 |
INTRODUCTION |
Deoxyribonucleoside analogs are among the most effective agents
for the treatment of cancer and viral diseases. Most of these compounds
exert their function by inhibiting cellular or viral DNA synthesis.
Since the naturally occurring nucleosides are in the
-D-configuration, most of the nucleoside analogs were
designed in that configuration. The discovery of
-L-2',3'-dideoxy-3'-thiocytidine (L-SddC,
3TC),1 a stereochemically
unnatural L-nucleoside analog (Fig. 1) inhibitor of HIV and
hepatitis B virus replication (2-6), defined a new category for
the design of nucleoside analogs. Among these,
-L-dioxolane-cytidine (L-OddC) (Fig. 1) is
the first to show potential anticancer activity (2, 3), and clinical
evaluation demonstrates its effectiveness against both leukemia and
solid tumors (4-7).
For activation, most nucleoside analogs require conversion to the
triphosphate metabolite by several cellular enzymes. Our previous
studies have shown that L-OddC can be phosphorylated by
deoxycytidine kinase to its monophosphate metabolite, which is
further phosphorylated by cellular kinases to its di- and triphosphate metabolites (2, 3). The latter form of L-OddC can be
incorporated into DNA by DNA polymerases 
, , 
, 
, and 
in vitro (8). Since L-OddC lacks a hydroxyl
group at the 3'-position, it causes premature termination of DNA
replication once incorporated and eventually leads to cell death. This
chain termination is probably the major mechanism of action of
L-OddC. We have also shown previously that the cytotoxicity
of this drug is directly related to the steady-state level of
L-OddC in DNA (2).
The steady-state level of L-OddC in DNA is not only
dependent on incorporation by DNA polymerases but also on the excision by DNA repair enzymes. We have previously observed that incorporated L-OddCMP could be excised from the DNA in cells (2).
However, of the several DNA exonucleases that have been studied to
date, all showed preference for D- over
L-configuration nucleoside analogs at the 3' termini of DNA
(8-15), making them unlikely candidates as the enzyme responsible for
the removal of L-OddC from DNA. In this study, we used the
nuclei from leukemic cells of patients to identify the major enzymatic
activity that removes L-OddCMP from DNA as human
apurinic/apyrimidinic (AP) DNA endonuclease (APE1), also known as HAP-1
or Ref-1 (1, 16-18).
APE1 is a 35-kDa monomeric protein that possesses multiple activities
related to DNA metabolism. APE1 belongs to the class II AP
endonucleases and is the major cellular enzyme responsible for
repairing AP sites in DNA (1, 16, 19-21). The DNA repair activities
reside in the C-terminal region (22). APE1 initiates the DNA base
excision repair by cleaving the DNA immediately adjacent to the
5' of an AP site to produce a hydroxyl group at the 3' terminus
of an unmodified nucleotide upstream of the nick and a 5'-deoxyribose
phosphate moiety downstream. The product of APE1 is further processed
by DNA polymerase to release 5'-deoxyribose phosphate with its
intrinsic lyase activity and to fill up the one-nucleotide gap with its
DNA polymerase activity (23). The nicked DNA is then sealed by DNA
ligase I or DNA ligase III/XRCC1 to complete this repair process. In
addition to the AP endonuclease activity, APE1 also possesses a 3'-5'
DNA exonuclease activity, a 3'-phosphodiesterase activity, a
3'-phosphatase activity, and a RNaseH activity (24). APE1 also contains
a number of potential phosphorylation sites for protein kinase C and
casein kinases I and II (25, 26). However, the impact of
post-translational phosphorylation on the DNA repair activity of APE1
is not clear at this time.
In addition to its DNA repair activities, APE1 plays a role in
regulating gene expression. The N-terminal domain of this protein has a
redox regulatory function (17, 27, 28). Through this domain, APE1
mediates the reductive activation of oxidized proteins such as c-Jun
and c-Fos as well as Myb, nuclear factor 
B, and members of CREB
family to facilitate their specific DNA binding ability (18, 29). A
mouse knock-out of APE1 resulted in an embryonic lethal phenotype (30,
31), further emphasizing the importance of this gene for normal
cellular functioning.
The 3'-5' DNA exonuclease activity was reported to be between 3 and 4 orders of magnitude less efficient than the AP endonuclease activity
(32) and hence was not considered biologically significant. In view of
our current study, this exonuclease activity could be much more
important than that previously suggested. Besides L-OddC
terminated oligonucleotide, we also examined the effectiveness of APE1
exonuclease activity against the antiviral deoxyoligonucleotides analogs L-SddC, L-Fd4C (33, 34), ddC (35, 36),
L-ddC (37), and the anticancer analogs such as dFdC
(38-41) and araC (42, 43) (Fig. 1). The
exonuclease activity of APE1 showed a preference for
L-configuration nucleoside analogs over the natural
D-configuration analogs. Moreover, the product of APE1
exonuclease could also be extended by polymerases, which suggested that
APE1 could be the key enzyme involved in removing
L-configuration nucleoside analogs from the 3' termini of
DNA. The potential importance of this APE1 exonuclease activity in
determining the toxicity of deoxyribonucleoside analogs is
discussed.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 1.
Structures of nucleoside analogs.
1, dCyd; 2, ddC; 3, L-ddC;
4, L-OddC; 5, araC; 6,
dFdC; 7, L-SddC; 8,
L-Fd4C.
|
|
| |
EXPERIMENTAL PROCEDURES |
Materials and Compounds--
Radiolabeled
[
32P]ATP, [32P]dATP, and
[32P]dCTP were purchased from Amersham Pharmacia
Biotech. L-SddCTP was a gift from Dr. R. F. Schinazi
(Veterans Affairs Medical Center, Decatur, GA). L-Fd4CTP
was synthesized by Vion Pharmaceuticals, Inc. (New Haven, CT). dFdCTP
was a generous gift from Dr. W. Plunkett (M. D. Anderson Cancer
Center). All other nucleoside 5'-triphosphate analogs were synthesized
in our laboratory as described previously (44). P11 cation and DE52
anion exchange resins were obtained from Whatman (Fairfield, NJ).
Phenyl-Sepharose resin was purchased from Amersham Pharmacia Biotech;
single-stranded DNA cellulose was from Sigma. T4 polynucleotide kinase
was purchased from New England Biolabs (Beverly, MA); E. coli DNA polymerase I Klenow fragment, Klenow exonuclease free
protein, deoxynucleotidyl terminal transferase, and HIV reverse
transcriptase were purchased from Amersham Pharmacia Biotech. The
recombinant APE1 protein was a generous gift from Dr. Bruce Demple
(Harvard University). f1-K12 DNA and tetrahydrofuran containing
oligonucleotide and its complementary strand were gifts from Dr. Zafer
Hatahet (University of Texas Health Center, Tyler, TX).
Oligonucleotide Substrates--
All oligonucleotides were
synthesized by Integrated DNA Technology, Inc. (Coralville, IA) and
further purified by electrophoresis on a 20% denaturing polyacrylamide
gel. To facilitate monitoring the 3'-5' DNA exonuclease activity that
removed L-OddC from the 3' termini of DNA and to avoid
possible false results introduced by contaminating 5'-phosphatase or
5'-3' DNA exonucleases in the purification process, oligonucleotide A
(Sequence 1) was 3'-labeled with [32P]dATP and
further extended with OddCMP by E. coli DNA polymerase Klenow fragment (exonuclease-free) enzyme in 50 mM
Tris-HCl, pH 7.5, 1 mM dTT, and 50 µg/ml bovine serum
albumin at 37 °C for 5 min. To monitor other exonucleases with
different substrate preferences, double-stranded oligonucleotide B
(Sequence 2) was labeled with [32P]dCTP using
E. coli DNA polymerase Klenow fragment in 50 mM
Tris-HCl, pH 7.5, 1 mM dTT, and 50 µg/ml bovine serum
albumin at 37 °C for 5 min. The products were purified using G-50
spin columns (Roche Molecular Biochemicals) to remove radioactive
nucleotides. The purity of the products was examined by a 12.5%
urea/polyacrylamide gel electrophoresis in 1× TBE buffer (89 mM Tris, 89 mM boric acid, 2.5 mM
EDTA, pH 8.3).
|
|
|
|
|
|
|
|
|
|
|
|
Since there was no 5'-phosphatase activity in the final enzyme
purification, to facilitate the characterization of APE1 exonuclease activity and substrate specificity, both oligonucleotides C and D were
5'-end-labeled with [32P]ATP using T4 polynucleotide
kinase. Following the kinase reaction, 50 µM
L-OddCTP and 1 unit of deoxynucleotidyl terminal
transferase (Amersham Pharmacia Biotech) was added to oligonucleotide C
(Sequence 3) and the mixture was incubated at 37 °C for 30 min.
Oligonucleotides C and D (Sequence 4) were then rendered
double-stranded by the addition of the 33-nucleotide complementary
strand.
The AP endonuclease activity was assayed with a synthetic
tetrahydrofuran (F) containing oligonucleotide E as listed in
Sequence 5.
Exonuclease and Endonuclease Assays--
The standard
exonuclease and endonuclease reaction (10 µl) mixtures contained 2.5 fmol of either 3'- or 5'-labeled oligonucleotides and 20 mM
Tris-HCl, pH 7.5, 2 mM MgCl2, 0.5 mM EDTA, 30 mM KCl, and 0.1 mg/ml bovine serum
albumin. 1 µl of different dilutions of enzyme was added to the
reaction mixtures for 30 min at 37 °C. The reaction was stopped by
adding 4 µl of loading solution (90% formamide, 1 mM
EDTA, 0.1% xylene cyanol, and 0.1% bromphenol blue) and
heating at 80 °C for 3 min. Samples (3 µl) were loaded onto a
12.5% polyacrylamide gel containing 8 M urea, which was electrophoresed in 1× TBE buffer. The gel was then dried under vacuum
and subjected to autoradiography and a phosphor-imaging screen
(Bio-Rad).
Purification of 3'-5' DNA Exonuclease Activity with Specificity
against L-()-OddCMP--
All purification procedures
were performed at 4 °C unless otherwise indicated. The buffers used
in purification were as follows: buffer A, 20 mM Tris-HCl,
pH 7.4, 300 mM KCl, 2 mM benzamidine, 1 mM PMSF, 1 mM DTT, 0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin A, 20% glycerol; buffer B, 20 mM
Tris-HCl, pH 7.4, 50 mM KCl, 2 mM benzamidine,
1 mM PMSF, 1 mM DTT, 0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin A, 20% glycerol; buffer C, 20 mM
Tris-HCl, pH 7.4, 600 mM KCl, 2 mM benzamidine,
1 mM PMSF, 1 mM DTT, 0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin A, 20% glycerol; buffer D, 20 mM Tris-HCl, pH 7.4, 1 M KCl, 2 mM benzamidine, 1 mM PMSF, 1 mM DTT, 0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin A, 20% glycerol; buffer E, 20 mM
Tris-HCl, pH 7.4, 2 mM benzamidine, 1 mM PMSF,
1 mM DTT, 0.5 µg/ml leupeptin, 0.5 µg/ml pepstatin A,
20% glycerol.
Crude Nuclei Extract--
The 3'-5' DNA exonuclease that removed
L-OddCMP was purified from the nuclei of acute lymphocytic
leukemia cells removed by leukophoresis from patients in blast crisis.
The procedures for purifying cellular nuclei from leukemic cells were
described (11). Frozen nuclei pellets (50 g) were thawed in 150 ml of buffer A. The nuclei debris was removed by centrifugation at
27,500 rpm for 60 min, and the supernatant was collected (fraction I)
for further purification.
DE52 Anion Exchange--
A DE52 anion exchange column was used
to remove nucleic acids from the supernatant of the crude extract. A
50-ml DE52 column was packed and equilibrated with buffer A. Fraction I
was loaded onto the column at a flow rate of 1.5 ml/min, the column was
washed with 3 bed volumes of buffer A. Both the flow-through and wash fractions were collected and pooled (fraction II).
P11 Cation Exchange--
Fraction II was dialyzed against buffer
B until it had the same conductivity as buffer B before it was loaded
onto a 50-ml P11 column preequilibrated with buffer B. The column was
washed with 3 column volumes of buffer B followed by a 500-ml linear gradient of buffer B to buffer C at a flow rate of 1 ml/min. Fractions were collected and analyzed for exonuclease activity. The active ones
were pooled (fraction III).
DE52 Anion Exchange--
Fraction III was dialyzed against
buffer B and loaded onto a buffer B-preequilibrated DE 52 column. The
column was washed with 3 column volumes of buffer B. The proteins were
eluted with a 300-ml linear gradient of buffer B to buffer C at a flow
rate of 1 ml/min. Fractions were collected and assayed for exonuclease activity. The active fractions were pooled and dialyzed against buffer
B (fraction IV).
Single-stranded DNA Cellulose Column--
A 15-ml
single-stranded DNA cellulose column was packed and preequilibrated
with buffer B. Fraction IV was loaded onto the column and washed with 3 column volumes of buffer B. The column was eluted with a 150-ml
linear gradient of buffer B to buffer C at a flow rate of 1 ml/min. Fractions were collected, analyzed, pooled, and dialyzed
against buffer D (fraction V).
Phenyl-Sepharose Column--
Fraction V was loaded onto a 2-ml
buffer D-preequilibrated phenyl-Sepharose column. The column was washed
with 3 column volumes of buffer D before the application of a 25-ml
linear gradient of buffer D to buffer E at a flow rate of 0.5 ml/min. The flow-through (fraction VI) and the active fractions eluted
from the column (fraction VII) were collected, pooled, and
dialyzed against buffer B.
Concentration on P11 Column--
Fractions VI and VII were
loaded onto two separate 1-ml P11 columns and eluted with a 10-ml
linear gradient of buffer B to buffer C. Active fractions were
collected and pooled (fractions VIII and IX, respectively).
SDS-PAGE--
Fractions VIII and IX were analyzed by
SDS-polyacrylamide gel. After electrophoresis, proteins were visualized
by Coomassie Blue staining.
Activity Gel Electrophoresis--
The 3'-5' DNA exonuclease
activity was detected in situ after SDS-PAGE by activity gel
electrophoresis, following the method of Longley and Mosbaugh (45) with
slight modifications. A 10% SDS-polyacrylamide gel was prepared with
the addition of 0.1 mM EDTA and 1 nM f1-K12
single-stranded DNA (Sequence 6, bottom), which was annealed to an
[32P]dATP-labeled oligonucleotide with
L-OddCMP at the 3' terminus (Sequence 6, top) as shown.
|
|
|
|
|
|
Samples (20 µl) from fractions VIII and IX were loaded on the
gel for electrophoresis at 4 °C. After electrophoresis, SDS was
extracted from the gel at room temperature by two 30-min incubations with 20 mM Tris-HCl, pH 7.5, 5 mM
2-mercaptoethanol, and 25% (v/v) isopropyl alcohol. To renature the
proteins, the gel was then shaken gently at 4 °C overnight in 20 mM Tris-HCl, pH 7.5, 5 mM 2-mercaptoethanol,
500 µg/ml bovine serum albumin, and 20% glycerol to renature the
proteins. The exonuclease activity was renatured by shaking the gel for
3 h at 37 °C in renaturation buffer supplemented with 2 mM MgCl2. The gel was subsequently dried, and
the exonuclease activity was visualized by autoradiography.
Matrix-assisted Laser Desorption Ionization (MALDI) Protein
Analysis--
The ~35-kDa polypeptide on 10% SDS-polyacrylamide
gel, which correlated with exonuclease activity from fraction IX in the activity gel analysis, was excised and sent for MALDI protein analysis
at the HHMI Biopolymer/Keck Foundation Biotechnology Resource
Laboratory (Yale University).
Western Blotting with APE1 Monoclonal Antibody--
To further
confirm the result of MALDI protein analysis, we performed a Western
blot experiment with a monoclonal APE1 antibody (Novus Biologicals,
Inc., Littleton, CO). Sample (20 µl) from fraction IX was loaded on a
10% SDS-polyacrylamide gel and electrophoresed. The proteins were
electrotransferred to a nitrocellulose membrane. The membrane was
blocked with 5% milk in PBS supplemented with 0.002% Tween 20 before
the addition of monoclonal APE1 antibody. The membrane was later
developed using the ECL system (Amersham Pharmacia Biotech) and
visualized by exposure to an Eastman Kodak Co. MR film.
Exonuclease Activity of APE1 toward Different Analogs at the 3'
DNA Terminus--
DNA substrates with different deoxyribonucleoside
analogs at the 3' terminus were made to assess the APE1 exonuclease
activity. The different deoxyribonucleoside analogs were incorporated
into 32P-5'-end-labeled oligonucleotide C by incubating 20 µM of the nucleoside 5'-triphosphate analogs and 1 unit
of HIV reverse transcriptase (HIV RT) at 37 °C for 30 min. The
desired oligonucleotide products were purified by electrophoresis on a
20% urea/polyacrylamide gel followed by excising the bands and shaking
them gently in 10 mM Tris-HCl, pH 7.5, buffer supplemented
with 1 mM EDTA overnight at room temperature to extract the
oligonucleotides. These single-stranded oligonucleotides were then
annealed to a 33-mer template to form partial double-stranded DNA
substrates. These DNA substrates were incubated with same amount of
APE1 under the standard exonuclease assay condition, and the result was
analyzed on a 12.5% urea/polyacrylamide gel. The gel was subsequently
dried followed by autoradiography.
Primer Extension Reaction--
To demonstrate that the DNA
product of APE1 was extendable by DNA polymerase after removing
L-OddCMP, a primer extension reaction was performed using
HIV RT. After a 30-min incubation of APE1 with L-OddC
terminated DNA in the exonuclease assay condition described earlier,
0.5 units of HIV RT and 30 µM dNTPs were added to the
reaction mixture to a final volume of 15 µl for another 5-min
incubation at 37 °C. The reaction was stopped by adding 6 µl of
gel loading solution and heating at 80 °C for 3 min before analysis
on a 12.5% urea/polyacrylamide gel.
| |
RESULTS |
Purification of the Exonuclease That Removed
L-()-OddCMP from 3' Termini of DNA--
The presence of
exonuclease activity that removed L-OddCMP was detected in
nucleic acid-free crude extract (fraction II). Two L-OddCMP
exonuclease activity peaks were observed in the fractions from the P11
column (Fig. 2a). The first
activity peak was eluted between 250 and 290 mM KCl, and
the second activity peak (fraction III) was eluted between 400 and 500 mM KCl. Meanwhile, there was only one broad activity
without a distinct peak when deoxycytidine-terminated oligonucleotide B
was used as substrate (Fig. 2b). The latter eluted from the
column between 250 and 570 mM KCl. The lower bands in Fig.
2a are the released [-32P]dAMP from
oligonucleotide A, an indication of the exonuclease action. The dNMPs
were sequentially excised after the removal of L-OddCMP by
this exonuclease. The 20-mer oligonucleotide product after the removal
of L-OddCMP was not always observed, perhaps due to the
excess amount of enzyme. However, the distribution of activities in the
eluted fractions from column chromatography were different when
oligonucleotide A or B was employed, as represented by the intensity of
released [-32P]dAMP at the bottom of the gel. This
result suggested that the exonucleases in different fractions had
different substrate specificity. Since our purpose was to purify the
major exonuclease, which removed L-OddCMP, fraction
III was chosen for further purification.
View larger version (70K):
[in this window]
[in a new window]
|
Fig. 2.
Exonuclease activity profile from the
fractions of P11 column. A 50-600 mM KCl linear
gradient was applied to the column to elute proteins. Aliquots from
every other fraction were assayed for exonuclease activity using either
oligonucleotide A with L-OddC at the 3' termini
(a) or Oligonucleotide B with deoxycytidine at the 3'
termini (b) under the conditions described under
"Materials and Methods." As indicated in a, the lower
band was [-32P]dAMP removed by DNA exonuclease
activity from oligonucleotide substrates. In b, the lower
band was [-32P]dCMP released by the exonuclease
activity. The amount of exonuclease activity was reflected by the
intensity of each band.
|
|
Fraction III was further purified using a DE52 anion exchange column.
As shown in Fig. 3a, most of
the observed exonuclease activity was not retained on the column
(fraction IV) when either L-OddCMP- or dCMP-terminated
oligonucleotide DNAs were used as substrates. An insignificant
exonuclease activity was found to be bound to the column and had a
preference for natural DNA substrate (oligonucleotide B) instead of
L-OddCMP (oligonucleotide A).
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 3.
Exonuclease activity profile from DE52,
single-stranded DNA, and phenyl-Sepharose columns. The exonuclease
activity was further purified sequentially using a DE52 anion exchange
column (a), a single-stranded DNA cellulose column
(b), and a phenyl-Sepharose column (c). As
described under "Experimental Procedures," a 50-600 mM
KCl linear gradient was applied to both DE 52 and single-stranded DNA
columns, and a 1.0-0 M KCl linear gradient was used
for the phenyl-Sepharose column to elute proteins. The exonuclease
activity was assayed with either oligonucleotide A or B (data not
shown) as described previously.
|
|
Fraction IV was further purified using a single-stranded DNA column.
The exonuclease eluted from the column between the concentrations of
220 and 350 mM KCl (fraction V) (Fig. 3b). This
was the only exonuclease activity peak detected when either
oligonucleotide A or B was used.
Fraction V was further purified using a hydrophobic phenyl-Sepharose
column. A fraction of the active protein was not retained on the column
(fraction VI). The majority of the activity, however, eluted from the
column between 540 and 300 mM KCl (fraction VII) (Fig.
3c). This exonuclease had activity against both
L-OddCMP and dCMP. Fractions VI and VII were each loaded
onto a 1 ml P11 column for concentration, and both exonuclease
activities eluted from P11 columns at 400 mM KCl (fraction
VIII and IX, respectively).
At this stage of purification, a single major polypeptide of
approximately 35 kDa and several minor contaminants were observed in
fraction IX by Coomassie Blue-stained SDS-PAGE analysis (Fig. 4, lane 1). In
fraction VIII, no major protein bands were discernible (data not
shown). The exonuclease activities from fractions VIII and IX were
stable at 20 °C when stored in 20 mM Tris-HCl, 300 mM KCl, 1 mM EDTA, 0.1 mM dTT, and
20% glycerol. The enzyme specific activity of fraction IX was
determined to be at least 3000 units/mg; the unit of enzyme was defined
as removal of 10 pmol of L-OddCMP at the 3' termini of DNA
at 37 °C for 30 min under our assay condition.
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 4.
Identification of the major DNA exonuclease
for the removal of L-OddC from the 3' terminus of DNA.
Lane 1, aliquot (20 µl) from fraction IX was subjected to
a SDS-PAGE analysis followed by Coomassie Blue staining. Lane
2, L-OddC-containing DNA substrate was embedded in a
10% SDS- polyacrylamide gel. An aliquot (20 µl) from fraction IX was
subjected to electrophoresis. The enzymatic activity that removed
L-OddC was visualized by autoradiography after the
renaturation treatment, as described under "Experimental
Procedures." Lane 3, same procedures as in lane
1, but the polypeptides were electrotransferred to a
nitrocellulose membrane after SDS-PAGE. The membrane was subsequently
probed with monoclonal APE1 antibody and developed using the ECL
system. The result was visualized by exposing to a Kodak MR film.
|
|
In Situ Activity Gel--
To identify the DNA exonuclease that
removes L-OddC from the 3' terminus of DNA, an in
situ activity gel technique (45) was used. This analysis confirmed
that the 3'-5' DNA exonuclease was associated with the major of 35-kDa
polypeptide in fraction IX (Fig. 4, lane 2). The
polypeptide was recovered from the gel, and a partial amino acid
sequence was determined using MALDI. The sequence matched that of APE1,
the major base excision repair enzyme. To further confirm the identity
of our purified protein, a commercial monoclonal antibody to APE1 was
used in Western blot analysis. The antibody identified the same 35-kDa
major polypeptide in fraction IX (Fig. 4, lane
3).
Since one of the major cellular functions of APE1 is believed to be
repair of abasic sites in DNA, we tested the activity of our partially
purified 35-kDa protein on a double-stranded oligonucleotide containing
the AP site analog tetrahydrofuran (F) (oligonucleotide E).
Reciprocally, purified recombinant APE1 was incubated with
L-OddC-terminated oligonucleotide C to examine its 3'-5'
exonuclease activity. As shown in Fig. 5,
the partially purified APE1 was able to cleave the furan
containing DNA, and the recombinant APE1 possessed 3'-5'
exonuclease activity to remove L-OddC from the 3' terminus
of DNA.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5.
Comparison of the activities of purified APE1
and recombinant APE1 on AP DNA and L-OddC-terminated
DNA. Tetrahydrofuran containing DNA (oligonucleotide E) or
DNA with L-OddC at the 3' termini (oligonucleotide C) were
used as substrates for testing the endonuclease and exonuclease
activities in fraction IX and recombinant APE1. Reaction conditions are
described under "Experimental Procedures." Lane 1,
oligonucleotide E; lane 2, oligonucleotide E and fraction
IX; lane 3, oligonucleotide E and recombinant APE1;
lane 4, oligonucleotide C; lane 5,
oligonucleotide C and fraction IX; lane 6, oligonucleotide C
and recombinant APE1.
|
|
A triplet polypeptide band with an approximate molecular mass of
44 kDa was also observed in addition to APE1 on Coomassie Blue-stained
SDS-polyacrylamide gel from our final preparation (fraction IX). The
triplet was also analyzed by MALDI and appeared to be derived from one
protein, p38-2G4, a cell cycle-modulated nuclear protein (46). A
previous report showed that the murine homologous p38-2G4 co-purified
with AP endonuclease (47) from mouse sarcoma cells. However, the
recombinant p38-2G4 had no major impact on the AP endonuclease enzyme
activity (47). Throughout our studies, all of the characteristics of
the partially purified APE1 and the recombinant APE1 were very similar
(data not shown).
Reaction Parameters of the 3'- 5' DNA Exonuclease Activity of
APE1--
No exonuclease activity was detected in 20 mM
Tris-HCl, pH 7.5, buffer supplemented with 30 mM KCl and
1.0 mM EDTA. This 3'-5' DNA exonuclease activity of APE1
can be restored by adding 2.5 mM divalent cation
Mg2+ or Mn2+ (data not shown). However,
Ca2+ or Zn2+ could not restore the activity.
The optimal salt concentration for APE1 exonuclease activity was 30 mM KCl, while 50% of the enzyme activity was inhibited
between 60 and 90 mM KCl (data not shown).
Since polyamines were shown to have impact on some DNA-associated
enzyme activities (48), the effect of polyamine on this exonuclease
activity of APE1 was also addressed. Spermine and spermidine showed
50% inhibition at concentrations of 0.1 and 0.2 mM,
respectively, but 2 mM putrescine was required to show 50%
inhibition (data not shown).
Substrate Specificity of APE1 Exonuclease Action--
Previous
reports indicated that the 3'-5' DNA exonuclease activity of APE1 is
3-4 orders of magnitude less efficient than the AP endonuclease
activity. In our study, the exonuclease activity of APE1 was the major
enzyme activity to remove L-OddC from the 3' termini of
DNA. Since L-OddC is an unnatural
L-configuration nucleoside analog, we compared the
substrate preference between L-OddCMP and dCMP terminated
DNA. As shown in Fig. 6a,
L-OddCMP-terminated DNA was a better substrate than
dCMP-terminated DNA. Furthermore, to elucidate whether DNAs with
L-deoxyribonucleoside analogs at the 3' termini were
preferable substrates for APE1 exonuclease activity over DNA with a
D-configuration deoxyribonucleoside, three different
D-configuration deoxyribonucleoside analogs (dFdC, araC,
and ddC) and four L-configuration analogs:
L-ddC, L-OddC, L-Fd4C, and
L-SddC were examined as substrates. The different analogs
at the 3' terminus caused slight differences in mobility on 12.5%
urea/polyacrylamide gel otherwise identical oligonucleotide (Fig. 6a). The relative removal efficiency of different
analogs from the 3' termini is reported in Table
I (data normalized to that of
L-OddC, which was considered 100%). The 3'-5' DNA
exonuclease activity of APE1 exhibited a substrate preference for the
L-configuration analogs at the 3' termini of DNA over the
D-configuration analogs. The 5-fold removal efficiency of
L-ddC over the ddC is a good example that the
L-configuration analog is a better substrate for the
exonuclease activity. Among these four L-configuration analogs, L-OddC was the best substrate for this
exonuclease. L-Fd4C, a nucleoside analog with the
introduction of a fluoride on the 5-position of cytosine and a double
bond between the 2'- and 3'-carbon of the L-deoxyribose
moiety, was also a good substrate of this exonuclease. Interestingly,
replacing the 2'-oxygen of sugar moiety of L-OddC with a
sulfur (L-SddC) reduced the removal efficiency 2-fold.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 6.
APE1 exonuclease activity toward partial
double-stranded DNA with different deoxycytidine analogs at the 3'
terminus. a, the DNA substrates with different deoxycytidine
analogs at the 3' termini were examined for removal efficiencies of
APE1. b, The purified APE1 was incubated with
oligonucleotide C for 30 min at 37 °C before the addition of HIV RT
and dNTPs.
|
|
View this table:
[in this window]
[in a new window]
|
Table I
The efficiency of removing various analogs from the 3' termini of DNA
by APE1 exonuclease
DNA oligonucleotides with various analogs at the 3' terminus were used
as substrates for APE1 exonuclease activity. The reaction conditions
were as described under "Experimental Procedures." The removal
efficiency was calculated by dividing the product radioactivity by the
total radioactivity of substrate and product. The efficiencies of
removal of different analogs were adjusted using the efficiency of
L-OddCMP removal as a 100% standard.
|
|
In comparison with the double-stranded DNA substrates, single-stranded
DNAs with different deoxyribonucleoside analogs at the 3' termini were
not substrates for the exonuclease activity of APE1 under the same
reaction conditions (data not shown).
The product of APE1 after removal of L-OddCMP from the 3'
termini was also examined to see if it could be extended by DNA polymerase-HIV RT in this study. As shown in Fig. 6b, in the
presence of dNTPs, HIV RT could not extend
L-OddCMP-terminated DNA. However, after the removal of
L-OddCMP by APE1, HIV RT was able to extend the product,
indicating the presence of a 3'-hydroxyl group.
| |
DISCUSSION |
One of the major mechanisms of action of deoxyribonucleoside
analogs is the incorporation into DNA by DNA polymerases followed by
cessation of cellular or viral DNA synthesis. However, the steady-state
level of deoxyribonucleoside analogs in DNA depends on several factors,
such as the enzymes involved in the process of metabolizing these
deoxyribonucleoside analogs to deoxyribonucleotides, the DNA
polymerases that incorporate analogs into DNA, and DNA repair enzymes
that can remove these analogs from DNA. The incorporation of the novel
anticancer drug L-OddCMP into DNA has been shown to
terminate DNA elongation (3). However, DNA with L-OddC at the 3' termini may be subject to actions of DNA exonucleases. Indeed, a
previous study indicated that the extent of L-OddC in DNA
was correlated with the cytotoxicity of L-OddC and that
L-OddCMP could be removed from DNA in cultured cells
(2).
Several non-DNA polymerase-associated exonucleases with the ability to
remove nucleoside analogs from the 3' terminus of DNA have been
identified (8-15). The physiological role of these exonucleases is
presumably the maintenance of DNA integrity; however, none of those
studied showed preference for the removal of
L-configuration deoxyribonucleoside analogs from the 3' end
of DNA (8-15). Here we report the purification and identification of
APE1 as the major DNA 3'-5' exonuclease from the nuclei of human
leukemic blast cells that can remove L-OddCMP from DNA preferentially.
APE1 is a multiple function enzyme, which in addition to its DNA
endonuclease and 3'-5' DNA exonuclease activity, also possesses 3'-phosphatase, 3'-phosphodiesterase, and RNaseH activity. The 3'-5'
DNA exonuclease activity was reported to be 3-4 orders of magnitude
less active than the endonuclease activity. Demple et al.
(32) reported that the exonuclease activity of APE1 is sensitive to the
reaction buffer condition and the nature of duplex DNA substrates. Our
results are consistent with the fact that DNA with dCMP at the 3'
termini is a poor substrate for APE1 exonuclease but DNA with
L-OddCMP at the 3' termini was a good substrate.
The proposed mechanism for the endonuclease activity of APE1 involves a
divalent cation stabilization of a transitional state leading to
hydrolysis by facilitating the leaving of the phosphate group. A
similar mechanism could be involved in the exonuclease activity, since
the latter also requires divalent cations. The observation that
requirement of Mg2+ or Mn2+ could not be
substituted by Ca2+ or Zn2+ might be due to a
less than ideal coordination between these metal ions and critical
amino acid residues in the catalytic site.
Polyamines also influenced APE1 exonuclease activity. Spermine,
spermidine, and putrescine inhibited the exonuclease activity at
physiological concentrations. Since polyamines are highly positive molecules, they could alter DNA structure and stabilize the staggering base pairing between L-OddC and deoxyguanosine and
therefore affect the cleavage by APE1 exonuclease activity.
The substrate preference for L-OddC over dCyd is likely to
be caused by the unnatural configuration of L-OddC.
According to computer modeling, the unnatural configuration of
L-OddC does not affect the base stacking stability and
hydrogen bonding in DNA. The substrate preference for L-
versus D-configuration analogs of APE1
exonuclease was clearly shown, when DNA substrates with D-
or L-ddC at the 3' termini were used. This APE1 exonuclease showed a 5-fold higher efficiency in removing L-ddC over
D-ddC, which is very different from the exonuclease
activities reported to date. The hypothesis that
L-configuration nucleoside analogs are better substrates
for the exonuclease activity of APE1 was further supported when
L-Fd4C- and L-SddC-terminated substrates were
tested. In general, L-configuration analogs are better
substrates than the D-configuration substrates for the
exonuclease activity of APE1. This is very different from the
exonuclease activity reported by others before. For example, TREX-1 was
shown to remove araCMP and dNMP from the 3' terminus of DNA (14), and
cytosolic and p53 associated exonucleases have preferences for removal
of D-configuration nucleoside analogs over L-
configuration analogs from 3' termini of DNA (9, 10). Crystal structure
analysis revealed that binding of APE1 to AP DNA resulted in a
flipped-out AP site and a racemized -anomer AP site (49).
Since the endonuclease and exonuclease activities of APE1 are believed
to share the same catalytic site (20), the unnatural structure of
L-OddCMP may mimic the structure of the transition state
during APE1 endonuclease action on AP sites, explaining why
L-OddC and other L-configuration nucleoside
analogs are better substrates of APE1 than dCyd at the 3' termini of DNA.
Although APE1 exonuclease has a preference for
L-configuration analogs, the removal efficiency was
different among the analogs examined. In the case of
L-SddC, the replacement of oxygen with a sulfur on the
3'-position of the deoxyribose moiety reduced the efficiency 2-fold
when compared with L-OddC. The introduction of a double bond between
the 2'- and 3'-position on the deoxyribose moiety in L-Fd4C
decreased the efficiency about 20% compared with that of
L-OddC. These results suggested that the configuration and
structure of the deoxyribose could play an important role in
determining the substrate specificity for APE1 exonuclease. Molecular
modeling and the kinetics of APE1 exonuclease toward different
deoxyribonucleoside analogs at the 3' terminus need to be further
studied to facilitate understanding of the substrate specificity.
Since the DNA product of the exonuclease activity of APE1 after removal
of L-OddC can be extended by DNA polymerases, it is likely
that no other enzyme is required during the excision step of the repair
process. It is possible that the intracellular activity of APE1 may
have an important influence in determining the pharmacological effects
of L-OddC and other L-configuration nucleoside
analogs but not for the D-configuration nucleoside analogs
such as dFdC, araC, or ddC.
Previous reports demonstrated the importance of APE1 during the early
embryonic development and growth in mice (30, 31). The discovery of
this novel activity of APE1 suggests that it may be important not only
for its AP endonuclease and redox activities but also as an
exonuclease. Correlation between APE1 expression level in the cell, the
redox state, the phosphorylation status, and the cytotoxicity of
deoxyribonucleoside analogs requires further investigation.
TOP
ABSTRACT
INTRODUCTION
To whom correspondence should be addressed: Department of
Pharmacology, Yale University School of Medicine, Sterling Hall of
Medicine, B315, 333 Cedar St., New Haven, CT 06520. Tel.: 203-785-7118; Fax: 203-785-7129; E-mail: Cheng.lab@yale.edu.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
