| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics |
Department of Experimental Oncology, Istituto Nazionale Tumori, 20133 Milan [M. B., R. F., E. B., G.C.], and G. Moruzzi Department of Biochemistry, Bologna University, 40126 Bologna [G. C.], Italy
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
promoted cleavage patterns that were much more similar to those of
Drosophila top2 (both in vitro and
in vivo) than human top2ß. Moreover, da-IDA showed a
marked site-dependent preference for human top2ß. Thus, DNA site
selection in vivo is different for the test
anthracyclines, and together with a degree of ß-form specificity, may
affect drug activity in human cells. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
For chemically related top2 poisons, a good correlation exists between the extent and persistence of cellular DNA cleavage and the cytotoxic potency of drugs (1, 3, 4, 6) . Interestingly, the cleavage:cytotoxicity ratio varies greatly among unrelated poisons and is lowest for anthracyclines (1, 3, 6, 7) . Therefore, DNA breaks promoted by anthracyclines might be more lethal than those promoted by other top2 poisons; however, this remains to be proved fully. top2 poisons stimulate in vitro DNA breaks in a sequence-selective manner (1–4) , and the site selectivity of anthracyclines has been studied extensively (3) . In particular, doxorubicin requires 5'TA3' dinucleotides at the 3' terminus of the strand cut, corresponding to -2 and -1 positions, respectively (3, 8) . In contrast, much less is known for base sequence preferences in the chromatin of living cells. We showed previously that dh-EPI, a potent anthracycline analogue, had a distinct locus and site specificity in the genome of Drosophila Kc cells (9, 10) . In contrast to VM-26 and clerocidin, dh-EPI was completely unable to stimulate cleavage in the heterochromatic satellite III DNA, whereas it was very active at the histone gene locus (9, 10) . However, drug nucleotide preferences remained to be established in vivo.
Thus, to determine the in vivo base preferences of
anthracycline activity, we have investigated at nucleotide levels
cleavage sites of dh-EPI and da-IDA (Fig. 1A
) at the histone gene cluster of Kc cells and in
vitro with Drosophila and human top2s. The two studied
analogues are closely related to doxorubicin structure (11, 12)
and mainly differ at the 3 and 4 position of the sugar (Fig. 1A
). The histone gene cluster was selected for this study
because top2-dependent DNA cleavage is much more easily detected
in vivo in multicopy DNA regions and is a locus of high
activity of top2 poisons (9, 10)
. dh-EPI and da-IDA had
different cleavage:cytotoxicity ratios in human HL60 leukemic cells; in
particular, DNA lesions stimulated by da-IDA were more lethal than
those of dh-EPI (6)
. Because the two analogues also
stimulated distinct cleavage patterns with murine top2 (11, 12)
, genomic localization of cleavage may be a significant
factor influencing drug activity. We show that the two analogues mainly
maintain the in vitro base requirements in nuclear chromatin
of Drosophila cells. Moreover, da-IDA showed a marked
site-dependent preference for human top2ß.
|
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-32P]ATP was purchased from Amersham
(Milan, Italy). Other enzymes were from New England Biolabs (Taunun,
Germany). Drosophila melanogaster top2 was purchased from
United States Biochemical Corp. (Cleveland, OH). Human top2s were
purified from Saccharomyces cerevisiae JEL-1
(13)
cells carrying YEpWOB6 (provided by C. Austin,
Newcastle-upon-Tyne University, Newcastle-upon-Tyne, United Kingdom) or
YEphTOP2ß-165S plasmids that expressed top2 and top2ß,
respectively (13, 14) . Enzymes were purified and stored as
described already (14, 15)
. YEphTOP2ß-165S plasmid,
which encodes for a wild-type top2ß, was obtained mutating the
Arg165 to Ser of YEphTOP2ß plasmid, which
encodes for a mutant enzyme (16)
.
Low-Resolution Mapping of Cleavage Sites.
Drosophila Kc cells were cultured as described (9)
.
Exponentially growing cells were treated with 10 µM
dh-EPI or da-IDA for 30 min at 25°C. Then, cells were centrifuged;
lysed in 0.1% SDS, 20 mM Tris-HCl (pH 8), and 0.5 mg/ml
proteinase K; and incubated overnight at 37°C. DNA was purified as
reported (9, 10)
and digested with HindIII to
generate the 5-kb major repeat of the histone gene cluster (Fig. 1C
). Agarose gel electrophoresis and Southern blotting were
performed by standard methods. 32P-labeled HP or
HB probes (Fig. 1C
) were used to map cleavage sites in the
SAR or H2B-H3 region, respectively (9)
.
High-Resolution Mapping of Cleavage Sites.
Mapping of DNA breaks at nucleotide levels was based on premature
termination of primer extension because of drug-dependent interruption
of templates and on thermostable, nonproofreading Taq DNA polymerase
(9, 10, 17) . Genomic DNAs from control and drug-treated
cells were digested with MspI and HaeII for the
intergenic H2A-H2B and H2B coding regions, respectively. Four primers
were used for the H2A-H2B intergenic region: upper strand, D10-1
(5'-CACTTTGCCACCTTTTCCACG) and D10-3 (5'-CGTCCAGACATTTTGCTTTGCG) from
160 to 180, and 179 to 200 positions, respectively; lower strand, D10-2
(5'-GTTTTCGGAGGCATTGTTCAC) and D10-4 (5'-CTTGGTGTTCTTCTGAGCC), from 429
to 409, and 481 to 460 positions, respectively (GenBank-EMBL X14215).
Primers for the H2B coding region were: D11-1 (upper strand,
5'-GTGAACAATGCCTCCGAAAACTAGTGG), from 409 to 435 positions, and D11-2
(lower strand, 5'-GCTTCGGCAGCAATTCGCTCG), from 639 to 619 positions.
Primer extension was performed using 1 µg of genomic DNAs and
5 x 106 cpm/sample of labeled
primers. Taq DNA polymerase (1.25 units) was added in 50 µl of 10
mM Tris-HCl (pH 8.3), 50 mM
KCl, and 1.5 mM MgCl2
supplemented with 10 µM of each deoxynucleotide
triphosphate. Identical results were obtained when using other
polymerases without proof-reading activity (9, 10)
.
Samples were subjected to 30 cycles of 1/1/3 min at 94/58/72°C,
followed by a final 15-min extension at 72°C. Then, samples were
extracted with phenol/chloroform; precipitated in ethanol; resuspended
in 2.5 µl of 80% formamide, 10 mM NaOH, 1
mM EDTA, and 0.1% dyes; heated at 90°C for 2
min; chilled on ice; and loaded onto 8% polyacrylamide gels. Sequence
ladders were obtained as described already (9)
. Dried gels
were exposed to phosphor screens, and images were taken with a
PhosphorImager 425 model (Molecular Dynamics, Sunnyvale, CA). DNA cuts
were mapped, taking into account that Taq polymerase adds an
untemplated nucleotide to 3' ends of extended primers (9, 18)
.
DNA Cleavage Assay with Purified Drosophila and
Human top2.
A H2A-H2B intergenic fragment was amplified by PCR with Pfu
polymerase and cloned into pUC18 plasmid. Plasmid DNA was purified,
restricted with BamHI or EcoRI, and uniquely
5'-end-32P-labeled with T4 kinase as described
previously (11)
. The labeled fragment was used as
substrate for cleavage reaction in 20 µl of 40
mM Tris-HCl (pH 7.5), 80 mM
KCl, 10 mM MgCl2, 0.5
mM DTT, 1 mM ATP, 15
µg/ml BSA, and drugs. Drosophila top2 (2–5 units/sample;
as defined by the supplier) or human top2 isoforms [30 units/sample;
one unit as defined by Cornarotti et al. (19)
]
were added at 30°C or 37°C for 20 min, respectively. Reactions were
stopped by 1% SDS, 0.1 mg/ml proteinase K and incubated at 42°C for
45 min. Cleavage sites were analyzed by sequencing gels as described
above.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), for 30 min at 25°C. The tested drugs
stimulated different cleavage patterns in vitro, with murine
top2 using SV40 DNA as substrate (11)
. Structure-activity
relationships established that structural changes at the sugar 3'
position were responsible for altered site selectivity (11, 12)
. dh-EPI, which has a 3'-OH group, has the same sequence
specificity of doxorubicin, whereas da-IDA, which has no 3'
substituent, stimulated distinct cleavage intensity patterns
(11)
.
Double-stranded cleavage sites were mapped by the indirect end-labeling
method (9, 10)
. Cleavage patterns were similar but not
identical between the two analogues in the SAR and H2A-H2B intergenic
region (Fig. 1B
). Overall, dh-EPI promoted more cleavage
sites than da-IDA, which was thus more selective. At the SAR, da-IDA
stimulated cleavage at a subset of the sites detected with dh-EPI (Fig. 1, B
, left panel, and C). In the
H2A-H2B region, da-IDA stimulated sites 9, 10, and 11, whereas dh-EPI
stimulated sites 9 and 10 and several others in the H3-H4 region (Fig. 1B
and not shown; see also Refs. 9
and
10
). It must be noted that site 11 was specific for
da-IDA, because dh-EPI-dependent cleavage was very weak (Fig. 1
B,
right panel), and mapped in the middle of the H2B gene
(Fig. 1C
). Thus, the molecular actions of two closely
related anthracyclines could be distinguished at the histone gene
cluster of Drosophila Kc cells.
Analogue-specific Sites at Nucleotide Levels.
To establish the in vivo sequence specificity of the studied
drugs, we sequenced cleavage sites by primer extension of genomic DNAs
(9, 10, 17) . We selected the H2A-H2B intergenic region and
the H2B coding sequence, which corresponded to low-resolution sites 10
and 11, respectively (Fig. 1C
). Cuts were mapped at
nucleotide levels in both strands. Each strand was extended with two
distinct and adjacent primers (see "Materials and Methods"), and
extension stops specifically present in the "drug lanes" were
detected with either primer at identical nucleotide positions (Fig. 2
and not shown). Thus, primer mispairing was unlikely to be the cause of
the observed drug-dependent stops. Rather, they were likely
attributable to the poison-dependent trapping of top2-DNA complexes
that yielded interrupted templates (9, 10, 17)
.
|
). dh-EPI, but not da-IDA, stimulated
other sites throughout the region. A rather high cleavage level was
observed close to the CAP box of H2A (sites i and
j; Fig. 3A
and not shown), and cleavage (site x) was weakly
detected at the TATA box of H2B (Figs. 2 and 3
A). A striking
difference was observed in the H2B coding sequence (Fig. 2B
). Site 
was the strongest one for da-IDA, whereas site
µ was strongest for dh-EPI. Moreover, sites stimulated by the two
analogues were not at the same position but 1 bp far apart from each
other (Figs. 2B
and 3B). dh-EPI also stimulated
weaker sites (
and 
; Figs. 2B
and 3B).
|
. Only in case of da-IDA, the site on the top strand, had
a 4-bp-staggered cut on the other strand (Fig. 3B
), in
agreement with the observed double-strand cut (site 11) in agarose gels
(Fig. 1B
). In the case of dh-EPI, the sites on
the two strands were not staggered by 4 bp. Thus, in contrast to
da-IDA, dh-EPI stimulated cleavage at least at two distinct and
alternative sites characterized by very weak cuts on one strand (Fig. 3B
). dh-EPI action at nucleotide levels was consistent with
an analogue-specific lack of double-stranded cleavage as observed in
agarose gels (Fig. 1B
). In the H2A-H2B intergenic region,
sites n and u had a corresponding cut on the other strand with a 4-bp
stagger for either analogue (Fig. 3A
), consistent with
double-stranded cleavage. It is likely that these two sites were mainly
responsible for site 10 (Fig. 1B
). Overall, 15 of 25 were
"single" sites because they were not paired with a corresponding
cut at the complementary strand. Because several of them were weak, the
break on the other strand might have not been strong enough to be
detected in our assay.
Base Preferences of in Vivo Anthracycline Action.
Forty-seven dh-EPI-stimulated sites were collected, and base
preferences could be determined. Nonrandom base distributions were
observed at positions -1 and +1 by 
2
analyses, where adenines and guanines were strongly preferred,
respectively (Table 1
). Of the 47 sites, 23 (49%) had at least an adenine at position -1,
and 3 (6%) had an adenine at the -1 position of the complementary
strand only. Of the sites without adenines at -1 positions of both
strands ([-1]noA sites), 14 (30%) were 1–2 bp close to
sites with adenines at -1 positions ([-1]A sites), and 7 (15%)
were more distant from other sites (Fig. 3)
. Indeed, some breaks
(20)
constituted nine clusters of two to three
adjacent cuts, 1–2 bp close to each other (Fig. 3)
, suggesting a
reciprocal influence among very close sites (see "Discussion").
Because the majority of in vivo sites had adenines at -1
positions, a further statistical analysis was then made by considering
only the base sequences of [-1]A sites (Table 1
, lower part).
Although the sequences were limited, we could derive an in
vivo consensus for dh-EPI action that completely agrees with the
in vitro doxorubicin consensus (3, 8)
showing a
5'TA dinucleotide highly preferred at both
3'-termini of strand cuts (Table 1)
. Interestingly, a guanine was
strongly preferred at position +1 (Table 1)
, and this was unexpected
based on in vitro data (3)
. These data show
that dh-EPI mainly maintained in nuclear chromatin the in
vitro-established base requirements (3, 8)
.
Nevertheless, a significant fraction of sites appeared not to follow
such a rule.
|
, and 
, had the expected base requirement on
both strands. Interestingly, site , which is a strong
double-stranded break site for da-IDA but not for dh-EPI (Fig. 1)
, has
5'GA3' at -2 and -1
positions at one of the two strand cuts. A guanine at position -2 has
been shown in vitro to be highly preferred by anthracycline
analogues with modifications at the 3' sugar position (3, 11)
.
Analogue-specific DNA Cleavage Sites with Purified Enzymes.
We then asked the question of whether the observed in vivo
sites were also cleaved by purified top2. DNA breaks were examined with
the Drosophila enzyme in a cloned intergenic H2A-H2B
fragment (Fig. 4
). Although the purified protein mediated strong cleavage with VM-26, it
was weakly sensitive to both the tested analogues, and four sites only
were detected in the whole region. Two of them (site n and ) were
stimulated in vivo as well (Fig. 2)
. Interestingly, da-IDA
was more specific than dh-EPI because it stimulated only site n,
whereas dh-EPI was active also at sites ß, , and (Fig. 4)
.
da-IDA was more selective than dh-EPI in an unrelated DNA fragment as
well (not shown), showing that analogue specificity was not restricted
to a particular substrate. Thus, drug interactions with
Drosophila top2 were analogue specific, supporting that they
may occur in vivo as well.
|
). da-IDA was again more selective (and potent) than dh-EPI. As expected
(3)
, all of the observed cuts had an adenine at -1
positions (Fig. 3)
, and 6 of 16 in vitro cuts were also
detected in nuclear chromatin (Figs. 3 and 5)
. The absence of sites
in vivo is explained by chromatin proteins that restrict DNA
accessibility to top2 (1, 3, 9, 21)
. Surprisingly, the
site selectivity of human isozymes was in part, but significantly
different, most notably with da-IDA (Fig. 5)
. The highest cleavage
level was always found at site n with top2. However, with top2ß
0.5 µM of dh-EPI stimulated higher cleavage
levels at site than n, and da-IDA at 0.5 µM
stimulated comparable levels at sites , , and site n (Fig. 5)
.
Differences between the two human proteins were more striking with the
more selective analogue, da-IDA. This compound exclusively stimulated
cleavage at site n with top2, whereas sites n, , 
,
, and e were detected with top2ß (Fig. 5)
. Thus, da-IDA had a
preference for top2ß at a number of sites.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Thus, DNA site selection in vivo, together with
isoform selectivity of some anthracyclines, may affect drug
cytotoxicity in human cancer cells. Strand breaks were mapped at nucleotide levels in vivo by genomic primer extension with a thermostable DNA polymerase, and the observed cuts were largely promoted by the endogenous top2 based on the following considerations: (a) nucleotide-selective stops of primer extensions were dependent upon drug treatments of cells and were consistently observed in independent experiments (see also Refs. 9 and 10 ); (b) extension of different but adjacent primers was stopped at identical nucleotide positions when using the same genomic DNA as template; (c) major in vivo break sites were also in vitro sites of top2-dependent DNA cleavage; and (d) local base preferences of poisons, as established in vitro (3) , were largely maintained in nuclear chromatin of Kc cells for anthracyclines (this work) and for VM-26 and clerocidin (9, 10, 17) .
Here, we have reported 47 in vivo sites of DNA cleavage stimulated by dh-EPI and 9 sites stimulated by da-IDA, 3 of which were specific for this compound. Many of the dh-EPI sites (55%) had an adenine at -1 positions; however, a consistent number of sites (45%) did not. Interestingly, many [-1]noA sites (30%) were very close (1–2 bp) to [-1]A sites. The observation of site clusters in chromatin may indicate that the enzyme can adjust the precise cut site of one or two nucleotides. In fact, if drug stabilization of enzyme-DNA complexes at one site causes a local enrichment of enzyme molecules, this may lead to an increased top2 cleavage activity at nearby sites as well. Modulation of cleavage levels by close sites was indeed observed in SV40 DNA with murine top2 (22) . Alternatively, the addition by Taq polymerase of two, instead of only one, untemplated nucleotides at 3' termini of a fraction of extended strands might result in extra bands in the gel (9, 18) .
However, 15% of [-1]noA sites do not apparently follow the in vitro rule and were not close to other sites, showing that anthracyclines can apparently stimulate in vivo DNA cleavage with a different sequence selectivity. The origin of these [-1]noA sites remains unclear, and some explanations can be proposed: (a) the lack of drug-preferred bases would result in a drug receptor for which the drug has a lower affinity (3) . Then, it might be possible that strong chromatin-enzyme interactions may determine the formation of ternary DNA-enzyme-poison complexes in these cases; (b) recently, we showed that anthracyclines can poison human top1 at relatively high concentrations (20) . Because these [-1]noA sites are apparently single-stranded cleavages, we cannot rule out the possibility that these breaks are generated by top1; (c) anthracyclines were shown to form a stable covalent adduct with DNA in the presence of formaldehyde or glyceraldehyde (23) . We do not know whether the adduct may occur in cultured cells at 10 µM (our experimental condition); however, it might lead to topoisomerase-independent strand breaks that could be detected by our assay.
Our results parallel previous observations with mammalian top1. Cleavage sites stimulated by camptothecin, a top1 poison, have been reported in the coding strand of the human 18S rRNA gene (24) . Four of seven sites matched the poison-preferred nucleotides (5'TG) at the cut site (25) , and the authors also showed that in vivo and in vitro sites were generally consistent with each other, although relative intensities were somewhat different (24) . An earlier study (26) investigated at nucleotide levels the in vivo sites stimulated by camptothecin in the SV40 genome during replication. Again, the large majority of the cuts had the drug-preferred nucleotides at the cut site. Thus, although some in vivo breaks cannot be predicted by in vitro-established, site-selective rules, altogether the results show that poisons of top1 and top2 mainly maintain in vivo the sequence selectivity shown in vitro (3, 9, 10, 24, 26) .
How the differences in base specificity of distinct poisons may project on different lesion lethality is a matter of further investigation. Nevertheless, it may affect the site selection of cleavage genome wide (9, 10) . This is best seen in the case of VM-26 and dh-EPI, two in vivo equally potent poisons. dh-EPI does not stimulate cleavage in the satellite III DNA, whereas VM-26 is weakly active in the histone gene region studied in this report (9, 10) . Site selection in nuclear chromatin is likely the result of a combination of several factors, including chromatin structure, enzyme site selectivity as well as drug sequence specificity (10) .
A second major finding of the present study is the isoform specificity
of da-IDA that shows a site-dependent preference for top2ß (Fig. 5)
.
This may add a further level of specificity to poison action in
mammalian cells; an altered balance of the two isozymes in tumor cells
may influence cleavage sites and cytotoxicity. Recently, it was
reported that a specific point mutation in yeast top2
(Ser740) to Trp confers resistance to quinolone
and hypersensitivity to etoposide, markedly altering the enzyme
sequence selectivity with and without poisons (27, 28)
.
The mutation may change enzyme-DNA interactions, thus influencing the
poison binding site at the protein-DNA interface (28)
. At
the equivalent positions, a serine residue
(Ser763) is present in top2, whereas an
alanine residue (Ala784) is present in top2ß.
It must be noted that Ser740 to Ala mutation has
been shown to confer a low degree of quinolone resistance to yeast top2
(27)
. Therefore, we hypothesize that the amino acid
difference, Ser763 versus
Ala784, is responsible for the different
site-dependent sensitivity to da-IDA of human top2 forms.
Interestingly, da-IDA-stimulated patterns with Drosophila
top2 were identical to those with top2, consistent with the presence
of a Ser in the insect enzyme at the equivalent position of the yeast
Ser740. The observations thus suggest that
anthracycline interactions may be subtly altered in the ternary complex
with top2ß.
With the filter elution technique, we observed that the two analogues were equipotent in stimulating double-strand breaks in human HL60 cells genome wide (6) . Nevertheless, different cleavage:cytotoxicity ratios were found for the two analogues; in particular, da-IDA-stimulated breaks were more lethal than those of dh-EPI (6) . Because the more selective analogue, da-IDA, generally stimulate cleavage at a subset of dh-EPI sites, sites stimulated by both analogues may be more lethal than those stimulated by dh-EPI only. Diverse antitumor drugs, including nitrogen mustards, alkylating compounds, cisplatin, and topoisomerase poisons, kill cells by producing sequence-selective DNA lesions (29–33) . Moreover, new pyrrole-imidazole polyamides have been designed that can specifically bind to genomic sites interfering with the biological activity of transcription factors (34, 35) . Among clinically used agents, mitomycin C is a highly specific alkylating agent of CpG sites (36) , and drug cross-linking activity has been shown to be greatly affected by local DNA structure (37) . This might result in heterogeneous genomic localization of the damage with consequences on the drug biological activity. Specific interactions of cisplatin-DNA adducts with nuclear proteins can also occur that might influence drug activity (38, 39) . A specific interaction of cisplatin-DNA adducts with a testis-specific protein (SRY) has been proposed to explain the high efficacy of cisplatin in the treatment of testicular tumors (29, 40) . Thus, the sequence selectivity of several nucleic acid-targeted drugs or their specific molecular interactions can be important determinants of the biological activity of clinically useful or promising compounds.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by grants from
Associazione Italiana per la Ricerca sul Cancro, Milan, and Ministero
dell’Università e della Ricerca Scientifica e Tecnologica, Rome,
Italy. 
2 Present address: Menarini Ricerche, via Speri
10, 00040 Pomezia, Italy.
3 To whom requests for reprints should be
addressed, at G. Moruzzi Department of Biochemistry, Bologna
University, via Irnerio 48, 40126 Bologna, Italy. Phone:
39-051-2094282. Fax: 39-051-2094283. E-mail: capranico{at}biocfarm.unibo.it
4 The abbreviations used are: top2 or top1, DNA
topoisomerase II or I; dh-EPI,
4-demethoxy-3'-deamino-3'-hydroxy-4'-epi-doxorubicin;
da-IDA,
4-demethoxy-3'-deamino-4'-deoxy-4'-epi-amino-daunorubicin;
VM-26, teniposide; SAR, scaffold-associated region.
Received 1/ 3/00. Accepted 5/11/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in yeast. Cancer Res, 53: 3591-3596, 1993.-dependent DNA cleavage and yeast cell killing by anthracycline analogues. Cancer Res, 58: 1886-1892, 1998. (p170) and IIß (p180). Mol. Pharmacol, 50: 1463-1471, 1996.[Abstract]
I1 determines its sequence specificity for DNA cleavage. Proc. Natl. Acad. Sci. USA, 88: 7464-7468, 1991.
B and DNA by a designed DNA-binding ligand. J. Biol. Chem, 274: 12765-12773, 1999.This article has been cited by other articles:
|
|
 |
|
  B.-G. Ju, V. V. Lunyak, V. Perissi, I. Garcia-Bassets, D. W. Rose, C. K. Glass, and M. G. Rosenfeld A topoisomerase IIbeta-mediated dsDNA break required for regulated transcription. Science, June 23, 2006; 312(5781): 1798 - 1802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Lemke, M. Wojciechowski, W. Laine, C. Bailly, P. Colson, M. Baginski, A. K. Larsen, and A. Skladanowski Induction of unique structural changes in guanine-rich DNA regions by the triazoloacridone C-1305, a topoisomerase II inhibitor with antitumor activities Nucleic Acids Res., October 27, 2005; 33(18): 6034 - 6047. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Minotti, P. Menna, E. Salvatorelli, G. Cairo, and L. Gianni Anthracyclines: Molecular Advances and Pharmacologic Developments in Antitumor Activity and Cardiotoxicity Pharmacol. Rev., June 1, 2004; 56(2): 185 - 229. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
