution) the preparations were alternately incubated with TRTTC-avidin D (5 u.g/ml, 3X30 min
at room temperature; Vector Laboratories), then biotin-conjugated goat antiavidin D antibodies (5 |ig/ml, 2X30 min at room temperature; Vector
Laboratories) and a final incubation with TRTTC-avidin D. Slides were washed
with 4X SSC, 0.05% Tween 20 (3X5 min) after each antibody incubation.
For multicolour FISH antibody incubations were for 30 min at 37°C if not
stated otherwise. Briefly, preparations were incubated with FTTC-avidin D (5
Hg/ml, 30 min at room temperature; Vector Laboratories), biotin-conjugated
goat anti-avidin D antibodies (5 ng/ml, 2X30 min at room temperature) and
FTTC-avidin D and then washed first with 4X SSC, 0.05% Tween 20 (2X5
min) and second with TNT (0.1 M Tris-HCl, pH 7.5, 0.15 M NaCl, 0.05%
Tween 20) (5 min). Next, preparations were alternately incubated twice with
mixtures of either mouse anti-digoxigenin (0.5 and 0.25 |ig/ml respectively;
Boehringer) and biotin-conjugated goat anti-avidin D antibodies (5 (ig/ml)
diluted in TNB (0.1 M Tris-HCl, 0.15 M NaCl, 0.5% Boehringer blocking
reagent) or FTTC-avidin D (5 ng/ml) and digoxigenin-conjugated sheep antimouse antibodies (2 and 1 ng/ml respectively; Boehringer) diluted in TNB.
For the last step, preparations were incubated with TRTTC-conjugated sheep
anti-digoxigenin antibodies (2 Hg/ml; Boehringer) diluted in TNB. After each
antibody incubation preparations were washed with TNT (3X5 min).
Finally, preparations were dehydrated, dried and mounted in Vectashield
(Vector Laboratories) containing 0.15 ng/ml 4',6-diamidino-2-phenylindole
(DAPI; Serva).
Microscopy and scoring
The distribution of CpG islands was generated by computer enhancement of
the FISH signal and the chromosomes were identified by computer-assisted
generation of the DAPI banding pattern. Slides were analysed for chromosome
aberration using a Zeiss microscope equipped with a triple filter to simultaneously visualize chromosomes stained red (TRTTC-detected, digoxigeninlabelled probes), green (biotinilated probes detected with FTTC), yellow
[biotimdigoxigenin (l:l)-labelled probes simultaneously detected with FITC
and TRITC] and blue (DAPI counterstaining). All aberrations were recorded,
including reciprocal and non-reciprocal translocations (terminal and interstitial), dicentrics, fragments and rings. Genomic frequencies of different
chromosomal aberrations were estimated per chromosome following the
formula:
Repair and chromosome gene density
Fc = Fp/2/pd - / p ) ,
where F o is the genomic frequency, Fp is the frequency of aberrations
observed with the stained chromosomes and fp is the fraction of the genome
stained by FISH. Interactions between stained chromosomes were considered
separately for each chromosome. The frequency of break points per chromosome was also calculated. Insertions (i.e. interstitial translocations) of fragments
derived from stained chromosomes were considered as two breaks. This
estimation allowed us to correct the number of break points in each chromosome
for its DNA content so that direct comparisons between chromosomes could
be done.
Results and discussion
Analysis of the hybridization patterns observed after FISH
with biotinilated CpG islands generated with the restriction
enzyme HpaH confirmed the observations reported by Craig
and Bickmore (1994), who showed that genes are clustered in
some, but not all, R bands. Based on computer-assisted
generation of the DAPI banding pattern, most of the FISH
signals were located in the light DAPI chromosomal bands
Qight G bands). CCD images of the metaphase spreads
counterstained with DAPI and hybridized with the CpG islands
were obtained. The DAPI banding pattern was digitally converted to black and white in order to obtain a pattern resembling
classical G banding, which facilited identification of the
chromosomes. Among the long chromosomes (groups A and
B), chromosome 4 showed hardly any hybridization signal,
whereas chromosome 1 was rich in CpG islands, above all in
the terminal part of the short arm and light G bands of the long
arm, surrounding the pericentromeric region. Chromosomes of
group C presented an intermediate level of CpG island density,
with some clusters in the telomeric regions. As expected, CpG
islands in the big acrocentric chromosomes (group D) were
clustered in the short arms (nucleolar organizer regions, NOR),
where the density of ribosomal genes is known to be very
high. Among the smaller chromosomes (16-22 and Y), chromosomes 18 and 19 showed lower and higher concentrations of
CpG islands respectively, hence confirming the findings of
Craig and Bickmore (1994). As CpG island density is a good
indicator of gene density, we can conclude that: (i) among the
long chromosomes, chromosomes 1 and 4 have higher and
lower gene densities respectively; (ii) among the smaller
chromosomes, chromosome 18 has little protein coding
information; (iii) chromosome 19 is that with relatively the
highest density of genes in the human genome.
Based on these observations, five chromosomes were
selected in order to check their differential response to the
DNA repair inhibitor Ara-C: chromosomes 1 and 4 as examples
of long chromosomes with high and low gene density respectively; chromosomes 18 and 19 as examples of small chromosomes with low and high gene density respectively;
chromosome 20 as an example of a small chromosome with
an intermediate level of gene density. Blood cultures were
either left untreated or treated with a combination of Ara-C
and EMS during most of the Gj phase. At the concentrations
and timings used, Ara-C and EMS given separately induced
hardly any clastogenic effect in human lymphocytes (Surralles
et al, 1997). Indeed, previous studies conducted in our
laboratory showed that >83% of the clastogenic effect induced
by co-treatments with Ara-C and EMS was due to a synergistic
interaction between the two chemicals (Surralles et al., 1997),
so the cultures that received Ara-C and EMS alone were
excluded from the FISH analysis of clastogenicity.
Clastogenicity induced by co-treatments with Ara-C and
EMS were mediated by a specific inhibition of the gap filling
step of excision repair of EMS-induced lesions by Ara-C
(Surralles et al, 1995, 1997). Ara-C inhibits DNA synthesis
and hence blocks the gap filling step of excision repair of
EMS-induced lesions (Mirzayans et al, 1994). To achieve
specific inhibition of repair synthesis, Ara-C action was
restricted in large part to the G! phase. Unfilled repair gaps
then reached S phase and were converted to chromosome
breaks (Fenech and Neville, 1992; Fenech et al, 1994). It is
important to note that with this experimental design the more
efficient DNA repair, the higher the frequency of breaks.
Chromosome breakage was measured in metaphase cells using
multicolour FISH with specific staining probes targeting chromosomes with diverse gene densities.
Data on spontaneous and Ara-C + EMS-induced chromosome breaks are presented in Table I. Spontaneous breakage
is based on 3464 metaphases analysed for chromosomes 1 and
4 and 3875-7048 metaphases for the smaller chromosomes
(18, 19 and 20). A higher number of cells was analysed for
the small chromosomes as they represent a smaller fraction of
the total genome in terms of their DNA content. The frequency
of spontaneous translocations was higher than dicentrics in all
chromosomes analysed, for both untreated and Ara-C +
EMS-treated cultures. This phenomenon has been extensively
discussed elsewhere (Natarajan et al, 1996b). The elevated
frequency of translocations when compared with dicentrics
could be related to the fact that each translocated chromosome
was scored as a separate event, leading to identification of two
types of translocated chromosomes that could actually be
related in origin (Tucker et al, 1995a), e.g. those with a
stained centromere and an unstained portion [t(Ba), using the
PAINT nomenclature system suggested by Tucker et al,
1995b] and those with an unstained centromere and a stained
portion [t(Ab)]. As this paper focuses on interchromosomal
differences in breakage and repair, the difference between the
ratio of dicentrics and translocations will not be further
discussed.
Interestingly, the overall frequency of spontaneous breaks
is much higher in the group of small chromosomes when
compared with big chromosomes (Table I). The biological
significance of this observation, if any, is not known. It could
very well be related to random differences due to the low
number of spontaneous aberrations found. When comparisons
were made within size groups, the chromosomes with very
low gene density (4 and 18) presented an increased spontaneous
fragility. However, this observation must be interpreted with
care, since the total number of breaks in untreated cultures
was rather low and false conclusions might be deduced.
Most convincingly, chromosomes with high gene density
are more sensitive to co-treatments with Ara-C and EMS than
chromosomes with low gene density (Table I), indicating than
the level of repair synthesis is positively correlated with gene
density. Thus high gene density chromosome 19 is the most
sensitive of all chromosomes analysed. This observation is
statistically significant and consistent in the two size groups.
When the two long chromosomes are compared, the number
of Ara-C + EMS-induced break points encountered in chromosome 1 (64.8%) is 2.1-fold higher than in chromosome 4
(30.3%). This data cannot be explained by the presence of the
breakage prone chromosome 1 heterochromatin region (band
Iql2), as previous studies have demonstrated that this band
has reduced DNA repair capacity and, therefore, is insensitive
to co-treatments with Ara-C and EMS (Surralles et al, 1997).
439
J.Surralles, S.Scbastlan and A.T.Natarqjan
Table I. Detection of chromosomal aberration in untreated or EMS (0.5 mM) and Ara-C (1 Hg/ml) co-treated 48 h whole blood cultures by multicolour FISH
Treatment
Chromosome no.
Cells scored
Type of chromosomal aberration detected by FISH
Translocations
Control
AraC + EMS
1
4
18
19
20
1
4
18
19
20
3464
3464
6639
7048
3875
1109
1109
1946
2248
725
Dicentrics
Fragments
Break points0
n
%b
n
%
n
%
n
%
1
1
8
5
4
54
24
18
26
10
0.20
0.26
2.37
1.79
2.32
32.71
18.60
18.16
29.54
32.07
1
0
2
1
1
29
10
10
21
2
0.02
0.00
0.59
0.36
0.60
17.53
7.74
10.07
23.72
6.27
4
6
4
1
0
20
5
3
12
0
0.81
1.46
1.18
0.36
0.00
12.09
3.87
2.96
13.52
0
6
7
15
8
6
107
39
32
65
13
1.17
1.72
4.54
2.81
3.48
64.82
30.26
32.38
73.72
41.60
•Bicolour junctions and fragments (interstitial translocations of stained chromosomes were scored as two breaks).
''Corrected for the whole genome.
#18
#04
#20
#01
#19
Qvomoaonwa wtth Ini !•—Inp o«n« dansfty
Fig. 1. Relationship between chromosomal gene density and
interchromosomal differences in repair synthesis inhibition of EMS-induced
DNA lesions in 48 h cultured human lymphocytes. Chromosomes are
ordered from low to high gene density.
Similarly, low gene density chromosome 18, with an induced
frequency of breakage of 32.4%, is 2.3-fold less sensitive than
high gene density chromosome 19, with as much as 73.7%
Ara-C + EMS-induced breakage. Chromosome 20, with an
intermediate level of gene density, also presents an intermediate
level of induced clastogenicity (41.60%).
The observed positive correlation between gene density and
sensitivity to inhibition of repair synthesis is illustrated in
Figure 1. Chromosomes with increasing gene density are
plotted against Ara-C + EMS-induced break points after
subtracting the spontaneous frequency of break points
observed. Considering that unfilled EMS-induced repair sites
are converted chromosome breaks due to Ara-C, our results
suggest that excision repair is preferentially directed to chromosomes with high gene density.
Our findings of a relatively low level of excision repair
synthesis in chromosomes with low gene density cannot be
explained by a lower alkylation level of those chromosomes,
since previous studies demonstrated that ethylation is found
even more frequently in satellite DNA and highly repetitive
sequences than in gene-rich euchromatin (Durante et al, 1989).
440
On the other hand, and consistent with the notion that base
excision repair of DNA alkylation is not coupled to transcription, the repair of alkylated bases is probably not strand biased
and not strongly affected by the transcription activity of the
region (Friedberg et al, 1995). Some studies have actually
demonstrated that there is no difference in repair of damage
produced by alkylating agents between transcriptionally active
and silent genes or between transcribed and non-transcribed
strands in Chinese hamster cells (Scicchitano and Hanawalt,
1995; Wang et al, 1995).
The elevated frequency of Ara-C + EMS-induced breaks
in high gene density chromosomes is probably better related
to an enhanced accessibility of EMS-induced DNA damage to
repair enzymes in open chromatin (Smerdon, 1991). This
enhanced accessibility would facilitate the excision of alkylated
bases from DNA and Ara-C-mediated formation of unfilled
gaps and chromosome breaks. It is known that artificial changes
in chromatin conformation have led to different rates of removal
of DNA lesions. For instance, repair synthesis occurring after
UV irradiation in human cells treated with sodium butyrate (an
inhibitor of histone deacetylases) was enhanced in nucleosome
cores containing hyperacetylated histones (Ramanathan and
Smerdon, 1989). This finding is consistent with the role of
hyperacetylation of histones in chromatin unfolding and in
determining the genetic activity of diverse chromosome
regions, as seen with antibodies against acetylated isoforms of
histone H4 (Jeppesen and Turner, 1993; Turner, 1993; Surralles
et al, 1996; Surralles and Natarajan, 1997). However, the
possibility of decreased accessibility to repair enzymes in
chromosomes with low gene density must be further studied
in specially designed experiments.
Following X irradiation of human lymphocytes and assessment at first mitosis, chromosome 4 is more often involved in
chromosome exchange (translocations) than expected, whereas
chromosome 1 is less often involved (Knehr et al, 1996, Boei
et al, 1997), a situation opposite to the findings reported here
for combined treatment with EMS and Ara-C. This apparent
contradiction can be easily explained in terms of differential
repair rates between these two chromosomes. Following X
irradiation, the chromosomes with low gene density would
repair the lesions slowly, thus allowing interaction with other
lesions to form exchanges. The opposite would be true for
chromosomes, such as 1 or 19, with high gene density, where
Repair and chromosome gene density
repair of lesions would be very quick and, therefore, the time
for lesions to interact and produce exchange aberrations would
be short. Another explanation is that in chromosomes with
high gene density there may be a higher probability of excision
occurring simultaneously on opposite strands of the DNA in
sufficiently close proximity to cause double-strand breaks.
This model has recently been proposed as a mechanism for
chromosome breakage due to uracil misincorporation after
folate deficiency (Blount et al, 1997).
Our finding of enhanced repair synthesis in chromosomes
with high gene density casts doubt on the assumption of a
random distribution of chromosome damage when FISH is
used for biomonitoring. It is therefore expected that human
exposure to agents affecting repair processes could enhance
interchromosomal differences in response, which could result
in a higher response in certain chromosomes which would be
more sensitive to given genotoxic exposures. Other endogeneous factors, such as an age-related decrease in DNA repair or
genetic syndromes presenting deficient DNA repair, could also
be a source of interchromosomal differences in chromosome
breakage detected by FISH.
Acknowledgements
The contribution of J.S. to this research and his visit to Leiden were made
possible by a long-term post-doctoral fellowship awarded by the Commission
of the European Union, contract no. EV 5V-CT-94-524O. Research in the
A.T.N. group is partly supported by the Commission of the European Union,
Radiation Protection and Environment and Climate programmes.
References
Bird.A.P. (1986) CpG rich islands and the function of DNA methylation.
Nature, 321, 209-213.
Blount.B.C, Mack,M.M., Wehr.C.M., MacGregorJ.T., Hiatt,R.A., Wang.G.,
Wickramasinghe.S.N., Everson.R.B. and Ames3.N. (1997) Folate
deficiency causes uracil misincorporation into human DNA and chromosome
breakage: implications for cancer and neuronal damage. Proc. Natl Acad.
Sci. USA, 94, 3290-3295.
BoeiJJ.W.A, Vermeulen.S. and Natarajan.A T. (1996) Detection of
chromosomal aberrations by fluorescence in situ hybridization in the first
three postirradiation divisions of human lymphocytes. Mutat. Res., 349,
127-135.
CraigJ.M. and Bickmore.W.A. (1994) The distribution of CpG islands in
mammalian chromosomes. Nature Genet., 7, 376-382.
Domfnguez,!., BoeiJJ.W.A., Balajee.A.S. and Natarajan.A.T. (1996) Analysis
of radiation induced chromosome aberrations in Chinese hamster cells by
FISH using chromosome specific DNA libraries. Int. J. Radiat. Biol., 70,
199-208.
Drapkinjt., Sancar,A. and Reinberg.D. (1994) Where transcription meets
repair. Cell, 77, 9-12.
DurantcM., Geri.C, Bonnati.S. and Parenti.R. (1989) Non-random alkylation
of DNA sequences induced in vivo by chemical mutagen. Carcinogenesis,
10, 1357-1361.
Fenech.M. and Neville.S. (1992) Conversion of excision repairable DNA
lesions to micronuclei within one cell cycle in human lymphocytes. Environ.
Mol. Mutagen., 19, 27-36.
Fenech,M., Rinaldi J. and Surralles J. (1994) The origin of micronuclei induced
by cytosine arabinoside and its synergistic interaction with hydroxyurea in
human lymphocytes. Mutagenesis, 9, 273-277.
Friedberg.E.C. (1996) Relationships between DNA repair and transcription.
Annu. Rev. Biochem., 65, 15-42.
Friedberg.E.C., Walker.G.C. and Siede.W. (1995) DNA Repair and
Mutagcnesis ASM Press Washington, DC.
Grigorova^M., Xiao.Y. and Natarajan,A.T. (1997) Analysis of radiation induced
chromosome aberrations in Chinese hamster splenocytes by FISH using
chromosome specific DNA libraries. In preparation.
Hagmar,L. et al. (1994) Cancer risk in humans predicted by increased levels
of chromosomal aberrations in lymphocytes: Nordic study group on the
health risk of chromosome damage. Cancer Res., 54, 2919-2922.
Jeppesen.P. and Turner.B.M. (1993) The inactive X chromosome in female
mammals is distinguished by a lack of histone H4 acetylation, a cytogenetic
marker for gene expression. Cell, 74, 281-289.
Knehr.S., Zitzelsberger.H., Brasselmann.H., Nahrstedt,U. and Bauchinger,M.
(1996) Chromosome analysis by fluorescence in situ hybridization: further
indications of a non-DNA proportional involvement of single chromosomes
in radiation induced structural aberrations. Int. J. Radiat. Biol., 70, 385—392.
LucasJ.N., Awa,A., Straume.T., Poggensee,M., Kodama,Y., Nakano.M.,
Ohtaki.K., Weier.H.U., PinkelJ)., GrayJ. and Littlefield,G. (1992) Rapid
translocation frequency analysis in humans decades after exposure to
ionizing radiation. Int. J. Radiat. Biol., 62, 53—63.
Mirzayans.R., Cubitt,S., EnnsJL, Karimian.K., Radatus.B., Htrani-Hojatti.S.,
Murthy.K. and Paterson.M.C. (1994) Comparative genotoxicity of 2,3'-Ocyclocytidine, p-xylocytidine and 1-fJ-D-arabinofuranosylcytosine in human
tumor cell line. Carcinogenesis, 15, 2319-2324.
Mullenders.L.H.F, Vrieling.H., VenemaJ. and van ZeelandAA. (1991)
Hierarchies of DNA repair in mammalian cells: biological consequence.
Mutat. Res., 250, 223-228.
Natarajan.A.T., Vyas.R.C, WiegantJ. and Jurado^M.P (1991) A cytogenetic
follow-up study of the victims of a radiation in Goiania (Brazil). Mutat.
Res., 247, 103-111
Natarajan,A.T., DarroudiJ\, Vermeulen.S. and WiegantJ (1992a) Frequency
of X-ray and neutron induced chromosomal translocations in human
peripheral blood lymphocytes as detected by m situ hybridization using
chromosome specific DNA libraries. In Sugahara,T., Sagan,L.A. and
Aoyama.T. (eds), Low Dose Irradiation and Biological Defence
Mechanisms. Elsevier Science, Amsterdam, The Netherlands, pp. 343—346
Natarajan^A.T., Vyas.R.C., DarroudiJ. and Vermeulen.S. (1992b) Frequencies
of X-ray induced chromosome translocations in human peripheral
lymphocytes as detected by in situ hybridization using chromosome specific
DNA libraries. Int. J. Radiat. Biol., 61, 199-203.
Natarajan.A.T, BoeiJJ.W.A., Darroudi.F, Van Diemen.P.C.M, DuloutJ1.,
Hande.M.P. and Ramalho.A.T. (1996a) Current cytogenetic methods for
detecting exposure and effects of mutagens and carcinogens. Environ, tilth
Perspect., 104, 445-448.
Natarajan.A.T. et al. (1996b) Mechanisms of induction of chromosomal
aberrations and their detection by fluorescence in situ hybridization. Mutat.
Res., 372, 247-258.
Pinkel.D., Straume.T. and GrayJ.W. (1988) Cytogenetic analysis using
qualitative, high sensitivity, fluorescence hybridization. Proc. Nad Acad.
Sci. USA, 83, 2934-2938.
Ramanathan.B. and Smerdon.MJ. (1989) Enhanced DNA repair synthesis in
hyperacetylated nucleosomes. J. Biol. Chem., 264, 11026-11034.
Sachs,R.K.,Awa, A., Kodama,Y, Nakano.M., Ohtaki.K. and LucasJ.N. (1993)
Ratios of radiation-produced chromosome aberrations as indicators of largescale DNA geometry during interphase. Radiat. Res., 133, 345-350.
Scicchitano.D.A. and HanawaluPC. (1995) Repair of A'-methylpurines in
specific DNA sequences in Chinese hamster ovary cells: absence of strand
specificity in the dihydrofolate reductase gene. Proc. Nad Acad. Sci. USA,
86, 3050-3054.
Shjeocevic.P. and Natarajan.A.T. (1994a) Distribution of X-ray induced G2
chromatid damage among Chinese hamster chromosomes: influence of
chromatid configuration. Mutat. Res., 323, 113-119.
Slijeocevic.P. and Natarajan.A.T. (1994b) Distribution of radiation-induced
Gl exchange and terminal deletion break points in Chinese hamster
chromosomes as detected by G-banding. Int. J. Radiat. Biol., 66, 747-755.
Smerdon.MJ. (1991) DNA repair and the role of chromatin structure. Curr.
Opin. Cell Biol, 3, 422^28.
SurrallesJ. and Natarajan.A.T. (1997) Analysis of XIST function and position
effect of translocations involving the inactive X-chromosome. Proc. Nad
Acad. Sci. USA, submitted for publication.
SurrallesJ., Xamena,N., CreusA and Marcos.R. (1995) The suitability of the
micronucleus assay in human lymphocytes as a new biomarker of excision
repair. Mutat. Res., 342, 43-59.
SurrallesJ., Jeppesen.P., Morrison.H. and Natarajan.A.T. (19%) Analysis of
loss of inactive X chromosome in interphase cells. Am. J. Hum. Genet., 59,
1091-1096.
SurrallesJ., Darroudi,F. and Natarajan.A.T. (1997) Low level of DNA repair
in human chromosome 1 heterochromatin. Genes Chromosomes Cancer,
in press.
TuckerJ.D. and SenftJR. (1994) Analysis of naturally occurring and radiationinduced break points locations in human chromosomes 1, 2 and 4. Radiat.
Res., 140, 31-36.
TuckerJ.D., Lee.D.A. and Moore.D.H. (1995a) Validation of chromosome
staining. II. A detailed analysis of aberrations following high doses of
ionizing radiation in vitro. Int. J. Radiat. Biol., 67, 19—28.
441
J-Sun-allfe, S-Sebastian and A.T.Natarajan
TuckerJ.D.,
Morgan.W.R,
AwaAA.,
Bauchinger.M.,
Blakey.D.,
ComforuvM.N., Littlefield.L.G., Natarajan.A.T. and Shasserre.C. (1995b) A
proposed system for scoring structural aberrations detected by chromosome
staining. Cytogenet. Cell Genet, 68, 211-221.
Turner.B.M. (1993) Decoding the nucleosome. Cell, 75, 5-8.
Verma,R.S. and Babu^A. (1995) Human Chromosomes. Principles and
Techniques. McGraw-Hill, New York, NY.
Wang,W., Sitaram.A. and Scicchitano.D.A. (1995) 3-Methyladenine and 7methylguanine exhibit no preferential removal from the transcribed strand
of the dihydrofolate reductase gene in Chinese hamster ovary B11 cells.
Biochemistry, 34, 1798-1804.
Received on April 25, 1997; accepted on June 25, 1997
442