Photochemistry and Photobiology, 2003, 77(6):
675–679
Rapid Communication
Cross-linking of Histone Proteins to DNA by UV Illumination
of Chromatin Stained with Hoechst 33342{
Sara K. Davis and Christopher J. Bardeen*
Department of Chemistry, University of Illinois, Urbana, IL
Received 7 April 2003; accepted 24 April 2003
ABSTRACT
nm (5). It binds to the minor groove of DNA, with larger affinity
for the AT base pairs, and this binding enhances its fluorescence
quantum yield by 10–40 times (5–7). Although it is known that
H33342 and related dyes can interfere with the action of nuclear
proteins on DNA, e.g. topoisomerase (8), it does not prevent cell
proliferation or growth (9). Because of this, it is commonly used
for cell-sorting applications where the stained cells are later used
for other biological purposes, for example, sex selection (10). But
although H33342 in its electronic ground state is relatively benign,
it is an open question whether the excited-state photochemistry
of H33342 can produce photochemical changes in biological
systems. A previous study found no evidence of direct DNA
damage when cells stained with H33342 were exposed to low
levels of light during a cell-sorting experiment (11). On the other
hand, Hoechst dyes are known to produce reactive species like
free radicals on photoexcitation (12), and H33342 has been
shown to cause cytotoxicity at high concentrations and extended
exposures (13).
In this article, we investigate the photochemistry of H33342
in both isolated cell nuclei and dilute chromatin in buffered solution. On excitation with near-UV light, H33342 can initiate the
formation of cross-links between nuclear proteins and DNA. This
cross-linking can be observed in a number of ways, most dramatically through the apparent fixation of nuclei stained with
H33342 and exposed to 365 nm light. Significant cross-linking
occurs at UV exposures of ;70 J/cm2, and gel electrophoresis
reveals that the cross-linked proteins are predominantly core
histones. The implications of this UV-induced protein–DNA crosslinking for live-cell studies using fluorescence microscopy or flow
cytometry will be discussed.
The photochemical effects of near-UV light on chromatin
labeled with the vital DNA dye Hoechst 33342 (H33342) are
studied. Several types of experiments demonstrate that
illumination at both 365 and 410 nm results in significant
cross-linking of proteins with the DNA. Fluorescence microscopy of dye-stained Xenopus XTC-2 nuclei shows that UV
illumination has effects similar to chemical fixation by formaldehyde. At 365 nm a dose of ~70 J/cm2 results in 50% of the
DNA being cross-linked, as measured by chloroform–sodium
dodecyl sulfate extraction. At 410 nm the efficiency of crosslinking was smaller by a factor of 3. Gel electrophoresis of the
cross-linked proteins shows them to be predominantly core
histones. The implications of these results for experiments on
live cells stained with H33342, for example, fluorescence
microscopy of nuclear dynamics or cell sorting, are discussed.
INTRODUCTION
Interest in DNA-containing chromatin has increased rapidly, because it now appears that chromatin structure and dynamics play an
important role in biological processes like gene transcription (1–3).
The ability to image DNA using fluorescence microscopy is critical for studying the structure and dynamics of chromatin in the
nuclei of live cells. This ability is made possible by the different fluorescent dye molecules that can bind DNA selectively and
with large affinity (4). Although there are many such molecules
available for labeling DNA in vitro, the inability of most of these
molecules to pass through live-cell membranes has limited the
number of live-cell DNA stains to a relatively small subset of
molecules. Of these, one of the most widely used is 2,59-bi-1Hbenzimidazole,29-(4-ethoxyphenyl)-5-(4-methyl-1-piperazinyl)-trichloride, commonly known as Hoechst 33342 (H33342), whose
structure is shown in the inset of Fig. 1. H33342 absorbs in the
near-UV region and has a broad emission centered at around 500
MATERIALS AND METHODS
Nuclear isolation. Xenopus XTC-2 cells were grown in phenol red–free
70% Dulbecco’s Modified Eagle’s Medium with F-12 nutrient mixture
(DME–F12) (Sigma, St. Louis, MO) supplemented with 10% fetal bovine
serum (GIBCO, Grand Island, NY) at room temperature. For nuclear
isolation, cells were resuspended in cell media and then rinsed with ice-cold
70% phosphate-buffered saline (BioWhittaker, Walkersville, MD). Subsequent steps were done on ice. After centrifugation, the cell pellet was
loosened, and mammalian cell lysis reagent (Pierce, Rockford, IL) was
added. The sample was agitated for 30 s, and then the cell membranes were
disrupted by resuspending the solution twice with a glass pipette. An excess
of cell media was added, and then the nuclei were centrifuged once and
resuspended in regular cell media. To examine salt-dependent morphology
changes, nuclei were treated (fixed or UV exposed) in regular cell media
{Posted on the website on 1 May, 2003.
*To whom correspondence should be addressed at: Department of
Chemistry, University of Illinois, 600 South Mathews Avenue, Urbana,
IL 61801, USA. Fax: 217-244-3186; e-mail: [email protected]
Abbreviations: H33342, Hoechst 33342; PAGE, polyacrylamide gel
electrophoresis; SDS, sodium dodecyl sulfate.
Ó 2003 American Society for Photobiology 0031-8655/03
$5.00 þ 0.00
675
676 Sara K. Davis and Christopher J. Bardeen
Figure 1. Absorption spectra of chromatin and H33342. Solid line is
chromatin with bound H33342 (20 base pairs per dye). Dashed line is
chromatin in solution with no H33342. Arrows denote 365 and 410 nm.
Inset is chemical structure of H33342.
and then resuspended in an excess of media with 2 M NaCl. After
incubating for about 30 min, this solution was spread on a coverslip and
imaged using a fluorescence microscope with a charge-coupled device
camera.
Chromatin isolation. After nuclear isolation, chromatin was isolated using
a slightly modified method of Bhorjee and Pederson (14). Briefly, nuclei were
rinsed three times, resuspended in 10 mM NaCl, 1.5 mM MgCl2 and 10 mM
Tris–HCl (pH 7.0) and then disrupted by sonication with three 10 s pulses at
20 kHz and 20 W (Sonics and Materials VCX 600, Danbury, CT). The
resulting solution was layered over 30% sucrose in NaCl–Tris buffer (10 mM
NaCl and 2.5 mM Tris–HCl, pH 7.2) and centrifuged at 4500 g for 15 min
(Sorvall RC-5 centrifuge, SS-34 rotor, Newtown, CT). The top layer was
removed, and aliquots of 5.5 mL were layered over 15.5 mL 60% sucrose in
NaCl–Tris buffer with 24 mM ethylenediaminetetraacetic acid in nitrocellulose tubes. The top two-thirds of the tubes were gently mixed, and then the
samples were centrifuged for 90 min at 131 000 g (Beckman L7-65 centrifuge,
60Ti rotor, Fullerton, CA). The pellet was resuspended in and then dialyzed
overnight against NaCl–Tris buffer. The absorbance ratio at 260 and 240 nm
of the resulting chromatin solution was 1.3.
H33342 staining. For each experiment, a freshly prepared solution of
H33342 (Sigma) in water was used. To stain isolated nuclei and chromatin,
an aliquot from the stock solution was added to the sample such that the final
H33342 concentration was 9 lM. For the chromatin samples, the resulting
solution had a base pair–dye ratio of 20:1, as calculated using e260 ¼ 7260
M1 cm1 for DNA in chromatin (15) and e348 ¼ 42 000 M1 cm1 for
H33342 (16). Samples were incubated for 15 min before UV irradiation.
UV irradiation. UV irradiation was done with samples (1 mL isolated
nuclei in cell medium or 0.25 mL chromatin in NaCl–Tris buffer,
A260 ¼ 1:3) in a 1 3 1 cm quartz cuvette, suspended over an open objective
socket in an Olympus IX-70 (Melville, NY) inverted microscope with
a fluorescence observation attachment. The sample was placed where the
lamp arc images were overlapped and collimated. The UV image was
slightly smaller than 1 3 1 cm, but we assume that diffusive mixing leads to
uniform sample exposure during the experiment. In this case, the intensity
of the light seen with the solution is 70 mW/cm2 at 365 nm and 60 mW/cm2
at 410 nm.
Chloroform–sodium dodecyl sulfate extraction. Quantitation of DNA–
protein cross-linking in isolated chromatin was done using a chloroform
extraction assay (15,17). After UV irradiation, samples were adjusted to
0.2% sodium dodecyl sulfate (SDS) and 1.0 M NaCl. Samples were split
into three tubes and incubated at 608C for 10 min. An equal volume of
chloroform–isoamyl alcohol (24:1) was added, and the solutions were
vortexed. Samples were then centrifuged at 1000 rpm for 5 min. With this
treatment, proteins and DNA cross-linked to proteins are at the cloudy
interface, and free DNA is in the aqueous phase. The A260 of the aqueous
phase was measured to quantify the amount of DNA that was not crosslinked and averaged for the three tubes of chromatin to eliminate any effects
from small differences in extraction conditions. The absorption spectra were
recorded in a microcell using a HP 8452A (Palo Alto, CA) diode array
spectrophotometer. The reported percentage of cross-linked DNA was
calculated by comparison with the A260 of a sample that had not been
irradiated.
SDS–polyacrylamide gel electrophoresis. SDS–polyacrylamide gel electrophoresis (PAGE) of isolated nuclei was done using a modified procedure
of Laemmli (18,19). Half of each sample was incubated in cell medium with
a total NaCl concentration of 2 M for 2 h at room temperature. All samples
were rinsed with serum-free medium without additional salt, and the pellets
were manually disrupted in 20 lL serum-free medium. Control samples did
not form a solid pellet after the 2 M NaCl treatment; the rinse solution was
removed such that the final volume was approximately 20 lL. An equal
volume of loading sample buffer was added to each sample, and the
solutions were incubated in boiling water for 10 min. After this, the nuclei were disrupted, and the solutions were clear. Fifteen microliters of
each sample was loaded onto a vertical 3% stacking gel (acrylamide–
bisacrylamide, 10:0.5% [wt/wt], in 0.12 M Tris–HCl, pH 6.8) with an 18%
separating gel (acrylamide–bisacrylamide, 30:0.15% [wt/wt], in 0.75 M
Tris–HCl, pH 8.8). Electrophoresis was performed at 30 mA, and the gel
was stained with Coomassie blue to detect the proteins that leave the
chromatin (which is trapped in the well) and migrate through the gel.
RESULTS
Figure 1 shows the absorption spectrum of H33342-stained
chromatin, with the characteristic peaks of H33342 at approximately 350 nm, the absorption of the DNA at 260 nm and the large
shoulder of the shorter-wavelength protein absorption at 220 nm
and below. There is also a contribution to the spectrum from the
histone proteins at 280 nm. The histone absorption leads to
absorption ratios A260:A280 and A260:A240 for unstained chromatin
that are different from those observed in bare DNA: in our samples,
the absorbance ratio A260:A280 is 1.5 (17), and the absorbance ratio
A260:A240 is 1.3 (14). We note that the absorption of Hoechst dyes
depends on the level of staining (20) and that high concentrations
can lead to aggregate formation and different binding motifs (21–
23). In all experiments reported in this study, the maximum
staining level is 1 dye per 20 base pairs, as determined by the
absorption spectrum. At this staining level, which is typical of what
is used in microscopy and cytometry experiments, the effects of
aggregation should be negligible (22). In Fig. 1, arrows mark the
two wavelengths used to excite the stained chromatin in our
microscope, namely 365 and 410 nm. The dashed line indicates the
absorption of the chromatin alone, which has no measurable
absorption at wavelengths greater than 350 nm. That these
wavelengths do not affect bare chromatin was confirmed using
control experiments, where neither the salt-dependent nuclear
morphology nor the amount of cross-linked DNA was affected by
prior illumination at 365 nm in the absence of H33342. In Fig. 2
we show the effect of protein–DNA cross-linking on the saltdependent morphology of the intranuclear DNA. The bright areas
of the images correspond to regions of high H33342 fluorescence
and thus high chromatin density. As the ionic strength of the
medium is increased, the binding of the histones to specific regions
of DNA is weakened because of electrostatic screening effects,
which leads to nucleosome sliding at low ionic strength and
eventually histone dissociation at high ionic strength (24,25). The
structured intranuclear chromatin visible in Fig. 2a at physiological
ionic strength disappears as the NaCl concentration increases, until
at 2 M NaCl it is completely lost, and the nuclear DNA forms
a uniform, diffuse mass as shown in Fig. 2b. This loss of structure
can be prevented by either treatment with formaldehyde (Fig. 2c)
or irradiation at 365 nm (Fig. 2d) or 410 nm (Fig. 2e). The fact that
both formaldehyde fixation and UV exposure prevent the saltinduced changes in nuclear morphology suggests that both cause
similar chemical changes. Formaldehyde is known to induce
Photochemistry and Photobiology, 2003, 77(6) 677
Figure 2. Fluorescence images of H33342-stained chromatin in isolated
nuclei. Light regions correspond to high fluorescence, indicating the presence of DNA. The dark circular spot visible in some of the nuclei is the
nucleolus. The nuclei are exposed to the following conditions: in cell media
(80 mM NaCl) (a), in cell media with 2 M NaCl (b), fixed for 15 min in 4%
formaldehyde solution in cell media and then stained and resuspended in
cell media with 2 M NaCl (c), stained, exposed to 365 nm UV irradiation
and then resuspended in cell media with 2 M NaCl (d) and stained, exposed
to 410 nm UV irradiation and then resuspended in cell media with 2 M
NaCl (e). The scale bar is 10 lm.
a broad variety of cross-links, including protein–protein, DNA–
DNA and protein–DNA cross-links. It appears that photoexcited
H33342 also produces cross-linked species that preserve chromatin
structure.
The amount of cross-linking can be quantified by exposing
H33342-stained chromatin to variable doses of UV light and then
extracting the cross-linked DNA using a SDS–chloroform solution.
Previous workers have shown that this is a very sensitive way to
measure the fraction of DNA linked to protein (15). Figure 3 shows
how the fraction of DNA removed from the aqueous phase varies
with exposure time under our experimental conditions. After 15–20
min, corresponding to a total dose of ;70 J/cm2, 50% of the DNA
can be removed from the aqueous phase. This does not necessarily
mean that 50% of the base pairs have formed covalent bonds to
proteins, because even one cross-link can result in a sizable length
of DNA being pulled into the chloroform phase with its attached
protein. It does suggest that at these fluence levels, H33342 is
capable of cross-linking proteins to a large fraction of the available
DNA. Figure 3 shows that the cross-linking yield levels off below
100%, which is probably due to the relatively low H33342 staining
level. Other studies have shown that the asymptotic level of crosslinking depends on the photosensitizer concentration (17).
It is instructive to calculate the number of absorption events
required to achieve significant cross-linking. Assuming low excitation rates, so that the probability that a molecule can be excited
twice during its excited-state lifetime is negligible, the number of
photons absorbed per H33342 molecule, Nabs, is given by the
relation (26)
I
t;
ð1Þ
Nabs ¼ r
hm
where r is the absorption cross section, I is the incident intensity, h
is Planck’s constant, m is the frequency and Dt is the period of
illumination. The absorption cross section for H33342 at 365 nm is
calculated from the molar absorptivity e by unit conversion from
Figure 3. Quantitation of DNA–protein cross-linking in isolated chromatin
after 365 nm irradiation at an intensity of 70 mW/cm2. The percent crosslinked is obtained by measuring the amount of DNA that remains in the
aqueous phase after chloroform–SDS extraction relative to the amount in
a sample that was not irradiated.
per molar per centimeter to square centimeter using Avogadro’s
number. This relationship is given by r ¼ 1:661 3 1021 e, and for
the case of H33342 at 365 nm, it is 7:0 3 1017 cm2 . Equation (1)
then leads to a value of 104 photons absorbed on average by
a H33342 molecule after a 20 min dose of 365 nm irradiation at 70
mW/cm2. Under irradiation at 410 nm, we find that the efficiency
of the cross-linking is lower by approximately a factor of 3, after
taking into account the reduced absorption at this wavelength. The
fact that the cross-linking efficiency is wavelength dependent may
be the result of energy-dependent branching ratios in the excited
state of H33342 or may indicate the presence of different types of
absorbing species, which would likely reflect the variation in
binding sites for the dye (21–23).
To determine the identity of the proteins cross-linked to DNA by
H33342, we take the chromatin from irradiated nuclei and perform
SDS-PAGE to look at individual protein bands. The gel electrophoresis of chromatin is well established, and the individual bands
corresponding to the various histones can be easily identified (19).
Figure 4 shows a gel of nuclear chromatin after 10 min of exposure
to 365 nm light. If the chromatin is loaded directly onto the gel,
only small differences are seen between the UV-exposed sample
and the unexposed control, as seen in Lanes 1 and 2. The main
change is the appearance of weak dimer bands due to histone–
histone cross-linking near the H1 band, similar to what is observed
in other cross-linking experiments involving formaldehyde (27) or
methylene blue (17,28). Histone–DNA cross-linking should lead to
a decrease in the intensity of the histone bands because the crosslinked proteins are expected to be bound to the chromatin in the
well and not be able to migrate down the gel. The fact that the
histone bands in Lane 2 are not attenuated is likely due to the
severe preparation conditions (boiling the sample for 10 min),
which can disrupt the DNA–protein cross-links. Such reversible
protein–DNA cross-links have been observed previously in
formaldehyde cross-linked systems (27). To circumvent this
problem, we used a 2 M NaCl solution to extract the free proteins
from the nuclei before gel preparation. In this case, only the
proteins that have been cross-linked to the intranuclear DNA will
remain in the sample after extraction. Indeed, this is what is
observed in Lanes 3 and 4, where the control sample has no
678 Sara K. Davis and Christopher J. Bardeen
served for extended periods of time. Prolonged exposure to the excitation light may affect the ability of the chromatin to move or
undergo conformational changes that require displacement of the
histones, with the result that the nuclear DNA may appear more
stationary than it would actually be in an unperturbed living
system. In flow cytometry experiments the exposure time is
a fraction of a second and the probability of damage is much
smaller in general, although it would depend on the laser power as
well. Thus, the ability of H33342 to photochemically cross-link
the histones and DNA should be taken into account in the interpretation of live-cell–imaging experiments or the design of cellsorting procedures. A final observation is that this photochemical
cross-linking could also be used to probe the role of nucleosomal
sliding and mobility in cellular processes by providing a way to
selectively alter such dynamics in localized regions of live cells.
Figure 4. SDS-PAGE of isolated nuclei. Lane 1, no UV irradiation; Lane
2, 10 min, 365 nm irradiation; Lane 3, no UV irradiation, after 2 M NaCl
histone extraction; and Lane 4, 10 min, 365 nm irradiation, after 2 M NaCl
histone extraction. The positions of the protein molecular weight standards
in kilodaltons are labeled on the right.
Acknowledgements—Support from the Research Corporation (Research
Innovation Award RI0436) and the Illinois Research Board is gratefully
acknowledged. Experimental assistance was provided by Andinet Amare.
We thank Caroline Christian for help with the gel electrophoresis and the
Gruebele research group for assistance with various biochemistry procedures.
REFERENCES
chromatin proteins remaining after extraction, whereas the UVexposed sample has significant amounts of the core histones,
especially H3, H2A, H2B and H4. The large protein bands at the
top of the gels, which are especially pronounced in the control
sample Lane 3, result from incomplete removal of the cell medium
serum proteins and are not related to the chromatin sample.
Interestingly, we do not observe an enhancement in the linker
histone H1 band. Minor groove binders like H33342 are known to
disrupt linker histone binding to DNA (29), and it may be that if
H33342 binds to the internucleosomal DNA, it displaces H1 and
cannot induce cross-links between H1 and the DNA.
In conclusion, we find that H33342 can photosensitize the crosslinking of core histone proteins and DNA with reasonable efficiency at staining levels similar to those used in fluorescence
microscopy experiments. This has immediate consequences for the
use of H33342 in live-cell experiments or cell-sorting procedures,
where the cell may be exposed to UV light for prolonged periods
of time. Although exposure of a H33342-stained cell to UV light
does not directly damage the DNA itself via strand cleavage or
interstrand cross-links, it does create chemical changes in the
chromatin, which contains the DNA in living systems. If the
chromatin experiences significant levels of cross-linking between
the DNA and histone proteins, the dynamics of the DNA may be
adversely affected. Cellular processes that rely on the chemical
accessibility of DNA, such as gene transcription and DNA
replication, may be slowed down or prevented by the presence of
immobile histones, which block access to the underlying DNA. For
example, nucleosomal sliding (30) is thought to be necessary for
various DNA reactions in living systems, and it is reasonable to
expect that the motion of the histone octamers would be prevented
by cross-linking to specific regions of the chromosomal DNA.
Clearly, the duration and intensity of UV exposure for live cells
stained with H33342 and related chromophores should be controlled to ensure that the chromatin retains its native structure
throughout the course of the experiment. This is a concern especially
in fluorescence microscopy experiments, where the cells are ob-
1. Cremer, T. and C. Cremer (2001) Chromosome territories, nuclear
architecture and gene regulation in mammalian cells. Nat. Rev. Genet. 2,
292–301.
2. Gasser, S. M. (2002) Visualizing chromatin dynamics in interphase
nuclei. Science 296, 1412–1416.
3. Horn, P. J. and C. L. Peterson (2002) Chromatin higher order folding:
wrapping up transcription. Science 297, 1824–1827.
4. Arndt-Jovin, D. J. and T. M. Jovin (1989) Fluorescence labeling and
microscopy of DNA. Methods Cell Biol. 30, 417–448.
5. Cosa, G., K.-S. Focsaneanu, J. R. N. McLean, J. P. McNamee and J. C.
Scaiano (2001) Photophysical properties of fluorescent DNA-dyes
bound to single- and double-stranded DNA in aqueous buffered solution. Photochem. Photobiol. 73, 585–599.
6. Latt, S. A. and J. C. Wohlleb (1975) Optical studies of the interaction of
33258 Hoechst with DNA, chromatin, and metaphase chromosomes.
Chromosoma 52, 297–316.
7. Sandhu, L. C., R. L. Warters and L. A. Dethlefsen (1985) Fluorescence
studies of Hoechst 33342 with supercoiled and relaxed plasmid pBR322
DNA. Cytometry 6, 191–194.
8. Smith, P. J., S. M. Bell, A. Dee and H. Sykes. (1990) Involvement of
DNA topoisomerase II in the selective resistance of a mammalian cell
mutant to DNA minor groove ligands: ligand-induced DNA-protein
crosslinking and responses to topoisomerase poisons. Carcinogenesis
11, 659–665.
9. Durand, R. E. (1982) Use of Hoechst 33342 for cell selection from
multicell systems. J. Histochem. Cytochem. 30, 117–122.
10. Johnson, L. A., G. R. Welch, K. Keyvanfar, A. Dorfmann, E. F. Fugger
and J. D. Schulman (1993) Gender preselection in humans? Flow
cytometric separation of X and Y spermatozoa for the prevention of Xlinked diseases. Hum. Reprod. 8, 1733–1739.
11. Catt, S. L., D. Sakkas, D. Bizzaro, P. G. Bianchi, W. M. C. Maxwell
and G. Evans (1997) Hoechst staining and exposure to UV laser during
flow cytometric sorting does not affect the frequency of detected
endogenous DNA nicks in abnormal and normal human spermatozoa.
Mol. Human Reprod. 3, 821–825.
12. Gorner, H. (2001) Direct and sensitized photoprocesses of bisbenzimidazole dyes and the effects of surfactants and DNA. Photchem.
Photobiol. 73, 339–348.
13. Durand, R. E. and P. L. Olive (1982) Cytotoxicity, mutagenicity and
DNA damage by Hoechst 33342. J. Histochem. Cytochem. 30, 111–
116.
14. Bhorjee, J. S. and T. Pederson (1973) Chromatin: its isolation from
cultured mammalian cells with particular reference to contamination by
nuclear ribonucleoprotein particles. Biochemistry 12, 2766–2773.
15. Mandel, R., G. Kolomijtseva and J. G. Brahms (1979) DNA-protein
Photochemistry and Photobiology, 2003, 77(6) 679
16.
17.
18.
19.
20.
21.
22.
interactions in nucleosomes and in chromatin. Eur. J. Biochem. 96,
257–265.
Lakowicz, J. R., I. Gryczynski, H. Malak, M. Schrader, P. Engelhardt,
H. Kano and S. W. Hell (1997) Time-resolved fluorescence spectroscopy and imaging of DNA labeled with DAPI and Hoechst 33342 using
three-photon excitation. Biophys. J. 72, 567–578.
Lalwani, R., S. Maiti and S. Mukherji (1995) Involvement of H1 and
other chromatin proteins in the formation of DNA-protein cross-links
induced by visible light in the presence of methylene blue.
J. Photochem. Photobiol. B: Biol. 27, 117–122.
Laemmli, U. K. (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680–685.
Thomas, J. O. (1998) Isolation and fractionation of chromatin and linker
histones. In Chromatin: A Practical Approach (Edited by H. Gould),
p.19. Oxford University Press, Oxford.
Latt, S. A. and G. Stetten (1976) Spectral studies on 33258 Hoechst and
related bisbenzimidazole dyes useful for fluorescence detection of
deoxyribonucleic acid synthesis. J. Histochem. Cytochem. 24, 24–33.
Stokke, T. and H. B. Steen (1985) Multiple binding modes for Hoechst
33258 to DNA. J. Histochem. Cytochem. 33, 333–338.
Loontiens, F. G., P. Regenfuss, A. Zechel, L. Dumortier and R. M.
Clegg (1990) Binding characteristics of Hoechst 33258 with calf
thymus DNA, poly[d(A-T)], and d(CCGGAATTCCGG): multiple
23.
24.
25.
26.
27.
28.
29.
30.
stoichiometries and determination of tight binding with a wide spectrum
of site affinities. Biochemistry 29, 9029–9039.
Breusegem, S. Y., S. E. Sadat-Ebrahimi, K. T. Douglas, R. M. Clegg
and F. G. Loontiens (2001) Increased stability and lifetime of the
complex formed between DNA and meta-phenyl-substituted Hoechst
dyes as studied by fluorescence titrations and stopped-flow kinetics.
J. Mol. Biol. 308, 649–663.
Widom, J. (1998) Structure, dynamics, and function of chromatin
in vitro. Annu. Rev. Biophys. Biomol. Struct. 27, 285–327.
Spadofora, C., P. Oudet and P. Chambon (1979) Rearrangement of
chromatin structure induced by increasing ionic strength and temperature. Eur. J. Biochem. 100, 225–235.
Andrews, D. L. (1986) Lasers in Chemistry. Springer-Verlag, Berlin.
Jackson, V. (1999) Formaldehyde cross-linking for studying nucleosomal dynamics. Methods 17, 125–139.
Liu, Z. H., B. Sargueil and C. W. J. Smith (2000) Methylene bluemediated cross-linking of proteins to double-stranded RNA. Methods
Enzymol. 318, 22–33.
Kas, E., L. Poljak, Y. Adachi and U. K. Laemmli (1993) A model for
chromatin opening: stimulation of topoisomerase II and restriction
enzyme cleavage of chromatin by distamycin. EMBO J. 12, 115–126.
Becker, P. B. (2002) Nucleosome sliding: facts and fiction. EMBO J.
21, 4749–4753.