(Received for publication, August 27, 1996, and in revised form, November 20, 1996)
From the Sealy Center for Molecular Science,
University of Texas Medical Branch, Galveston, Texas 77555, and the
§ Department of Chemistry, New York University,
New York, New York 10003
The Escherichia coli UvrB and UvrC
proteins play key roles in DNA damage processing and incisions during
nucleotide excision repair. To study the DNA structural requirements
and protein-DNA intermediates formed during these processes,
benzo[a]pyrene diol epoxide-damaged and structure-specific 50-base
pair substrates were constructed. DNA fragments containing a
preexisting 3 incision were rapidly and efficiently incised 5
to the
adduct. Gel mobility shift assays indicated that this substrate
supported UvrA dissociation from the UvrB-DNA complex, which led to
efficient incision. Experiments with a DNA fragment containing an
internal noncomplementary 11-base region surrounding the
benzo[a]pyrene diol epoxide adduct indicated that UvrABC
nuclease does not require fully duplexed DNA for binding and incision.
In the absence of UvrA, UvrB (UvrC) bound to an 11-base
noncomplementary region containing a 3
nick (Y substrate), forming a
stable protein-DNA complex (Kd ~5-10
nM). Formation of this complex was absolutely dependent
upon UvrC. Addition to this complex of ATP, but not adenosine
5
-(
,
-iminotriphosphate) or adenosine
5
-(
,
-methylene)triphosphate, caused incision three or four
nucleotides 5
to the double strand-single strand junction. The ATPase
activity of native UvrB is activated upon interaction with UvrC and
enhanced further by the addition of Y substrate. Incision of this Y
structure occurs even without DNA damage. Thus the UvrBC complex is a
structure-specific, ATP-dependent endonuclease.
The Escherichia coli UvrABC endonuclease mediates the
repair of a broad spectrum of DNA adducts during the process of
nucleotide excision repair (for reviews, see Refs. 1-4). This process
can be viewed as a series of integrated steps. The UvrA protein (103.8 kDa) exists in equilibrium between a monomer and dimer; the latter interacts in solution with UvrB (76.1 kDa) to form a heterotrimer, UvrA2B (5, 6). The binding of UvrA2B to the
site of damage results in a conformational change in the protein-DNA
complex, the dissociation of UvrA2, and the formation of a
stable preincision UvrB-DNA complex, in which the DNA is unwound and
bent (7, 8). After UvrA is released from the damaged site, the UvrC (68.5 kDa) protein interacts with the COOH terminus of UvrB in the
UvrB-DNA intermediate and is believed to trigger an endonuclease activity in UvrB which cleaves a phosphodiester bond 4-7 phosphates 3
to the damage (9-11). After the 3
incision, a second incision occurs
at the eighth phosphate backbone 5
to the damaged residue (2). The
oligonucleotide containing the damaged nucleotide is then removed from
the DNA and the gap filled by the dual action of DNA helicase II (UvrD)
and DNA polymerase I. The final step of nucleotide excision repair is
performed by DNA ligase, which joins the two ends of the nick.
One of the most remarkable aspects of the UvrABC system, and nucleotide excision repair in general, is its ability to repair a large variety of DNA lesions (1, 12). Although significant progress has been made in understanding the overall mechanism of the E. coli nucleotide excision repair system, little is known about the precise molecular interactions involved in damage recognition by the UvrABC system. Evidence has shown that specific binding of UvrA to damage is not directly correlated to incision efficiency (10, 13). This implies that other Uvr proteins may be involved in the damage recognition process.
Experiments by several laboratories (for review, see Ref. 14) have
shown that UvrB, once loaded onto DNA by UvrA, can serve as a scaffold
for the binding of UvrC, leading to subsequent 3 and 5
incision. To
understand both the contribution to damage recognition and the nuclease
mechanisms of the UvrB and UvrC proteins, it would be beneficial to
study the activity of the UvrB and UvrC proteins in the absence of
UvrA.
We have defined a series of structure-specific DNA substrates and
studied the protein-DNA intermediates formed between these substrates
and the UvrB and/or UvrC proteins in the presence or absence of UvrA.
Results presented here demonstrate that the UvrB and UvrC proteins
promote the assembly of a complex that incises DNA 5 to the adduct in
an ATP-dependent manner and thus acts as a bonafide
nuclease. Our study also shows that UvrB is involved in damage
recognition, as part of a multilevel discrimination system, and that
the 3
incision is a rate-limiting step in the overall mechanism of the
UvrABC nuclease system.
Chemicals
Tris base, boric acid, EDTA, and MgCl2 were
purchased from Sigma. Acrylamide, ammonium persulfate,
N,N-methylenebisacrylamide, and urea were
obtained from Life Technologies, Inc. [
-32P]ATP was
purchased from DuPont NEN. Racemic BPDE1
was purchased from the National Cancer Institute Chemical Carcinogen Reference Standard Repository.
Purification of Proteins
Purification of UvrAUvrA was purified from E. coli strain MH1 UvrA containing the overproducing plasmid pSST10 (graciously supplied by L. Grossman, Johns Hopkins University), which is under the control of the heat-inducible PL promoter. UvrB and UvrC were overproduced from E. coli strain CH296 containing plasmids pUNC211 and pDR3274, respectively (graciously supplied by A. Sancar, University of North Carolina). The UvrA was purified to homogeneity as described previously (10). The UvrB and UvrC proteins were purified using modified procedures as described below.
Purification of UvrBE. coli cells
(DIQ) containing the expression plasmid pUNC211 were grown
to 1.0 OD in super broth and induced with
isopropyl-1-thio--D-galactopyranoside (1 mM)
for 3 h in a 5-liter fermenter, collected by centrifugation, and
resuspended in 100 mM KCl buffer A (50 mM MOPS,
pH 7.5, 1 mM EDTA, 10 mM
-mercaptoethanol,
10% glycerol) at 5 ml/g of packed cells. A mixture of protease
inhibitors (1 mM phenylmethylsulfonyl fluoride, 1 µg/ml
leupeptin, 1 µg/ml pepstatin) was added, and the cell slurry (~75
g) was brought to 20% (w/v)
(NH4)2SO4 saturation by the
addition of solid (NH4)2SO4. The
E. coli cells were then disrupted by sonication at 4 °C
and centrifuged at 15 K for 60 min to pellet the cell debris. The
supernatant was brought to 60%
(NH4)2SO4 saturation by the
addition of solid (NH4)2SO4 over 30 min. The precipitated protein was pelleted by centrifugation (12 K, 20 min), resuspended in a minimal volume of 20% saturated (NH4)2SO4 in buffer A, and loaded
onto a phenyl-Sepharose column. The column was eluted with a linear
gradient of (20% saturated (NH4)2SO4 in buffer A) to 50 mM KCl in buffer A. The fractions containing UvrB, as
determined by SDS-PAGE, were pooled and dialyzed against 50 mM KCl in buffer A. The dialysate was loaded onto a blue
Sepharose column and developed with a linear gradient of KCl (0.05-1
M) in buffer A; UvrB eluted at ~250 mM KCl.
The fractions containing UvrB, as determined by SDS-PAGE, were pooled
and dialyzed against 50 mM KCl in buffer A. The dialysate
was loaded onto a fast protein liquid chromatography Mono Q (10/10)
column that was eluted with a shallow linear gradient of KCl (50-500
mM) in buffer A; UvrB eluted off the column at 300 mM KCl. The fractions containing UvrB were pooled and
brought to 20% (NH4)2SO4
saturation by the addition of solid
(NH4)2SO4. The protein pool was
loaded onto a fast protein liquid chromatography phenyl-Sepharose
column (5/5) and eluted with a gradient from 20% saturated
(NH4)2SO4 in buffer A to 50 mM KCl in buffer A. Purity was determined by SDS-PAGE
analysis of 10 µg of UvrB and Coomassie Blue staining. The fractions
containing pure UvrB were pooled, dialyzed into storage buffer (50 mM Tris, pH 7.5, 100 mM KCl, 1 mM
EDTA, 0.1 mM dithiothreitol, 50% glycerol) and stored at
80 °C.
E. coli cells (CH296)
containing the expression plasmid pDR3274 were grown to 1.0 OD, induced
with isopropyl-1-thio--D-galactopyranoside (1 mM) for 3 h in a 5-liter fermenter, collected by
centrifugation, and resuspended in 5 ml/g of cells 300 mM
KCl buffer C (100 mM KPO4, pH 7.5, 1 mM EDTA, 20% glycerol). A mixture of protease inhibitors
(1 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 1 µg/ml pepstatin) was added to the cell slurry. The E. coli
cells were then disrupted by sonication at 4 °C and centrifuged at
15 K for 60 min to pellet the cell debris. The supernatant was loaded onto a Q-Sepharose Fast Flow column equilibrated in 300 mM
KCl buffer C. The flow-through fractions were collected and applied to
a newly prepared single-stranded DNA-cellulose column (60 ml) equilibrated in 300 mM KCl buffer C. The column was
developed with a 300-ml linear gradient of KCl (0.3-1.5 M)
in buffer C. UvrC eluted off the column at ~600 mM KCl in
buffer C. The fractions containing UvrC (as determined by SDS-PAGE)
were pooled and diluted to 150 mM KCl by the addition of
buffer C. The protein pool (~200 ml) was then applied to a fast
protein liquid chromatography Mono S (5/5) column and eluted with a
50-ml linear gradient of KCl (0.15-1 M) in buffer C; UvrC
eluted off the column at 200 mM KCl. Purity was determined
by SDS-PAGE analysis of 10 µg of UvrC and Coomassie Blue staining.
The fractions containing pure UvrC were pooled, dialyzed into storage
buffer, as defined earlier, and stored at
80 °C.
Other Enzymes
All restriction and modifying enzymes were obtained from Promega or New England Biolabs unless otherwise indicated.
Construction of DNA Substrates
Oligodeoxynucleotides were synthesized on an Applied BioSystem
394 DNA/RNA synthesizer. After synthesis, all oligomers were purified
by PAGE under denaturing conditions. The oligodeoxynucleotide 11-mer
(Fig. 1) containing a single
(+)-cis-anti-benzo[a]pyrene adduct was synthesized,
purified, and characterized as described previously (15). The
BPDE-adducted 11-mer was then used to construct all single
adduct-containing 50-bp substrates used in this study (Fig. 1).
Construction was carried out by procedures described previously (10).
Briefly, the phosphorylated BPDE-11-mers (20 pmol) were incubated with
equal mol of the 20-mer (5-terminally labeled with 32P),
the phosphorylated 19-mer (for S2 and S5, Fig. 1), or none (for S3 and
S6, Fig. 1), and the complementary bottom strand 50-mer in a 50-µl
solution containing 50 mM Tris-HCl, pH 7.8, 10 mM MgCl2, 10 mM dithiothreitol, 1 mM ATP, and 50 µg/ml bovine serum albumin. The mixture
was brought to 75 °C for 5 min, cooled slowly to room temperature,
then cooled further to 16 °C. T4 DNA ligase (0.2 unit) was added and
the ligation reaction carried out at 16 °C for 12 h. After
ligation, the sample was denatured by the addition of 8 M
urea and rapid chilling on ice, then purified on a 12% (w/v)
polyacrylamide sequencing gel under denaturing conditions. Bands
identified by autoradiography as BPDE-50-mer and BPDE-31-mer were
excised, eluted, and precipitated with ethanol. The purified adducted
DNA strand (50-mer) was then annealed with the 50-mer complementary
strand, then purified on a 10% polyacrylamide nondenaturing gel. The
31-mer was annealed in the presence of the 19-mer and the complementary
50-mer. All nondamaged substrates were constructed directly by
annealing PAGE-purified complementary oligomers.
The double-stranded character and homogeneity of the 50-bp substrates
were examined by a restriction enzyme assay in which the substrate (2 nM) was incubated with RsaI or HaeIII
(6 units) at 37 °C for 30 min. Reaction products were then analyzed
on a 12% polyacrylamide sequencing gel under denaturing conditions as
described before. Concentrations of the BPDE-adducted DNA duplexes were
determined by UV absorption at 260 nm with 260 = 660,000 M
1 cm
1 and confirmed by the
same specific activity of the 32P-labeled substrates as
quantified by liquid scintillation counting (Beckman LS6000SC) and the
DNA concentration.
The substrates were sequenced by standard Maxam-Gilbert procedures (16) with a DuPont NEN sequencing kit.
DNA Binding Assay
Binding of the 50-bp DNA substrates by UvrA, UvrB, and UvrC proteins was assayed qualitatively by gel mobility shift. Typically, the substrate (2 nM) was incubated with the indicated concentrations of UvrA, UvrB, UvrC, or some combination at 37 °C for 15 min in 20 µl of UvrABC buffer (50 mM Tris-HCl, pH 7.5, 50 mM KCl, 10 mM MgCl2, 5 mM dithiothreitol) in the presence or absence of 1 mM ATP. After incubation, 2 µl of 80%(v/v) glycerol was added, and the mixture was loaded immediately onto a 3 or 3.5% native polyacrylamide gel in TBE buffer (10 mM Tris borate, pH 8.1, 1 mM EDTA) and electrophoresed at 80 volts at room temperature. The gel was dried and exposed for an appropriate time to an x-ray film with intensifying screen or a PhosphorImaging screen (Molecular Dynamics).
DNA Incision Assay
The 5-terminally labeled DNA substrate (2 nM) was
digested with UvrA, UvrB, and UvrC, or with UvrB and UvrC in the UvrABC buffer (± 1 mM ATP) at 37 °C for 15 min. Uvr subunits
were diluted and premixed into storage buffer before mixing with DNA.
Reactions were terminated by adding EDTA (20 mM) or heating
to 90 °C for 3 min. The samples were denatured with formamide (50%
v/v) and heated to 90 °C and then quick chilled on ice. The digested
products were analyzed by electrophoresis on 12% polyacrylamide
sequencing gel under denaturing conditions with TBE buffer and
quantified as described above.
ATP Hydrolysis Assay
The conversion of ATP to ADP by the UvrABC system was determined on a Pharmacia Ultra III spectrophotometer using a coupled enzyme system consisting of pyruvate kinase and lactate dehydrogenase to link the hydrolysis of ATP to the oxidation of NADH. The assay mixture consisted of 50 mM Tris acetate, pH 7.5, 100 mM potassium acetate, 10% glycerol, 4 mM MgCl2, 0.1 mM dithiothreitol, 0.15 mM NADH, 2 mM phosphoenolpyruvate and the UvrA, UvrB, and/or UvrC in the presence or absence of substrate 8 (Fig. 1) in various combinations and concentrations as indicated in Table I. The reaction mixture (0.5 ml) was allowed equilibrate to 33 °C, and the assay was initiated by the addition of ATP (0.5 mM). The rate of ATP hydrolysis was calculated from the linear change in absorbance at 340 nm over 30 min which accompanied the oxidation of NADH.
|
Western Blotting Analysis
After the gel mobility shift assay described above, the complex formed by Uvr proteins interacting with the DNA substrates was identified by Western blotting analysis. Briefly, the gel was blotted to a nitrocellulose immobilization membrane (Schleicher & Schuell) using a Hoefer electrotransfer unit and the manufacturer's instructions. The membrane was then treated with UvrB antiserum (graciously supplied by L. Grossman, John Hopkins University) or UvrC (graciously supplied by E. Tang, University of Texas M. D. Anderson Cancer Center) antiserum for ECL Western blotting (Amersham).
Quantification of Incision and DNA Binding Products
Incision products were quantified using a PhosphorImager 425 (Molecular Dynamics) as described previously (10).
To probe
the structure of protein-DNA intermediates that form during the
assembly of the UvrABC nuclease, we have synthesized a series of eight
structure-specific DNA substrates; substrates 2, 3, 5, and 6 contain a
single BPDE adduct at position 26 in the 50-bp sequence (Fig. 1). The
double-stranded character of all substrates was confirmed by a
restriction assay. All substrates were 5-terminally labeled on the top
strand except for S8, which was 5
-terminally labeled on the bottom
strand.
Incision of each of the 5 end-labeled BPDE-DNA substrates
(S1-S6) by the UvrA, UvrB, and UvrC proteins is shown in Fig.
2. Incision was not observed for the two nondamaged
substrates S1 and S4. Although S4 contains a region of 11 mismatched
bases, it is not a substrate for the UvrABC nucleotide excision repair system. However, the addition of a BPDE adduct in the middle of the
unpaired region (S5) results in normal incision at the eighth phosphodiester bond (P8) 5
to the adduct as reported previously (10).
The mismatched region did not affect the incision efficiency, with S2
and S5 being incised at 1.3 and 1.1 fmol/min, respectively. Compared
with the normal substrate (S2), the substrate containing a preexisting
3
nick (S3), which is one nucleotide further away 3
from the adduct
than the major 3
incision (P5; 10), shows a increase in its 5
incision efficiency, 2.3 fmol/min (Fig. 2). The same observation has
been obtained by Moolenaar et al. (11) for a cisplatin-GG
diadduct. In spite of the shift of the 3
incision site 1 base in the
3
direction, the 5
incision of the substrate remains at the eighth
phosphodiester, as in the normal substrate S2 (Fig. 2), which is
consistent with the results reported previously (10, 11). The fact that
a 3
nick facilitates the 5
endonuclease activity suggests that there
is a kinetically slow step through which the UvrABC nuclease system
must proceed to make the 3
incision.
Among the substrates incised by the UvrABC system, S6 displayed 5
incision at a site one nucleotide further away 5
(P9) from the adduct
and a much weaker incision site at P10 two nucleotides further away,
versus the other substrates, which were incised at the P8
position 5
from the adduct (Fig. 2); S6 was cleaved with the same
incision efficiency as the normal substrate (S2). Given that S6
contains a nicked unpaired 11-base region around the adduct and that
UvrB unwinds and bends the DNA helix at the adduct when forming the
preincision complex UvrB-DNA, the shift of the 5
incision site may
imply that the position and size of the unpaired region play a role in
determining the position of the 5
incision.
To
examine any correlation between the formation of stable
protein·BPDE-DNA complexes and incision, binding of UvrA and/or UvrB
to the defined substrates (Fig. 1) was examined by gel mobility shift
assays. Fig. 3 shows the interactions of the UvrA and
UvrB proteins with the S2, S4, and S5 substrates. Comparison with the incision results in Fig. 2 suggested that the lack of incision of S4 by
UvrABC is due to the lack of formation of the UvrB-DNA complex. The
data in Fig. 3 suggest that the S2 and S5 substrates are incised
equivalently (Fig. 2), probably because they support the production of
approximately equal amounts of UvrB-DNA complex.
Comparison of the protein complexes formed on substrates S2 and S3
indicates that the 3-nicked substrate, S3, supported the formation of
more UvrB-DNA complex than S2, whereas in contrast S2 generated much
more UvrA2B-DNA complex (Fig.
4A). This dramatic increase in the formation
of the UvrB-DNA complex for S3 is most likely due to the 3
nick
causing a destabilization of UvrA2 from the protein-DNA
complex, so that UvrA2-S3 is less stable than UvrA2-S2. Incision of the same substrates by UvrABC
nuclease (under these experimental conditions) (Figs. 4B)
showed that S3 supports a much higher incision efficiency than for S2
(~5-fold), which formed much more UvrA2B-DNA complex than
S3 (~5-fold, Fig. 4A). These results suggest a direct
correlation of the formation of UvrB-DNA complex with incision of DNA
damage by UvrABC nuclease.
Incision of Defined Substrates by UvrB and UvrC Proteins
The
major role of UvrA in nucleotide excision repair is to provide about 3 orders of magnitude in damage discrimination and to load UvrB to the
damage site. Previous studies of interaction of UvrB and UvrC with
damaged DNA have only been possible in the presence of UvrA. To probe
the interaction of the UvrB and UvrC proteins with damaged DNA, we
wanted to find substrates that do not require UvrA involvement. If the
UvrA2B interaction leads to a locally unwound and melted
structure, we reasoned that the need for UvrA in the incision process
might be obviated by placing the damaged site in a locally unpaired
region (S5, S6). Thus it may be possible to demonstrate a direct
interaction of UvrB and/or UvrC with certain DNA structures resembling
preincision intermediates. To test this hypothesis, UvrB and UvrC were
allowed to react with the 5-terminally labeled substrates S2-S6 and
S8 (Fig. 1) in the absence of UvrA (Fig. 5).
S6, the Y substrate containing a nicked unpaired region surrounding the
BPDE-DNA adduct, was the only substrate that supported incision by the
UvrBC nuclease (lane 5). Although S3 is a 3
-nicked
substrate, it contains a fully base paired region surrounding the
damage (in contrast to S6), and no incision was observed (lane
2). These results also show that the Y substrate-specific incision
was strand-specific because no incision was observed on the other
strand (S8, lane 7). Furthermore, the lack of incision in
the absence of ATP (lane 6) indicated that ATP hydrolysis or
binding of hydrolyzed products to the proteins was necessary for UvrBC
incision. In control reactions, (Fig. 6), no 5
incision
was produced by the individual Uvr proteins. DNA sequence analysis
indicated that the UvrBC nuclease incised the Y substrate at the 9th
and 10th phosphodiester bonds 5
to the adduct (Fig. 5, lane
5, and Fig. 6, lane 7), which is identical to the
incision sites by UvrABC (Fig. 6, lane 6). These sites have
been found to be located in the double-stranded region three and four
nucleotides 5
to the junction of the duplex and the unpaired region.
Since the junction of the duplex unpaired region appears to be a structural characteristic recognized by UvrB-UvrC proteins, it was reasonable to ask whether the nondamaged Y structure such as S7 (Fig. 1) is a target of UvrB-UvrC nuclease. As shown in Fig. 6, S7 was incised by UvrB-UvrC proteins at the same positions as was S6 substrate but with lower efficiency.
Binding of UvrB and UvrC to Nicked Unpaired SubstrateUvrBC
nuclease incision of the Y structures (substrates S6 and S7) strongly
suggests that the UvrB-UvrC heteromeric complex is able to recognize
and bind to the DNA substrates under appropriate conditions. As
shown in Fig. 7A, the interaction of UvrA,
UvrB, and/or UvrC proteins with the Y structure, substrate S7, occurred in the absence of ATP. The nonadducted Y structure was recognized by
UvrA, as shown in lanes 2 and 3. However, neither
UvrB nor UvrC bound alone to the substrate (lanes 4 and
5). Similar to the normal damaged substrate S2, UvrA and
UvrB together bound to the S7 substrate to form UvrA2B-DNA
complex (lane 6). Most importantly, a complex formed when
both the UvrB and UvrC proteins were incubated with the S7 substrate
(lanes 7-10). Since UvrC was found to cause DNA
aggregation, a 5:1 ratio of UvrB to UvrC was used for the assay.
Western blotting with UvrB antiserum (Fig. 7B) revealed that
UvrB was a participant in the complex with DNA. Unfortunately the
antibodies to UvrC also cross-reacted with UvrB, so we could not rule
out the presence of UvrC in this complex. The observed formation of
UvrB/UvrC-DNA complex agrees with the results of the 5 incision made
by UvrB and UvrC proteins on the S6 and S7 substrates.
ATP Hydrolysis in the UvrBC Incision
The results described above (Fig. 5) indicated that incision of the nicked unpaired Y substrate by the UvrB and UvrC proteins is an ATP-dependent process. Therefore, incision experiments were performed in the absence and presence of ATP and its analogs including ADP, AMP-PNP, and AMP-PCP, to determine whether ATP binding or hydrolysis was necessary for the incision process. We found that both ATP and ADP can facilitate incision, 100 and 77%, respectively, whereas no incision was observed in the presence of nonhydrolyzable AMP-PNP or AMP-PCP (data not shown). The near full activity of incision with ADP implies that binding of ADP may be involved in the incision mechanism.
An ATP hydrolysis assay was also employed to monitor the ATPase
activity of the UvrB protein during the incision process. Analysis of
the sequence of UvrB displays one Walker-type ATPase motif (GKT) near
the NH2 terminus of the protein. It has been reported that
native UvrB shows no ATPase activity unless it is cleaved at residues
607-610 near the COOH terminus, yielding a 69-kDa protein, UvrB* (11,
17). The truncated UvrB protein is still able to form normal
preincision complexes with UvrA on DNA (5, 11, 18) but loses its
nuclease activity. The data shown in Table I indicate that the ATPase
activity of the native UvrB is activated when the protein interacts
with UvrC or the Y structure alone and is stimulated greatly by the
addition of both UvrC and the Y structure, which is consistent with the
incision results shown above. These data, combined with the fact that
UvrB and UvrC interact to form a protein-DNA complex on the S6, S7, and
S8 in the absence of ATP and that ATP hydrolysis is required for
incision, suggest that ATP hydrolysis or binding of ADP to UvrB is
required for 5 incision produced by the UvrABC system.
Experiments performed in this study illustrate several important
points regarding incision of the defined substrates by the UvrABC
nuclease system (Fig. 2). 1) Neither of the undamaged S1 and S4
substrates is incised by the UvrABC nuclease, even though the latter
contains an unpaired region of 11 bases. 2) The presence of an adduct
in the middle of the unpaired region (S5) supports incision equivalent
to that of the normal adducted substrate S2. 3) S3, which contains a
nick 3 to the adduct, was incised about 3-5 times as efficiently as
the normal substrate S2. 4) S6, a nicked unpaired Y substrate, is
incised as well as S2, the normal substrate. 5) Incision of the Y
substrate occurs in the absence of a DNA adduct although with lower
efficiency than either S2 or S6.
It has been shown that the local conformation of the DNA adduct is an important determinant for UvrABC nuclease recognition and incision (10, 19). Formation of the preincision complex of UvrB-DNA results in unwinding and bending of the DNA helix at the adduct (9, 20). It seems that DNA conformational changes induced by the modified base and by protein-DNA complex formation play an important role in the process of damage recognition and nucleotide excision repair. In this report, we studied the interaction of UvrABC with DNA substrates containing partially single-stranded structure (S4 and S5). Although the 11-base unpaired substrate, S4, is recognized by UvrA nearly as well as the BPDE and nicked substrates (Fig. 3), no incision of this substrate was observed (Fig. 2A). This result indicates that UvrA recognizes confirmational alterations in the DNA duplex and confirms earlier suggestions that there is no direct correlation between specific binding of UvrA to DNA and UvrABC incision (13). However, positioning a BPDE adduct in the middle of the unpaired region (S5, Fig. 1) results in the same incision efficiency as for S2 (Fig. 2). These results strongly suggest that in addition to the adduct-induced DNA conformation change, the presence of a bulky foreign (non-base) molecule in the DNA sequence is essential for recognition by the UvrABC nuclease. This indispensable requirement is probably indicative of a direct interaction of Uvr proteins with the modified base. Results of these experiments also indicate that large departures from normal B form DNA in the vicinity of the adduct (11-base unpaired region) do not affect recognition and incision by the UvrABC nuclease.
The binding of UvrB to damaged DNA has been suggested as being involved
in the recognition (10, 13, 21). In this study, we demonstrated the
direct correlation between the formation of the UvrB-DNA complex and
the incision (Figs. 2, 3, 4). We propose that there are at least two
levels of discrimination. 1) The specific binding of UvrA detects a
general DNA distortion, and this interaction helps bring UvrB to the
site of the damaged base, since UvrB itself normally has no ability to
recognize DNA damage in double-stranded DNA; 2) The formation of the
UvrB-DNA complex through the UvrA2B-DNA intermediate
provides the second level of recognition which monitors modified bases,
especially intercalating adducts. The binding of UvrB to the damaged
DNA may be mediated through hydrophobic stacking interactions between
the aromatic amino acid side chains and the bases (10, 12, 21). These
two levels, plus a third level of discrimination, as discussed below,
determine the specificity of the 3 incision near the adduct.
The dramatic increase in the 5 incision of a BPDE-damaged substrate
containing a 3
-preincised substrate (S3) suggests that a rate-limiting
step in the UvrABC mechanism of nucleotide incision has been overcome
(Figs. 2 and 4). This suggestion is supported strongly by the fact that
the UvrB-nicked DNA complex is much more stable than the
UvrA2B-DNA complex (Fig. 4A). These data suggest
that the 3
incision is a rate-limiting step and is a consequence of
successful damage recognition. It has been reported that the 3
incision is triggered by the coil-coil domain interaction of the COOH
terminus of UvrB with UvrC (11). It appears that the slow formation of
the UvrBC-DNA complex may provide the third level of damage
recognition.
Although the UvrABC nuclease has been studied widely as a model of
nucleotide excision repair, much remains to be determined regarding the
mechanism of the post-3 incision process, especially the direct
interaction of UvrB and/or UvrC with the DNA. In this study,
structure-specific DNA substrates containing nicked unpaired Y
structures (S6-S8) were used to investigate the mechanism of the
post-3
incision process. The data presented here demonstrate for the
first time that UvrB and UvrC proteins work together to form a
protein-DNA complex with the nicked unpaired Y substrates in the
absence of UvrA and ATP (Fig. 7). Formation of this UvrB-containing protein-DNA intermediate seems to be specific to the Y structure, as no
similar complex formation was been observed for other substrates (S1-S5) used in this study (data not shown). As a result, the formation of this Uvr-DNA complex leads to efficient 5
incision of the
adduct in the presence of ATP (Fig. 5), which occurs at the
double-stranded region 3-4 bases to the duplex-single strand junction.
This endonuclease activity is strand-specific because no incision has
been observed on the other (bottom) strand, thus the incision is
correlated to the specific feature of the duplex-single strand junction
and is strand orientation-dependent. The specific recognition of the
duplex-single strand feature is confirmed further by the 5
incision of
the nonadducted Y substrate (substrate S7) by the UvrB and UvrC
proteins (Fig. 6). Similar results have been reported for the
Rad1-Rad10 protein complex in Saccharomyces cerevisiae (22,
23). It has been shown that the Rad1-Rad10 nuclease plays a role both
in the cleavage of specific recombination intermediates and in the 5
incision of DNA damage during nucleotide excision repair (22, 23). The
functional similarity between UvrB-UvrC and Rad1-Rad10 nucleases in the
cleavage of duplex-single strand junctions suggests a conservation in
function between E. coli UvrB-UvrC and yeast Rad1-Rad10
complexes.
The requirement of ATP for the 5 endonuclease activity is consistent
with that for the 3
incision by the Uvr proteins. The involvement of
ATP hydrolysis was confirmed further using ATP analogs in the study.
Although previous studies have indicated that protease-truncated UvrB*
has ATPase activity, intact UvrB was believed not to have an associated
ATPase in the absence of UvrA. Our results indicate that the silent
ATPase activity in UvrB can be activated by UvrC and/or a Y type DNA
structure in the process of producing a 5
incision in DNA. The data
also show that the hydrolysis occurs between the
and the
phosphorus atoms. There are two possible roles for ATP hydrolysis in
the final incision: 1) hydrolysis provides energy for the incision reaction, and 2) the binding of ADP to UvrB activates the nuclease.
In conclusion, we have presented the novel finding that the UvrBC
nuclease, in the absence of UvrA, can bind to and perform the 5
incision on the Y type structure in an ATP-dependent
manner. We have shown that damage discrimination and incision by the
UvrABC nuclease system occur in several steps. UvrA recognizes helical distortions and helps bring UvrB to the site of conformational change.
If the distorted region contains no adduct, the UvrAB complex
dissociates. However, if the altered DNA contains an adducted base,
there is a slow conformational change in the protein-DNA complex (19)
such that the DNA is bent, becomes unpaired, and interacts strongly
with UvrB. This causes UvrA to dissociate, leaving behind a stable
UvrB-DNA complex. Recognition of the UvrB-DNA complex by UvrC causes an
allosteric change in the protein-DNA complex to induce an incision 3
to the adducted base. That the 5
incision that rapidly follows
dissociation of UvrA is favored by a substrate containing a nick 3
to
the adduct indicates that the rate-limiting step in the incision
process is the formation of the preincision UvrB-DNA complex and/or
UvrC binding. We favor the hypothesis that UvrA must dissociate before
the binding of UvrC.
We thank Dr. David Konkel for critical reading of this manuscript.