[CANCER RESEARCH 5.1, 5908-5914.
December 15, 1993]
Interaction between Replication Forks and Topoisomerase I-DNA Cleavable
Complexes: Studies in a Cell-free SV40 DNA Replication System1
Yeou-Ping Tsao, Alicia Russo, Gill Nyamuswa, Robert Silber, and Leroy F. Liu2
Department of Pharmacology, University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School, Piscataway, New Jersey 08854 [Y-P. T., L. F. L.];
Department of Molecular Biology and Genetics, Johns Hopkins School of Medicine, Baltimore, Maryland 21205 [A. R.¡; and Department of Medicine, New York University
School of Medicine, New York, New York 10016 /G. N., R. S.J
ABSTRACT
The extreme S-phase-specific
cytotoxicity
of camptothecin
has been
shown to involve active DNA replication. To investigate the role of DNA
replication in camptothecin cytotoxicity, we have studied the interaction
between the DNA replication machinery and the topoisomerase I-camptothecin-DNA ternary cleavable complex in a cell-free SV40 DNA repli
cation system. The formation of topoisomerase I-camptothecin-DNAcleavable
complexes
on the replication
template efficiently
and
irreversibly inhibited DNA replication. Two aberrant forms of replication
products were produced whose abundance varied with the concentrations
of exogenously added topoisomerase I and camptothecin. At low concen
trations of topoisomerase I and camptothecin, the major aberrant DNA
replication product was close-to-unit-length-linear
DNA, while at higher
concentrations the predominant product was close-to-dimer-size-linear
DNA. Analysis of these aberrant replication products has suggested a
"collision" model in which the interaction between an advancing replica
tion fork and a topoisomerase
I-camptothecin-DNA-cleavable
complex
by camptothecin has been shown to be mediated through inhibition
p34cdc2/cyclin B activation: both tyrosine dephosphorylation of
p34cdc2 and the activation of histone HI protein kinase are delayed or
abolished upon acute camptothecin treatment (15). In order to under
stand the molecular mechanism of DNA synthesis-dependent cellular
responses of camptothecin, we have studied the interaction between
replication forks and topoisomerase I-camptothecin-DNA-cleavable
complexes in a cell-free SV40 DNA replication system. Our results
suggest that the interaction between advancing replication forks and
topoisomerase I-cleavable complexes results in irreversible arrest of
replication forks, breakage of the replication forks, and the conversion
of the reversible cleavable complex into a topoisomerase I-linked
DNA break. One or several of these events could be the trigger for cell
death and G2 arrest.
MATERIALS
results in irreversible arrest of the replication fork and the formation of a
double-strand DNA break at the fork. Concomitant with fork arrest and
fork breakage, the reversible cleavable complex was converted into a
topoisomerase I-linked DNA break. We propose that one or several of
these events triggers S-phase-specific cell killing and G2-phase cell cycle
arrest.
INTRODUCTION
Camptothecin inhibits eukaryotic DNA topoisomerase I by trapping
a reversible enzyme-inhibitor-DNA ternary complex, termed the cleav
able complex (1). Like topoisomerase II-cleavable complexes, topoi
somerase I-cleavable complexes are a unique form of cellular DNA
damage in that they are highly reversible (1-3). Because of this unique
property of topoisomerase-mediated DNA damage, we have been able
to investigate the cellular machineries which interact with these
transient "damaged" DNA sites to transform them into permanent
DNA lesions.
In addition to reversible fragmentation of chromosomal DNA and
inhibition of DNA and RNA synthesis (reviewed in Ref. 4), campto
thecin induces a number of cellular responses such as S-phase-specific
cytotoxicity (5, 6), G2-phase cell cycle arrest (7), cellular differentia
tion (8, 9) and sister-chromatid exchange (7). Most, if not all, of these
responses are probably due to the formation of reversible topoisom
erase I-camptothecin-DNA
ternary cleavable complexes in cells
treated with camptothecin (1, 2, 10, 11). At least two of these cellular
responses to camptothecin, the S-phase-specific cytotoxicity and the
G2 arrest, have been shown to require active DNA synthesis (12, 13).
Simultaneous inhibition of DNA synthesis with replication inhibitors
during acute camptothecin treatment has been shown to abolish both
S-phase cytotoxicity and G2 arrest (14, 15). G2-phase arrest induced
Received 8/5/93; accepted 10/11/93.
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.
1This work was supported by NIH Grants CA39962 and CA40884. L. F. L. is the
recipient of the George Hitchings Award from Burrough Wellcome Co.
2 To whom requests for reprints should be addressed, at Department of Pharmacology,
UMDNJ-Robert Wood Johnson Medical School, 675 Hoes Lane, Piscataway, NJ 08854.
AND METHODS
Materials. Plasmid pUC.HNO(SV40 ori + ) contains the SV40 origin of
DNA replication including the T-antigen-binding sites and an A+T-rich re
gion (16). Calf thymus topoisomerase I, SV40 large T-antigen, and HeLa
cytosolic extract were obtained as described previously (17, 18). Camptothe
cin sodium salt and camptothecin lactone were obtained from the Drug Syn
thesis and Chemistry Branch, Division of Cancer Treatment, National Cancer
Institute. Camptothecin was dissolved in dimethyl sulfoxide and stored in
aliquots at -20°C.
DNA Replication Assay in a Cell-free System. Replication of SV40 ori
gin-containing plasmid DNAs in a HeLa cytosolic extract supplemented with
purified SV40 T-antigen was performed as described previously (18). Each
reaction (25 ml) containing 10 ml extract, 0.5 /^g SV40 T-antigen, and 50 ng
plasmid DNA in a standard reaction mixture (40 mM 4-(2-hydroxyethyl)-lpiperazineethanesulfonic
acid, pH 7.5-8 mm MgCl2-0.5 mm dithiothreitol-100
JAMeach of dATP, dTTP, and dGTP-25 JU.MdCTP-40 /IM phosphocreatine-1
mg/ml creatine phosphokinase-0.025%
Nonidet P-40-3 mM ATP-25 JIM [a32P]dCTP with a specific activity of 12 X IO3 dpm/pmol) was incubated at
37°Cfor 2 h. Reactions were terminated by the addition of sodium dodecyl
sulfate (final concentration, 1%) and proteinase K (final concentration, 1
mg/ml) and incubated at 37°Cfor an additional 2 h. DNA was then precipitated
with ethanol and dissolved in 25 ml of loading buffer with bromophenol blue
and sucrose. Electrophoresis in 1% agarose gel was carried out in Tris-borateEDTA buffer (89 mM Tris-borate, pH 8.3-2 mM EDTA). Gels were dried, and
autoradiography was performed.
Two-Dimensional Gel Electrophoresis with Differential Voltage Gradi
ents. Two-dimensional gel electrophoresis in 1.2% agarose gel was carried out
in 0.5 x tris-phosphate-EDTA buffer (19). The first-dimensional gel electro
phoresis was carried out at 5 V/cm for 4 h. The gel was then turned 90 degrees,
and the second-dimensional gel electrophoresis was at 1.2 V/cm for 15 h.
Mapping of Double-Strand Breaks on Aberrant Replication Products.
The slower migrating aberrant replication product (X molecules) generated in
the presence of high concentrations of camptothecin and topoisomerase I was
gel isolated and digested with either EcoRI or Hindlll. The digested DNA was
analyzed by 1% agarose gel in Tris-borate-EDTA electrophoresis buffer.
Mapping of Topoisomerase I Cleavage Sites in Vitro. pUC.HNO (ori + )
DNA was digested with the restriction enzyme Xbal. Linearized DNA mol
ecules were then labeled at either their 3' ends using the Klenow polymerase
or their 5' ends (dephosphorylated with calf intestine phosphatase) using T4
polynucleotide kinase. Labeled DNA molecules were made uniquely end la
beled by digesting with either Hindlll or £coRI to remove a short fragment
from either the origin-proximal side or the origin-distal side (see Fig. 3B).
5908
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
REPLICATION
FORKS
AND TOPOISOMERASE
Topoisomerase I cleavage hot spots on both strands were mapped using these
two uniquely end-labeled DNA templates. Cleavage mapping was done using
HNO DNA was incubated with calf thymus DNA topoisomerase I and camp
tothecin at 37°Cfor 30 min. The reactions were terminated by the addition of
cules (designated X) and the form III replication product were quan
tified by densitometric analysis.
The ratio of "smear X" to form III increased with increasing camp
sodium dodecyl sulfate (final concentration, 1%) and proteinase K (final con
centration, 0.2 mg/ml), followed by incubation at 37°Cfor an additional 2 h.
followed by
RESULTS
Topoisomerase I-Camptothecin DNA-cleavable Complexes In
hibit SV40 DNA Replication and Produce Two Forms of Aberrant
Replication Products. To produce high levels of cleavable com
plexes on replicating DNA templates, excess calf thymus DNA topoi
somerase I was added to the HeLa cytosolic replication extract
(supplemented with SV40 T-antigen) in the presence of increasing
concentrations of camptothecin (Fig. 1). In the absence of camptothe
cin, replication of pUC.HNO DNA in this topoisomerase [-supple
mented extract resulted in the formation of monomeric closed, circular
daughter molecules with various degrees of supercoiling (Fig. L4 ). As
discussed previously, endogenous topoisomerase activities in this ex
tract are sufficient to support replication (17). Excess topoisomerase I
does not alter replication efficiency but can increase formation of
monomeric daughter molecules and reduce formation of the high
molecular weight replication product (21). In the presence of increas
ing concentrations of camptothecin, DNA synthesis was progressively
inhibited as evidenced by an overall decrease in total band intensities
(Fig. 1, D).
Two groups of aberrant replication products were observed in the
presence of camptothecin. At low concentrations of camptothecin, a
group of aberrant replication products migrating as a slightly diffuse
band near the position expected for linearized pUC.HNO DNA (form
III) was observed (Fig. 1, lanes B-E). Previous studies have shown
that this aberrant replication product has a size close to unit length and
is covalently linked to topoisomerase I (17). Only replicating DNA
template was converted to this linear form (17). Furthermore, the
covalently linked topoisomerase I is not in the form of a reversible
cleavable complex but, instead, is irreversibly linked to DNA (17).
Fig. 1. Topoisomerase I-cleavable complexes ar
rest DNA replication and produce aberrant replica
tion products in a cell-free SV40 DNA replication
system. Left, pUC.HNO DNA was replicated in a
HeLa cytosolic extract supplemented with SV4Ü
T-antigen and purified calf thymus DNA topoisom
erase I (20 ng/reaction). The replication products
were analyzed by neutral agarose gel electrophore
sis. Lanes A-E, 0, 8, 40, 200, and 1000 JIMsodium
camptothecin, respectively. X and HI, aberrant rep
lication products. The aberrant replication product
(III) has been previously identified as close-to-unitlength linear DNA which is slightly heterogeneous
in size. The aberrant replication product (X) repre
sents a heterogeneous population of linear replica
tion products which are longer-than-unit-Iength but
shorter-than-dimer-size-linear
pUC.HNO
DNA
(see text). The positions of nicked (form II) and
supercoiled pUC.HNO DNA (form I) are marked.
Right, Quantitation of various replication products
was done by densitometric tracing of the autora
diogram. The Y axis is in arbitrary units. D, sum of
all replication products; A and O, amounts of ab
errant replication products X and III, respectively;
•,X/1II ratio.
COMPLLXKS
At higher concentrations of camptothecin, a second group of aber
rant replication products was observed, which migrated as a smear
(marked X) with an upper limit at about dimer linear length (Fig. 1,
lanes B-E). The amounts of this heterogenous group of DNA mole
20 ng of purified calf thymus topoisomerase I and 0.5 /AMcamptothecin in a
standard topoisomerase I reaction (20). Briefly, uniquely end-labeled pUC.
DNA samples were analyzed by alkaline gel electrophoresis,
autoradiography (19).
CLEAVABLE
tothecin concentration (Fig. 1, •).Furthermore, as the abundance of
form III decreased, the amount of form X increased. Like the form III
aberrant replication product, X molecules were covalently and irre
versibly linked to topoisomerase I as indicated by their preferential
extraction into the phenol phase (Fig. 5, lanes A and B; 17).
Aberrant Replication Products in the Presence of Higher
Camptothecin Concentrations Are Linear Molecules with Sizes
Ranging from Unit Length to Twice the Unit Length. We observed
empirically that, depending on the voltage gradient, the electrophoretic mobilities of the X molecules changed significantly relative to
those of the circular forms of DNA. To test whether X molecules are
linear DNA, replication products were analyzed by two-dimensional
gel electrophoresis with differential voltage gradients in the two di
mensions. Under these conditions, linear and circular DNA mol
ecules migrated very differently. By comparison with linear and cir
cular DNA markers, X molecules migrated at positions expected for
linear DNA molecules (Fig. 2, bracketed area with the dashed center
line delineates the location of linear markers) but not circular mol
ecules (Fig. 2, bracketed area without the dashed center line delin
eates the location of circular DNAs). This result suggests that X
molecules are linear molecules. Compared to the linear DNA mark
ers, the upper limit of the size of the X molecules is dimeric linear
pUC.HNO DNA.
X Molecules Are Generated by Double-Strand DNA Cleavage at
or near Topoisomerase I-cleavable Complex Sites. To map the ends
of the X molecules, gel-isolated X molecules were digested with
different restriction enzymes that cut near the SV40 replication origin
on the plasmid DNA. A number of discrete fragments were produced,
suggesting that the X molecules have specific ends (Fig. 3/4, lanes
B-E). The major end was localized to a site about 2 kilobases from the
EcoRl site (Fig. 3A, lane D). The unique position of this site was
determined by analyzing the cleavage pattern produced by Hindlll
digestion (Fig. 3A, lane E). Hindlll digestion produced a 1.8-kilobase
A B C D E
l l l l l
600
-1.5
(D
hO.5
O
i
I
il
8
40
200
1000
camptothecin
5909
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
x
REPLICATION
FORKS
AND TOPOISOMERASE
CLEAVABLE
COMPLEXES
B
Fig. 2. The aberrant replication products are lin
ear DNA molecules. The aberrant replication prod
ucts were identified as linear DNA molecules by a
two-dimensional gel electrophoresis method (see
"Materials and Methods"). Top left, top right, and
bottom left, samples prepared from replication re
actions containing 0, 0.4, and 50 ¿IM
camptothecin,
respectively; bottom right, linear DNA markers;
32P-labeled Hind\\\ digested À-DNA.The four pan
els in A were duplicated (B) and are marked by one
or two rectangles. Larger rectangle, electrophoretic
mobilities of circular DNA molecules; smaller rect
angle (with a dotted line in the center), electropho
retic mobilities of the linear DNA molecules.
A.
-ii •M
t
? Ç
9 f
M.W.
(Kb)
X molecule
lllll-
Hind III digest
EcoRIi
Fig. 3. Replication-specific linearization occurs at specific sites
which correspond to topoisomerase 1cleavage hot spots. A, mapping
of camptothecin-induced double-strand breaks on replication prod
uct X (lanes A—E).Lane A, gel-isolated replication product X\ lane
B, X digested with EcoRI; lane C, X digested with HindlU; lanes D
and £,the same as lanes B and C, respectively, except with longer
exposure times during autoradiography. B, mapping of topoisomer
ase I cleavage hot spots using purified topoisomerase I and camp
tothecin (see "Materials and Methods"). Lane F, A'fcal-linearized
pUC.HNO DNA was end labeled with Klenow polymcrase and
[a-3:P]dATP and then digested with EcoRI to remove one small
labeled end. This uniquely 3'-end-labeled DNA (labeled at the ori
gin-proximal end) was reacted with HeLa DNA topoisomerase 1 (20
ng) and camptothecin (0.5 (IM). Lane G, ATwI-linearized pUC.HNO
DNA was end labeled with Klenow polymerase and [a-<2P]dATP
and then digested with Hindlll to remove the other small labeled
end. This uniquely 3'-end-labeled DNA (labeled at the origin-distal
— 0.6
end) was reacted with HeLa DNA topoisomerase I and camptothe
cin. Lane H, Jffcal-linearired pUC.HNO DNA was 5'-end labeled
with T4 polynucleotide kinase and [y-32P]ATP and then digested
with EcoRI to remove one small labeled end. This uniquely 5'-endlabeled DNA (labeled at the origin-proximal end) was reacted with
HeLa topoisomerase I and camptothecin. Lane I, .Vial-linearized
DNA was 5'-end labeled with T4 kinase and [-y-^PJATP and then
B.
digested with Hind\\\ to remove one labeled end. This uniquely
5'-end-labeled DNA (labeled at origin-distal end) was then reacted
nl W (Kb)
pUC.HNQ DNA
Xt>tl
Ori
Hind !
EcoRI
—I3'
19 kb
with HeLa DNA topoisomerase I and camptothecin. Samples for
lanes A-E were electrophoresed in Tris-borate-EDTA buffer (pH
8.3). Samples for lanes F—lwere electrophoresed under a denaturing
condition (30 mw NaOH and 3 mM EDTA). *, major end of the X
molecule [2.0 kilobases (kb) from the EcoRI site at the origin-distal
end] and a major hot spot for a topoisomerase I-mediated DNA
cleavage site (1.9 kilobases from the Xba\ site on the origin-distal
end) on pUC.HNO DNA. Righi, schematic diagram of the strategy
used for mapping double-strand breaks on X produced in replication
extracts (top) and single-strand breaks produced by purified HeLa
DNA topoisomerase I and camptolhecin (bottom).
0.9kb
fragment as the major band (Fig. 3,4, lane E), consistent with the 0.2
kilobases separation between EcoRI and HindlU sites on pUC.HNO.
The major end of the X molecule is, therefore, located about 2.0
kilobases from the EcoRI site distal to the SV40 replication origin.
To test whether the major end of the X molecules corresponds to a
major site the of camptothecin-induced
topoisomerase I-cleavable
complex, sites of camptothecin-induced cleavage were mapped on
naked pUC.HNO DNA using purified topoisomerase I (Fig. 35).
5910
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
REPLICATION
FORKS
AND TOPOISOMERASE
Alkaline gel electrophoresis was performed to map single-strand
breaks produced by topoisomerase I-camptothecin-DNA-cleavable
complexes (see "Materials and Methods"). A major single-strand
CLEAVABLE
COMPLEXES
also be generated. However, in this case, the two covalently linked
topoisomerase I molecules would be situated at each end of the linear
DNA molecules (labeled Y). (c) If fork breakage occurs randomly at
either the leading-strand side or the lagging-strand side of the two
forks, a circular DNA and a shorter-than-unit-length-linear
DNA
would be generated in addition to the products predicted in (a) and
(b). Based on our results, (c) seems least likely because the population
of nicked DNA (form II) was less than the population of X molecules
(see Fig. 1, lanes D and E, and Ref. 22). In addition, the majority of
the nicked DNA (form II) was in the form of reversible cleavable
complexes and could be converted to closed circular DNA when
heated to 65°C(22).
DNA cleavage site was located about 2 kilobases from the SV40
replication origin (1.9 kilobases from the Xbal site) on the (-) strand
(Fig. 3ß).This single-strand break is distal to the SV40 replication
origin. No major cleavage hot spot was found to be located about 2
kilobases from the origin on the (+) strand. This result suggests that
the major end of the X molecules corresponds to a major cleavable
complex site on the DNA template.
Possible Models for Fork Interaction with Cleavable Com
plexes. To explain our results, we consider the following models. At
low camptothecin concentrations (low levels of cleavable complexes),
only one of the two replication forks would encounter the cleavable
complex on the template (Fig. 4A). This interaction between the
replication fork and the topoisomerase I-camptothecin-DNA-cleav
able complex leads to replication fork arrest, fork breakage at the site
of the cleavable complex, and the formation of a protein-linked break.
The other unimpeded replication fork would continue replication and
generate a topoisomerase I-linked linear DNA molecule with a size
close to unit length and a completely replicated monomeric covalently
closed DNA circle. At higher camptothecin concentrations (high lev
els of cleavable complexes), when there is a higher probability of each
fork encountering a cleavable complex in the correct orientation, both
replication forks would encounter the cleavable complexes.
There are three possible outcomes that can be considered to explain
how the larger aberrant replication products (X molecules) may be
formed (Fig. 4ß).(a) If fork breakage occurs only at the leadingstrand side of both forks, greater-than-unit-length
(but less-thandimer-size) linear molecules would be generated. The two covalently
linked topoisomerase I molecules would be situated in the center part
of the molecule between the two origins (labeled X). (b) If fork
breakage occurs only at the lagging-strand side of both forks, greaterthan-unit-length (but less-than-dimer-size) linear molecules would
Replication Fork Breakage Occurs Preferentially at the Lead
ing-Strand Side of the Replication Fork. To distinguish between
possibilities (a) and (b), we designed a phenol extraction procedure
(Fig. 5). Depending on whether the topoisomerase I molecules are
covalently bound at the termini of the X molecules (Fig. 5B) or within
the sequence between the two SV40 replication origins (Fig. 5^4),
EcoRl digestion followed by phenol extraction would be expected to
produce different products in the aqueous and phenol phases.
Without EcoRl digestion, X molecules and a majority of the aber
rant replication products which migrated at the form III position were
removed from the aqueous phase by phenol extraction because of their
covalent association with topoisomerase I (Fig. 5, lanes A—C).This
covalent association could not be reversed by the normal procedures
(e.g., large dilution of the reaction mixture with drug-free buffer,
challenge with excess DNA or brief heat treatment at 65°C)(2, 14, 22)
used to reverse cleavable complexes (data not shown). All form I
molecules and part of the form II molecules were not affected by
phenol extraction and remained in the aqueous phase.
After EcoRl digestion and phenol extraction, the shorter-than-unitlength DNA was recovered in the aqueous phase but not the phenol
phase (Fig. 5, lane E), consistent with prediction (a). Furthermore,
unit-length-linear DNA was recovered in the phenol phase (Fig. 5,
A: SINGLE COLLISION EVENT:
on
Fig. 4. Possible outcomes due to interaction between advancing
replication forks and camptothecin-trapped, topoisomerase I-DNA
cleavable complexes. This schematic demonstrates several possible
outcomes due to interaction between replication forks and topoisom
erase I cleavable complexes. At lower levels of cleavable complexes
(due to lower camptothecin concentrations), only one of the two
replication forks encounters the cleavable complex (single collision).
Such an encounter leads to irreversible arrest of the replication fork,
breakage at the replication fork, and the conversion of the reversible
cleavable complex into the cleaved complex. The other replication
fork continues unimpeded and eventually produces a completely
replicated circular daughter molecule and a linearized replication
product with topoisomerase I covalently linked to one end. At higher
concentrations of cleavable complexes (due to higher camptothecin
concentrations), both replication forks will encounter the cleavable
complexes in the orientation that blocks fork progression. Depending
on the polarity of fork breakage (breakage at the leading- or laggingstrand sides), three different outcomes are possible. If fork breakage
occurs on the strand complementary to the leading strand of DNA
synthesis (case B, a), one expects the replication product to be
greater-than-unit-length
but shorter-than-dimer-size-linear
DNA
with two topoisomerase I molecules covalently bound to the DNA at
internal sites (labeled X). If fork breakage occurs on the strand
complementary to the lagging strand of DNA synthesis (case B, b),
one expects similar size replication products except that topoisom
erase I molecules are covalently bound to each end of the linear
product (labeled Y). If fork breakage occurs randomly on either side
of the fork (case B, c), one expects the replication products to be
equal amounts of X, Y, and two other new replication products
consisting of a topoisomerase I-linked nicked circle and a shorterthan-unit-length linear DNA that is also linked to topoisomerase I.
on
B: TWO COLLISION EVENTS.
•
: Doubl«»Inn* brMkigt
feeding ilrind iM«.
unii
length
Untar
occur»«Ittwj
greeur
Him
unit length
Iliuir
unit
Unfit
(X)
grt»Ur
thin
Itnglh
(V)
c:
Doubl«Urinal brtek«gtoccurt
or toggingetrendIK*.
»rtdomjy«teither Ih«
»m«lt»r
lh«n unii lenglk
llneer
5911
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
REPLICATION
Fig. 5. Fork breakage occurs when topoisomerase I is trapped on the strand complementary to the
leading strand of DNA synthesis. If fork breakage
is specific, depending on the polarity of fork break
age, each replication product may contain two covalently bound topoisomerase 1 molecules either
internally (A) or externally (B). A method involv
ing EcoRl digestion followed by phenol extraction
was used to distinguish between these two possi
bilities. A, schematic showing possible products
generated by fork breakage at the strand comple
mentary to the leading strand of DNA synthesis and
their fate after enzyme digestion and phenol extrac
tion. B, possible products generated by fork break
age at the strand complementary to the lagging
strand of DNA synthesis and their fate after enzyme
digestion and phenol extraction. Far right, experi
mental results. Lane A, total replication products;
lane B, DNA in the aqueous phase after phenol
extraction; lane C, DNA in the phenol phase after
phenol extraction; lanes D-F, the same as lanes
A-C except that replication products were digested
with EcoRI before phenol extraction.
FORKS
AND TOPOISOMERASE
CLEAVABLE
(A)
COMPLEXES
(B)
DNA breakage occurs
only at trie leading strand
side
ECO
Rl
ECO
Rl
DNA breakage occurs
only at the lagging strand
side
E.coRI
EcoRl
ABODE
F
lililÃ-
(unit
length)
Eco Rl
digestion
EcoRl
digestion
Phenol
extraction
Phenol
extraction
Phenol
phase
on
Ori
Aqueous
phase
(unit
1 length)
Ori
Ori
lane F), again consistent with the prediction of (a). These results,
therefore, strongly suggest that fork breakage occurred preferentially
at the leading-strand side of replication forks.
Encounter between a Replication Fork and a Topoisomerase
I-cleavable Complex Leads to Irreversible Arrest of the Replica
tion Fork. Previous studies have shown that both forms of the aber
rant replication products, the greater-than-unit-length
linear DNA
(marked X in Fig. 1) and the close-to-unit-length DNA molecules
(marked III in Fig. 1) were covalently bound to topoisomerase I.
These covalent topoisomerase I-DNA complexes have properties dif
fering from those of the cleavable complexes. While cleavable com
plexes are known to be highly reversible, the covalently bound to
poisomerase I molecules in these aberrant replication products were
not dissociated or converted to noncleavable complexes under
reversing
conditions
(e.g., 65°C, excess salts, excess
challenge
DNA, and large dilution of the reaction with drug-free buffer; data
not shown).
To test whether the formation of these irreversibly bound topoi
somerase I molecules prevented further DNA synthesis, a pulse-chase
experiment was performed. After replication reactions had proceeded
for 30 min, excess unlabeled dCTP (2 min) was added, and aliquots
were withdrawn every 30 min. As shown in Fig. 6, neither aberrant
replication product could be chased to other replication products. In
contrast, the late Cairn's replication intermediate (indicated by Rl in
Fig. 6), which apparently had not interacted with the cleavable com
plexes, was chased to other replication products (Fig. 6, lanes E-H).
These results suggest that the formation of aberrant replication prod
ucts is accompanied by irreversible arrest of the replication forks.
ABCD
! I I I
R.l-fft«
EF GH
I I I I
ff
Fig. 6. Irreversible fork arrest upon the collision between advanc
ing forks and topoisomerase I-cleavable complexes. pUC.HNO
DNA was replicated in HeLa extracts supplemented with SV40
T-antigen, purified HeLa DNA topoisomerase I (20 ng), and camptothecin (50 IJLM).Lanes A-D, replication proceeded for 30, 60, 90,
and 120, min, respectively; lanes F-H, the same as lanes A-D except
that 2 mM unlabeled dCTP was added 30 min after replication reac
tion was started. Rl, position corresponding to the late Cairn's rep
licative intermediates.
ABCD
Pulse
Chase
30' 60' 90'\2Q'
EFGH
301 301 301 30'
30' 60' 90'
5912
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
REPLICATION
Noncleavable
- Camptothecin
FORKS
AND TOPOISOMERASE
complex
of cleavable complexes since they cannot be dissociated by the normal
reversing procedures (e.g., dilution of the reaction mixture, challeng
ing with excess DNA, or brief heating to 65°C). (c) They form
Camptothecin
specifically on replicating DNA templates. Unreplicated DNA or nonorigin-containing DNA template is not converted to the linear form.
These preliminary results led to the consideration of a "fork collision
model" in which an advancing replication fork transforms the revers
B
D
fork arrest;
fork
Cell Death;
breakage
cleaved
complex
G2 Arrest
Fig. 7. A model describing orientation-specific interaction between a replication fork
and a topoisomerase I-cleavable complex. In this model, topoisomerase I is represented in
an asymmetric horseshoe shape. Camptothecin interacts with topoisomerase I forming a
ternary cleavable complex on the replicating DNA template. The free 5'-hydroxyl group
is minimally protected by topoisomerase I. The major protein-DNA contact is on the DNA
segment upstream of (he breakage site. The collision occurs when the cleavable complex
and the approaching replication fork are in such an orientation that the strand containing
the break within the cleavable complex is complementary to the leading strand of DNA
synthesis. The immediate consequence of such a collision is replication fork arrest, fork
breakage, and the conversion of the reversible cleavable complex into a cleaved complex.
One or several of these events at the arrested replication fork is presumed to trigger cell
death, cell cycle arrest, and other cellular responses.
DISCUSSION
COMPLEXES
plexes on DNA synthesis ( 17). Most strikingly, a linearized replication
product, which is slightly heterogeneous in size and is slightly shorter
than the template DNA, is prominently produced in the presence of
topoisomerase I and Camptothecin (17). This aberrant linear replica
tion product is unusual for the following reasons: (a) All of them are
covalently linked to protein (most likely the exogenously added to
poisomerase I), (b) The protein-DNA complexes are not in the form
5'
Replication
CLEAVABLE
ible topoisomerase I-camptothecin-DNA-cleavable
complex into a
form of cleaved complex which concomitantly arrests the replication
fork irreversibly.
In our present studies, we observed an additional aberrant replica
tion product in the same cell-free replication system at higher con
centrations of Camptothecin and/or topoisomerase I. This additional
aberrant replication product has properties very similar to those of the
aberrant replication product observed previously at low concentrations
of Camptothecin and/or topoisomerase I. One detectable difference is
their sizes; in contrast to the predominantly unit-length-linear mole
cules at lower Camptothecin concentrations, the aberrant replication
product at higher concentrations of Camptothecin are in the range
between unit and dimer length linear. This aberrant replication product
has a sharp cutoff at the linear dimer length. The formation of this
unusual aberrant replication product at higher concentrations of camptothecin and/or topoisomerase I is consistent with the fork collision
model in which both forks of a bidirectional replication intermediate
interact with topoisomerase I-camptothecin-DNA-cleavable
com
plexes, and fork breakage occurs at both replication forks in a nonrandom fashion; the parental strand that is complementary either to the
leading strand or to the lagging strand (but not both strands) of DNA
synthesis is broken irreversibly by the collision event. The localization
of the bound protein (topoisomerase I) at sites internal to the termini
of the X molecules supports the notion that fork breakage occurs
preferentially on the parental strand that is complementary to the
leading strand of DNA synthesis (Fig. 7).
The reason for this asymmetrical fork breakage is not clear. One
possible explanation is that the major protein-DNA contact in the
topoisomerase I-camptothecin-DNA ternary cleavable complex is on
the same side as the topoisomerase I-linked 3'-phosphoryl end, and
the strand containing the 5'-hydroxyl is minimally protected by to
poisomerase I (Fig. 7). The lack of significant protein-DNA contact on
the 5'-hydroxyl-containing strand may allow easy displacement of the
5'-hydroxyl-containing
strand from the ternary complex by the ad
The impressive antitumor activities of Camptothecin and its ana
logues against a large number of human tumors in xenografts are not
vancing replication apparatus, resulting in fork breakage at the paren
fully understood (23-25). It is known, however, that Camptothecin is tal strand that is complementary to the leading strand of DNA syn
highly selective in killing S-phase cells (6, 12, 26). The S-phasethesis. Topoisomerase I has been suggested to contact DNA
specific cytotoxicity cannot be solely explained by the formation of predominantly upstream from the topoisomerase I-linked 3'-phosphoryl end from studies of the "consensus" sequence of topoisomerase I.
cleavable complexes, since topoisomerase I is present in cells at all
phases of the cell cycle (27) and can be trapped by Camptothecin into
Most of the sequence information is contained on the DNA which is
on the same side as the topoisomerase I-linked 3'-phosphoryl end (28,
cleavable complexes (6). More recent studies have indicated that
active DNA replication is intimately involved in the S-phase-specific
29). Studies of spontaneous cleavage also suggested minimal proteinDNA contact between the 5'-hydroxyl-containing
strand and topoi
cytotoxicity of Camptothecin.
The involvement of active DNA replication in the cytotoxicity of somerase I (30, 31).
The nature of the arrested "broken fork" is not clear. The cleavable
Camptothecin has prompted our studies using a cell-free SV40 DNA
replication system (17). In our preliminary studies, SV40 DNA rep
complex is probably converted to a form of cleaved complex upon
lication was shown to be strongly inhibited only when both purified
interaction with the advancing replication fork. We have attempted to
detect the putatively free 5'-hydroxyl end without success. It is pos
topoisomerase I and Camptothecin are simultaneously present in the
sible that the 5'-hydroxyl-containing strand of the cleavable complex
replication extract, suggesting an inhibitory effect of cleavable com
5913
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
Kl l'I KAIION KìRKSANI) lOl'OISOMliRASI. (IIAVAHI.l:
is displaced by the replication apparatus but is still contained within
the replication complex and not free to diffuse in solution. It is also not
clear whether camptothecin is still reversibly bound to the cleaved
complex. Regardless of the exact nature of the cleaved complex,
replication can no longer continue. Further experiments are necessary
to establish the nature of the interaction between the cleavable com
plex and the replication apparatus and the pathways leading to its
cellular consequences.
The presence of distinct fork breakage sites is consistent with the
fork collision model. These fork breakage sites may correspond to
sites of cleavable complexes which are known to be nonrandomly
distributed. One of the major fork breakage sites has been mapped to
be at, or close to, the major site of camptothecin-induced topoisomerase I-cleavable complexes mapped on naked DNA. Our preliminary
studies have suggested that the major fork breakage site is the same as
the major cleavable complex site at the sequence level/1 Although we
are not certain whether all fork breakage sites are at the cleavable
complex sites, the frequency and intensity of these sites suggest that
they may be related.
The effect of camptothecin on SV40 DNA replication has also been
studied in cultured monkey cells (32-34). Camptothecin is highly
effective in blocking SV40 DNA replication in vivo (32). Linear SV40
DNA replication products have been observed in camptothecin-treated
cells (33, 34). Interestingly, these linear replication products are un
stable and can be chased into other forms of DNA (33). Because of the
extraction procedure (Hirt extract) used for the isolation of these
replication products in these experiments, it is not clear whether these
replication products are protein linked. It is, therefore, uncertain
whether these linear replication products observed in vivo correspond
to the linear replication products we have observed in vitro. Campto
thecin-induced breaks on replicating SV40 DNA have also been
mapped in vivo. Contrary to our findings in vitro, these breaks are
preferentially mapped to the strand that is complementary to the
lagging strand of DNA synthesis near the origin of replication (35).
The origin of this discrepancy is not clear. The fact that SV40 DNA is
in chromatin form in vivo and that topoisomerase I-camptothecinDNA-cleavable complexes are known to be enriched in the tran
scribed region may in part explain the discrepency. Another recent
report has shown that fork breakage induced by camptothecin treat
ment in vivo occurs on both the leading and lagging strand of DNA
synthesis (36). Again, the Hirt extraction procedure, which is known
to remove protcin-DN A complexes, was used for the isolation of these
replication products in these experiments. It is possible that only the
protein-free population of the linear replication products were ana
lyzed in these experiments. Whether these protein-free replication
products represent the processed form of the protein-linked replication
products or are produced by a different mechanism requires further
experimentation.
REFERENCES
1. Hsiang. Y-H.. Hertzberg. R.. Hecht. S., and Liu. L. F. Camptolhecin induces proteinlinked DNA breaks via mammalian DNA topoisomerase I. J. Biol. Chem., 260:
14873-14878, 1985.
2. Hsiang. Y-H., and Liu. L. F. Identification of mammalian DNA topoisomerase I as an
intracellular target of the anticancer drug camptothecin. Cancer Res., 48: 1722-1726,
1988.
3. Hsiang, Y-H., and Liu. L. F. Evidence for the reversibility of cellular DNA lesion
induced by mammalian topoisomerase 11 poisons. J. Biol. Chem., 264: 9713-9715,
1989.
4. Horwitz, S. B.. Corcoran. J. W., and Hahn, F. E. (eds.). Camptothecin in Antibiotics,
pp. 48-57. Berlin: Springer Verlag. 1975.
5. Gallo, R. C., Whang-Peng. J.. and Adamson, R. H. Studies on the antitumor activity,
mechanism of action, and cell cycle effects of camptothecin. J. Nati. Cancer Inst., 46:
789-795, 1971.
1Y-P. Tsao and L. F. Liu, unpublished results.
COMPLEXES
Horwitz, S. B.. and Horwitz, M. S. Effects of camptothecin on the breakage and repair
of DNA during the cell cycle. Cancer Res., 33: 2834-2836, 1973.
Tbbey, R. A. Effects of cytosine arabinosidc, daunomycin, mithramycin. azacytidine.
Adriamycin, and camptothecin on mammalian cell cycle traverse. Cancer Res., 32:
2720-2725, 1972.
Aller, P., Rius, C, Mata, F., Zorrilla, A., Cabanas, C.. Bellon. T.. and Bernabeu, C.
Camptolhecin induces differentiation and stimulates the expression of differentiationrelated genes in U-937 human promonocytic leukemia cells. Cancer Res.. 52: 12451251, 1992.
Chou. S., Kaneko, M.. Nakaya, K., and Nakamura, Y Induction of differentiation of
human and mouse myeloid leukemia cells by camptothecin. Biochem. Biophys. Res.
Commun., 166: 160-167, 1990.
10. Eng, W-K., Faucelte, L., Johnson, R. K., and Sternglanz, R. Evidence that DNA
topoisomerasc I is necessary for the cytotoxic effects of camptothecin. Mol. Pharmacol., 34: 755-760, 1989.
11. Nitiss. J., and Wang. J. C. DNA topoisomerase-targcting antitumor drugs can be
studied in yeast. Proc. Nail. Acad. Sci USA, 85: 7501-7505. 1988.
D'Arpa, P.. Beardmore. C., and Liu, L. F. Involvement of nucleic acid synthesis in cell
12.
killing mechanisms of topoisomerase poisons. Cancer Res., 50: 6919-6924. 1990.
13. Holm. C., Covey, J. M.. Kerrigan, D., and Pommier. Y. Differential requirement of
DNA replication for the cytotoxicity of DNA topoisomerase I and II inhibitors in
Chinese hamster DC3F cells. Cancer Res., 49: 6365-6368, 1989.
Zhang. H., D'Arpa, P., and Liu, L. F. A model for tumor cell killing by topoisomerase
14.
poisons. Cancer Cells, 2: 23-27, 1990.
Tsao, Y. P.. D'Arpa, P.. and Liu. L. F. The involvement of active DNA synthesis in
15.
camptothecin-induced G2 arrest: altered regulation of p34cdc2/cyclin B. Cancer Res.,
52: 1823-1829, 1992.
16. Cheng. L., Workman, J. L.. Kingston, R. E., and Kelly. T. J. Regulation of DNA
replication in vitro by the transcriptional activation domain of GAL4-VP16. Proc.
Nati. Acad. Sci. USA, 89: 589-593, 1992.
17. Hsiang, Y. H., Lihou, M. G., and Liu, L. F. Arrest of replication forks by drugstabilized topoisomerase I-DNA cleavable complexes as a mechanism of cell killing
by camptothecin. Cancer Res., 49: 5077-5082, 1989.
18. Li, J. J., and Kelly. T. J. Simian virus 40 DNA replication in vitro. Proc. Nati. Acad.
Sci. USA, 81: 6973-6977, 1984.
19. Maniatis. T.. Fritsch, E. F., and Sambrook, J. Molecular Cloning: A Laboratory
Manual. Ed. 2. Cold Spring Harbor. NY: Cold Spring Harbor Laboratory, 1989.
20. Liu, L. F., and Miller. K. G. Eukaryotic DNA topoisomerases. Two forms of type 1
DNA lopoisomerases from HeLa cell nuclei. Proc. Nati. Acad. Sci. USA, 78: 34873491, 1981.
21. Yang, L., Wold, M. S., Li, J. J., Kelly, T. J., and Liu, L. F. Roles of DNA topoisom
erases in SV40 DNA replication in vitro. Proc. Nati. Acad. Sci. USA, 84: 950-954,
1987.
22. Herlzberg, R. P., Caranfa, M. J., and Hecht, S. M. On the mechanism of topoisomerase-I inhibition by camptothecin—evidence for binding to an enzyme DNA com
plex. Biochemistry, 28: 4629-4638, 1989.
23. Giovanella. B. C., Stehlin, J. S., Wall, M. E., Wani, M. C., Nicholas, A. W., Liu, L.
F., Silber, R., and Potmesil, M. DNA topoisomerase I-largeled chemotherapy of
human colon cancer in xenografts. Science (Washington DC), 246: 1046-1048, 1989.
24. Giovanella, B. C., Hinz, H. R., Kozielski. A. J., Stehlin, J. S. J., Silber, R., and
Potmesil. M. Complete growth inhibition of human cancer xenografts in nude mice by
a treatment with 20-(S)-camptothecin. Cancer Res., 51: 3052-3055, 1991.
25. Potmesil, M., Giovanella, B. C., Liu, L. F., E.. W. M., Silber, R., Stehlin, J. S., Hsiang,
Y-H.. and Wani, M. C., Preclinical studies of DNA Topoisomerase I-targeted 9-amino
and 10,11-methylenedioxy camptothecins. In: M. P. Potmesil and K. Kohn (eds.),
DNA Topoisomerases in Cancer, pp. 299-311. New York: Oxford Press, 1991.
26. Li, L. H., Fräser,T. J., Olin, E. J., and Bhuyan, B. K. Action of camptothecin on
mammalian cells in culture. Cancer Res., 32: 2643-2650, 1972.
27. Hsiang. Y-H., Wu. H-Y., and Liu, L. F. Proliferation-dependent regulation of DNA
topoisomerase II in cultured human cells. Cancer Res., 48: 3230-3235, 1988.
28. Edwards, K. A.. Halligan. B. D.. Davis, J. L., Nevera, N. L.. and Liu. L. F. Recog
nition sites of eukaryotic DNA topoisomerase I. DNA nucleotide sequencing analyses
of topo I cleavage sites on SV40 DNA. Nucleic Acids Res.. 10: 2565-2576, 1982.
29. Been, M. D., Burgess. R. R., and Champoux, J. J. Nucleotide sequence preference at
rat liver and wheat germ type 1 DNA topoisomerase breakage sites in duplex SV40
DNA. Nucleic Acids Res., 12: 3097^(014, 1984.
30. Halligan, B. D., Davis, J. L., Edwards, K. A., and Liu, L. F. Intra- and intermolecular
strand transfer by HeLa DNA lopoisomerase I. J. Biol. Chem., 257: 3995^4000, 1982.
31. Svejslrup, J. O.. Christiansen, K., Andersen, A. H., Lund, K., and Westergaard, O.
Minimal DNA duplex requirements for topoisomerase-I-mediated cleavage in vitro, i.
Biol. Chem., 265: 12529-12535, 1990.
32. Yang, L., Liu, L. F., Li, J. J., Wold, M. S., and Kelly, T. J. The roles of DNA
topoisomerases in SV40 DNA replication. UCLA Symp. Mol. Cell. Biol.. 47: 315326, 1986.
33. Avemann, K., Knippers, R., Koller, T., and Sogo, J. M. Camptothecin, a specific
inhibitor of type I DNA topoisomerase, induces DNA breakage at replication forks.
Mol. Cell. Biol., 8: 3026-3034, 1988.
34. Champoux, J. J. Topoisomerase I is preferentially associated with isolated replicating
simian virus 40 molecules after treatment of infected cells with camptothecin. J.
Virol., 62: 3675-3683, 1988.
35. Porter, S. P., and Champoux. J. J. Mapping in vivo topoisomerase I sites on simian
virus 40 DNA: asymmetric distribution of sites on replicating molecules. Mol. Cell.
Biol., 9: 541-550, 1989.
36. Shin, C. G., and Snapka. R. M. Exposure to camptothecin breaks leading and lagging
strand simian virus 40 DNA replication forks. Biochem. Biophys. Res. Commun..
160: 135-140, 1990.
5914
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1993 American Association for Cancer Research.
Interaction between Replication Forks and Topoisomerase
I-DNA Cleavable Complexes: Studies in a Cell-free SV40 DNA
Replication System
Yeou-Ping Tsao, Alicia Russo, Gill Nyamuswa, et al.
Cancer Res 1993;53:5908-5914.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/53/24/5908
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. © 1993 American Association for Cancer Research.