ABSTRACT
Drosophila
Rrp1 (recombination repair protein 1) is a DNA repair enzyme whose nuclease
activities include AP-endonuclease, 3
'
-exonuclease, 3
'
-phosphodiesterase and 3
'
-phosphatase. This study investigates the sequence specificity of the dsDNA
3
'
-exonuclease activity of Rrp1. We demonstrate that the activity is more
efficient in purine-rich regions of dsDNA than in pyrimidine-rich regions. Rrp1 exonuclease activity is examined at 3
'
-terminal homopurine or homopyrimidine tracts, at junctions between purine- and pyrimidine-rich sequences and upon encountering repeated dinucleotide runs. The data show that purine-purine and 3
'
-pyrimidine-5
'
-purine dinucleotide bonds are cleaved faster than 3
'
-purine-5
'
-pyrimidine or pyrimidine-pyrimidine bonds. Thus, the base occupying the penultimate position
in the 3
'
-terminal dinucleotide may be important in determining the relative
efficiency of bond cleavage by Rrp1. These findings may reflect upon specific
DNA-protein interactions in the enzyme active site.
The fundamental role played by DNA as the source of genetic information
necessitates the existence of DNA repair pathways. Through the action of DNA
repair enzymes, genomic integrity is preserved despite exposure to exogenous and endogenous sources of DNA damage.
One component of damage is mismatches and errors introduced during DNA replication or recombination. Other damage results from
interactions with intracellular and extracellular chemical compounds. Reactive oxygen species and alkylating agents are two such sources
of damage to DNA. Oxygen-dependent DNA damage can be generated by environmental factors such as radiation, heat and some antitumor drugs. The importance of DNA repair has been
increasingly recognized in recent years, due to the mounting evidence
correlating DNA damage with cancer and aging in humans (
1
,
2
).
One enzyme involved in the repair of oxidative and alkylation damage to DNA is
the recombination repair protein 1 (Rrp1) from
Drosophila melanogaster
. Oxidative and alkylation damage generates AP sites, which are one of the most
common DNA lesions (
3
,
4
). Rrp1 cleaves AP sites, one of the essential steps in base excision repair (
5
,
6
). Other DNA nuclease activities of Rrp1 include dsDNA 3'-exonuclease, 3'-phosphodiesterase and 3'-phosphatase (
5
,
7
). These activities are associated with the C-terminal region of Rrp1, which is homologous to
Escherichia coli
exonuclease III and a number of eukaryotic homologs (
6
). Rrp1 has a unique N-terminal region which is associated with two additional enzyme functions:
DNA strand transfer and ssDNA renaturation (
8
).
In this report, the sequence specificity of the dsDNA 3'-exonuclease activity of Rrp1 is investigated. Rrp1 and the related protein
E.coli
exonuclease III demonstrate a 3'-exonuclease activity with several similar properties; both enzymes show a strong preference for recessed dsDNA 3'-ends compared with protruding ssDNA 3'-ends and have no detectable activity on
either ssDNA or dsDNA 5'-ends (
6
,
8
). Previous studies using dsDNA fragments with heterogeneous DNA sequence (naturally occurring sequences) indicate that the specific activity of the Rrp1 3'-exonuclease is at least two orders of magnitude lower than the
specific activity of exonuclease III (
8
). We demonstrate that the cleavage rate of the 3'-terminal nucleotide by Rrp1 can vary by up to two orders of
magnitude depending on the specific nucleotide and its sequence context. 3'-Terminal runs of purines are removed at a faster rate than 3'-terminal runs of pyrimidines; thus with some DNA
substrates Rrp1 displays a sequence-dependent strand specificity, cleaving only the purine-rich DNA strand. The activity of Rrp1 encountering purine-pyrimidine dinucleotides and junctions was also examined. We
speculate that the sequence specificity of Rrp1-catalyzed phosphodiester backbone cleavage is determined by specific
protein-DNA interactions in the enzyme active site.
Oligonucleotides were purchased from Oligos Etc. Inc. The sequences of
oligonucleotides used are as follows:
A18,
5'-GGGGGGAAGGAAAAAAAA-3';
T26,
5'-TTTTTTTTTTTTCCTTCCCCCCCCCC-3';
G18,
5'-AAAAAAAAGGGGGGGGGG-3';
C26,
5'-CCCCCCCCCCCCCCTTTTTTTTTTTT-3';
C19,
5'-TTTTTTTTCCCCCCCCCCC-3';
C21,
5'-TTTTTTTTCCCCCCCCCCCCC-3';
G27,
5'-GGGGGGGGGGGGGGGAAAAAAAAAAAA-3';
T18,
5'-CCCCCCCCTTTTTTTTTT-3';
T20,
5'-CCCCCCCCTTTTTTTTTTTT-3';
A26,
5'-AAAAAAAAAAAAAAGGGGGGGGGGGG-3';
TAEX-1,
5'-CGGCGGTGGCGGTTTTTTTTCCCCTTTTTT-3';
TAEX-2,
5'-GGGGGGAAAAAAGGGGAAAAAAAA-3';
TGJ1,
5'-GGGGGGGGTTTTTTTG-3';
TGJ3
,
5'-GGGGGGGGTTTTTTTGGG-3';
TGJ5,
5'-GGGGGGGGTTTTTTTGGGGG-3';
J2,
5'-CCCCCCCCCAAAAAAACCCCCCCCCCCC-3';
TG19,
5'-GGGGGGGGTGTGTGTGAAA-3';
CA27,
5'-TTTTTTTCACACACACCCCCCCCCCCC-3';
AG19,
5'-GGGGGGGGAGAGAGAGAAA-3';
CT27,
5'-TTTTTTTCTCTCTCTCCCCCCCCCCCC-3';
TC19,
5'-GGGGGGGGTCTCTCTCAAA-3';
GA27,
5'-TTTTTTTGAGAGAGACCCCCCCCCCCC-3';
TA19,
5'-GGGGGGGGTATATATAAAA-3';
AT27,
5'-TTTTTTTTATATATACCCCCCCCCCCC-3'.
Oligonucleotides were labeled with T4 polynucleotide kinase in the presence of [[gamma]-
32
P]adenosine triphosphate using standard conditions (
9
). The specific activity was determined by running an aliquot of labeled
oligonucleotide on a 16% polyacrylamide gel and determining the amount of
32
P incorporated into the appropriate gel slice containing the oligonucleotide.
The labeled oligonucleotide was purified by preparative gel electrophoresis on
a 16% denaturing polyacrylamide gel and eluted from the gel using an Elutrap
device (Schleicher & Schuell). The recovery was calculated and a dsDNA substrate was prepared by
annealing the substrate and a 2.5-fold molar excess of a complementary unlabeled oligonucleotide. Annealing
was carried out at 80oC for 1 min in a buffer containing 100 mM Tris-HCl, pH 8.0, 20 mM NaCl followed by slow cooling to room
temperature. The proposed structure of each dsDNA substrate (shown in Figs
1
A-
3
A and
5
A and Table
1
) was confirmed by carrying out a primer extension reaction in the presence of
Klenow polymerase (1 U) and the appropriate dNTP (100 [mu]M). The reaction was carried out at 30oC for 5 min in 50 mM Tris-HCl, pH 8.0, 20 mM NaCl, 0.05 mg/ml BSA, 0.2 mM EDTA, 5 mM MgCl
2
, 2 mm DTT. The reaction was stopped by the addition of proteinase K to 0.05 mg/ml and EDTA to 25 mM and was incubated for 10 min at 42oC. Aliquots of 5 [mu]l formamide gel loading buffer (95% formamide, 20 mM EDTA, 0.05%
bromphenol blue, 0.05% Xylene Cyanol FF) were then added to the reaction to
prepare it for gel electrophoresis. The products were analyzed on a 16%
denaturing polyacrylamide gel. In all cases, the results of the Klenow assays
confirmed the expected DNA structure; for example, in the presence of dGTP,
Klenow polymerase extends *TGJ1 by 8 nt, *TGJ3 by 6 nt and *TGJ5 by 4 nt (data
not shown). Minor product and substrate species were detected at 10% or less of
the total
32
P-labeled material.
Reactions were done with the labeled oligonucleotides to investigate the
exonuclease activity of Rrp1, exonuclease III and Klenow polymerase. The
reactions were carried out in a buffer of 50 mM Tris-HCl, pH 8.0, 10 mM NaCl, 0.05 mg/ml BSA, 0.2 mM EDTA, 5 mM MgCl
2
in a final volume of 10 [mu]l. DTT (2 mM) was added for reactions involving Klenow polymerase. Samples
of 0.5 ng (4-8 nM) labeled oligonucleotide were used per reaction. The amount of
enzyme used and the time of the reaction were varied depending on the
particular assay. All reactions were carried out at 30oC and were stopped by the addition of proteinase K to 0.05 mg/ml and EDTA to 25 mM and were incubated for 10 min at 42oC. Aliquots of 5 [mu]l formamide gel loading buffer were added to each sample to
prepare it for gel electrophoresis. The products were separated on a 16% denaturing polyacrylamide gel. Dried gels were analyzed and quantitated using a Molecular Dynamics
Phosphorimager.
Phosphorimager analysis was used to determine the relative fraction of each reaction product. To calculate fmol nucleotide released, the fraction of each product species was multiplied by the fmol oligonucleotide
in the reaction and the difference in nucleotide length between the product and
substrate. Total fmol nucleotide released was calculated by summing over all the product species.
To investigate the exonuclease activity of Rrp1 using defined sequence
oligonucleotides, we tested homopolymeric sequences, including 3'-terminal repeated deoxynucleotides of each of the four bases,
adenine, guanine, cytosine and thymine. Throughout the experiments presented
here, assays were carried out using a dsDNA substrate containing one
32
P-labeled oligonucleotide annealed to an unlabeled complementary
oligonucleotide (except where otherwise indicated). The oligonucleotide
sequences and structure for each of the dsDNA oligonucleotide substrates are
shown either in Table
1
or with the appropriate gel analyses presented in Figures
1
-
4
. The duplex substrate is referred to using the name of the labeled strand
preceded by an asterisk (i.e. *A18), while the single-stranded oligonucleotide of the same name lacks the asterisk. Both ends of
the labeled strand are recessed in the double-stranded substrate, with the exception of *TAEX-1 and *TAEX-2, as noted below. Based on earlier studies (
6
), it was assumed that Rrp1 acts only on dsDNA; thus the single-stranded 3'- and 5'-ends of the unlabeled complementary strand
prevent it from being degraded by Rrp1 (additional tests confirming the
validity of this assumption are also presented below). The polarity of the Rrp1
exonuclease is 3' -> 5' and thus Rrp1 is only able to act on the dsDNA 3'-end of the labeled strand. In this paper,
sequence patterns or dinucleotides are described 3' -> 5'. For example, the terminal dinucleotide GT refers to the 3'-terminal G and the penultimate T.
After initial protein titrations with the oligonucleotide substrates, the exonuclease rate was measured by sequence gel analysis and quantitation (Table
2
). As reported previously, the Rrp1 exonuclease demonstrates a lower specific activity than exonuclease III (
8
). In addition, the activity of Rrp1 differs >100-fold for the oligonucleotides tested. Oligonucleotides with nine or 11 3'-terminal deoxycytosines (*C19 or *C21) and oligonucleotides
with 8 or 10 3'-terminal thymidine nucleotides (*T18 or *T20) are poor substrates,
while 3'-terminal deoxyadenosine or deoxyguanosine tracts are efficiently
cleaved. The poor exonucleolytic activity of Rrp1 through pyrmidine tracts is
detected only at high enzyme concentration and after long incubation times. At
lower ratios of enzyme to DNA and shorter incubations, no detectable activity
is seen (for thymidine tracts, compare Figs
1
B and
2
B). The exonuclease rate for Rrp1 in 3'-terminal thymidine tracts is 250- to 300-fold lower than its rate in 3'-deoxyadenosine tracts. The activity of
E.coli
exonuclease III is slightly lower in repeated thymidine nucleotides, but the
magnitude of the effect is much smaller (8- to 25-fold; Table
2
). In terms of dinucleotide sequences, these results imply that Rrp1 cleaves AA or GG dinucleotides more quickly than
CC or TT dinucleotides.
The kinetics of the Rrp1 exonuclease reaction for oligonucleotides *A18 and *T18, containing 8 or 10 nt tracts of deoxyadenosine and thymidine
respectively, are shown in Figure
1
B. *T18 was treated with 40 or 20 ng Rrp1, each for three time points (6, 12 and
18 min), and *A18 was treated with 0.5 or 0.25 ng Rrp1, each for three time
points (1.5, 3 and 6 min). The stoichiometry of enzyme monomer to DNA 3'-ends is 6.6, 3.3, 0.08 or 0.04 for the reactions with 40, 20, 0.5
and 0.25 ng Rrp1 respectively. A similar rate of degradation of the *A18 and
*T18 substrates was observed under these conditions, in which the enzyme
concentration differed by 80-fold, demonstrating a strong sequence preference for adenine over thymine
nucleotide tracts.
To confirm the observation of strong sequence dependence of the Rrp1 exonuclease
rate, a double-stranded substrate was designed to measure the relative rate of
exonuclease reaction through 3'-terminal runs of adenine and thymine simultaneously. The
oligonucleotide TAEX-1 is a 32mer containing a 3'-terminal run of adenine nucleotides and the complementary
oligonucleotide, TAEX-2, is a 24mer containing a 3'-terminal run of thymine nucleotides. When TAEX-1 and TAEX-2 hybridize to one another, the 3'-end of both strands are recessed;
therefore Rrp1 exonuclease can act at the 3'-end of either strand. The oligonucleotide substrates *TAEX-1 and *TAEX-2 are structurally identical, but differ in which of the two strands is 5'-
32
P-labeled (Fig.
2
A).
Table 2
Four oligonucleotides were designed to investigate the activity of Rrp1 on
substrates with different repeating dinucleotide runs (Fig.
5
A). Both ends of the labeled strand of all four substrates are recessed. All
contained a 3'-terminal run of three adenines followed by four consecutive
dinucleotides: GT (TG19), GA (AG19), CT (TC19) and AT (TA19). Each labeled
strand is a 19mer and the complementary strand of each is a 27mer (CA27, CT27,
GA27 and AT27 respectively).
The activity of Rrp1 on each dinucleotide substrate is shown in Figure
5
B. Rrp1 demonstrated the highest activity on *AG19. This is consistent with the
preference of Rrp1 for adenine and guanine tracts. In a reaction with the
substrate *TC19, which contains alternating TC pyrimidines after the three 3'-terminal adenines, Rrp1 did not remove any nucleotides beyond the
three adenines. Both TC and CT dinucleotides were poorly cleaved, similarly to
CC and TT dinucleotides. *TA19 and *TG19 both contain alternating purines and
pyrimidines after the three 3'-end terminal adenines. For these two substrates, the cleavage
efficiency appeared intermediate between that of the purine and pyrimidine dinucleotide repeat substrates. The pattern of cleavage of both *TA19 and *TG19 showed alternating heavier and lighter bands throughout
the dinucleotide repeats. Since the heavier bands are purine-terminated species (see Fig.
5
B), this suggests that Rrp1 repeatedly slows down at the purine-pyrimidine dinucleotide (as seen above with a GT dinucleotide) and then more quickly
cleaves the following pyrimidine-purine dinucleotide. Rrp1 also demonstrated prefered cleavage of the AG over the GA dinucleotide in the
*AG19 substrate. The cleavage efficiencies of both *TG19 and *TA19 are
significantly less than that of *AG19, suggesting that the lower activity in the alternating purine-pyrimidine tracts can be attributed to the presence of the T residues. Analysis of the kinetics of Rrp1 action on these dinucleotide substrates
also demonstrated similar results. Additionally, *TC19 was still not degraded
into the pyrimidine tract after a 20 min incubation (data not shown). These
results are consistent with those described using homopolymeric and GT junction-containing substrates and indicate strong inhibition of Rrp1 3'-exonuclease by pyrimidine-pyrimidine dinucleotides.
The four dinucleotide substrates were also treated with exonuclease III for 10
min. The results were different from those obtained with Rrp1. Exonuclease III
was active at a much lower concentration than Rrp1. Although like Rrp1 it was
less active on *TC19 than on *AG19, the degree of difference was much less and it did not show a preference for pyrimidine-purine dinucleotides with the alternating pyrimidine-purine substrates *TG19 and *TA19 (data not shown).
Previous experiments demonstrated that Rrp1 requires a duplex DNA substrate for all its nuclease functions, including 3'-exonuclease activity (
5
,
6
; data not shown). In the experiments described here, it is assumed that a
duplex DNA structure is maintained, since no exonuclease activity is expected
on the unlabeled complementary strand oligonucleotides which carry four base
ssDNA extensions on both the 5'- and 3'-ends. Two experiments were carried out to test the
validity of this assumption. In the first experiment, we assayed for the
integrity of the DNA duplex after an intial Rrp1 treatment in which the labeled
DNA strand was degraded. *TC19 and *AG19 were treated with Rrp1 for 10 min,
after which the NaCl concentration was increased to inhibit Rrp1 exonuclease activity. Then Klenow polymerase and the appropriate dNTPs were added. The reaction was allowed to proceed for another 5 min. Both
labeled DNA strands were extended by the polymerase to the full length of the
complementary strand which was utilized as the DNA template in this reaction
(data not shown). This result confirms that for these two substrates both a DNA
duplex structure and the integrity of the complementary unlabeled DNA strand
are conserved after incubation with active Rrp1.
An experiment was also carried out to test directly for the presence or absence
of ssDNA 3'-exonuclease activity in the Rrp1 enzyme fraction. Under conditions
identical to those used in the above experiments, several ssDNA 5'-
32
P-labeled oligonucleotides were tested as Rrp1 substrates (substrates G18, A18, T18 and C19). G18
has a run of 10 guanines at its 3'-end, A18 has a run of 8 adenines, T18 has a run of 10 thymines and
C19 has a run of 11 cytosines. As expected, Rrp1 displayed no detectable
activity on these substrates (data not shown), again confirming the assumed
integrity of the unlabeled strand. The same substrates were efficiently degraded by the exonuclease of Klenow polymerase, which prefers ssDNA substrates (data not shown).
This paper demonstrates the strong sensitivity of the 3'-exonuclease of
Drosophila
Rrp1 to DNA sequence. The rate of cleavage of the 3'-terminal nucleotide by Rrp1 varied by up to two orders of magnitude
depending on the specific nucleotide and its sequence context (Table
2
). Purine-rich sequences containing AA, GG, AG or GA dinucleotides were cleaved much
more efficiently by Rrp1 than pyrimidine-rich sequences containing CC, TT, CT or TC dinucleotides (Table
2
and Figs
1
and
4
), while intermediate effects were seen in alternating purine-pyrimidine tracts. The more rapidly cleaved dinucleotides all have a 5'-purine nucleotide; thus, it appears that the penultimate nucleotide may
influence the cleavage efficiency. The data presented do not support a strong
influence of DNA sequence more distal from the 3'-end, however, such an influence cannot be completely ruled out at
present.
Throughout the experiments the same pattern of sequence specificity was
observed. The substrates used include the following configurations: 3'-terminal homopolymeric runs; thymine tracts recessed from the 3'-terminus by 1, 3 or 5 guanine residues; dinucleotide
repeat patterns; a substrate with two recessed 3'-termini susceptible to Rrp1 exonuclease simultaneously. With the
latter substrate, Rrp1 appeared to act as a strand-specific 3'-exonuclease; Rrp1 degraded the A-rich strand and did not degrade the T-rich strand of this duplex oligonucleotide (Fig.
2
). The suitability of all these substrates for other 3'-exonuclease activities (
E.coli
exonuclease III and Klenow polymerase 3'-exonuclease) also argues against the possibility that a defect
present in some of the substrates determines the observed cleavage specificity in
reactions with Rrp1 (Fig.
4
D and data not shown). Together these results strongly imply that the demonstrated sensitivity of Rrp1 to DNA sequence is an intrinsic property of Rrp1 protein.
The data reported here are consistent with several other properties of the Rrp1 3'-exonuclease. Earlier experiments indicated that the specific activity of the Rrp1 3'-exonuclease is approximately two orders of magnitude lower than that of
E.coli
exonuclease III (
7
,
8
). The specific activity of the Rrp1 3'-exonuclease was previously determined using naturally occurring random sequence DNA and that value lies in between the values reported here for
pyrimidine-rich and purine-rich DNA sequences. While in purine-rich DNA Rrp1 exonuclease activity is only one order of
magnitude less active than exonuclease III (
7
; Table
2
), it is two to three orders of magnitude less active than exonuclease III in
pyrimidine-rich DNA (Table
2
). Thus, the inability of Rrp1 to progress through pyrimidine-rich regions may determine its low activity level. When presented with DNA
fragments in the size range 300-1000 bp, Rrp1 degrades the 3'-ends to a limited extent before stopping, while exonuclease
III can proceed until the two strands fall apart (data not shown). Pyrimidine-tract inhibition could readily account for this observation.
Linxweiler and Horz (
10
) investigated the sequence specificity of
E.coli
exonuclease III and concluded that C residues are released most rapidly, A and
T residues are released at an intermediate rate and G residues are released at
the slowest rate. This pattern deviates from the observations made here, which
indicate that T residues are removed most slowly by exonuclease III (8- to 25-fold more slowly) and a slightly increased removal rate for C
residues (Table
2
). These differing results may stem from the dissimilarity of the DNA substrates
(random sequence or hexamer repeat DNA versus homopolymeric oligonucleotides)
used in these two studies. Similarly to our results with exonuclease III, the
magnitude of the sequence effect on its 3'-exonuclease rate observed in their study was small (up to 3-fold). Although the strongly decreased cleavage rate through
thymidine nucleotides is moderated in exonuclease III (8- to 25-fold) compared with Rrp1 (~250-fold), a common pattern is not seen for relative cleavage
efficiency of deoxycytosine nucleotides by the two enzymes. Thus, pyrimidine-dependent inhibition is not a characteristic that is conserved among these
two related proteins. It is possible that the large unique N-terminal region of Rrp1 (427 amino acids), for which no similar protein
region exists in exonuclease III, influences the enzymatic differences between
exonuclease III and Rrp1 noted here. Our attempts to characterize a Rrp1 mutant with a truncated N-terminus in order to address this possibility have not yet succeeded, due to
difficulties in purifying this truncation mutant in an active soluble form.
The basis for the pyrimidine-dependent inhibition of nuclease action by Rrp1 is not clear. Specific
interactions of the protein and its substrate in the active site may be
involved. The importance of potential major groove interactions between Rrp1
and the purine N-7 position was tested using an oligonucleotide containing 7-deaza adenine residues embedded in an adenine tract (data not
shown). This substitution had a very weak effect, reducing the exonuclease rate
by only ~2-fold, and thus did not support a role for major groove hydrogen bonds
involving N-7 purine. Several potential active site amino acid residues of Rrp1 have
been identified through a mutagenesis screen (
11
). These residues coincide with analogous conserved amino acids in the exonuclease III protein, which have also been placed in the active site by their proximity
in the crystal structure of that protein to both a bound dCTP nucleotide and the single catalytic metal ion (
12
). However, in the absence of a crystal structure with a bound dsDNA substrate, it is
difficult to evaluate the potential interactions in the active site that might
influence the sequence specificity of
Drosophila
Rrp1. Further, differences between the amino acid sequence or tertiary
structures of Rrp1 and exonuclease III may be important contributors to this
specificity. If a mechanistic basis for the Rrp1 sequence specificity is
determined, it may not be conserved between these two related enzymes.
The family of proteins related by protein homology to Rrp1 includes the
prokaryotic enzymes
E.coli
exonuclease III and
Streptococcus pneumoniae
exonuclease A, which are much more potent exonucleases than Rrp1 (
13
,
14
), and several mammalian proteins, including human APE1, mouse APEX and bovine
BAP, which are extremely weak exonucleases (
15
-
20
). In contrast, the high specific activity AP endonuclease function of the
members of this enzyme family is highly conserved. The lack of conservation of
the 3'-exonuclease suggests a possible shift in biological function for
these different proteins in their respective organisms. In addition, we
speculate that it may be possible for a 3'-exonuclease capable of only limited DNA degradation to be
biologically advantageous. However, the biological importance of the sequence
dependence of the Rrp1 exonuclease is not at present clear. An area of future
interest is whether DNA sequence context exerts a strong influence on other
Rrp1-catalyzed functions, including endonucleolytic cleavage of abasic sites or
removal of 3'-terminal phosphoglycolate.
One of the best studied nucleases demonstrating sequence-dependent DNA cleavage is the enzyme DNase I. DNase I acts on a dsDNA substrate and
nicks one DNA strand in a highly sequence-dependent manner; variation in cleavage efficiency of 400-fold has been reported (
21
,
22
). Several high resolution co-crystals of DNase I and DNA substrates have suggested a structural basis
for its sequence dependence that involves primarily DNA substrate minor groove
width and helix stiffness (
23
,
24
). These co-crystals reveal strong induced bending of the DNA substrate, but very
little protein conformational change upon DNA binding. The possibility that DNA structure influences the Rrp1 exonuclease cleavage efficiency, as it does the endonucleolytic
cleavage of DNase I, is speculative at present. However, the unusual result of
this work, the strand-specific sequence preference demonstrated by Rrp1, implies that a strand-specific structural conformation might be involved in this effect.
We anticipate that structural information, when it becomes available, may help
explain the unusual sequence dependency pattern exhibited by Rrp1.
We gratefully acknowledge Richard Sinden, Leroy Worth, Tom Kunkel and Rick Paules for their thoughtful comments on this work.
+
Present address: Cell Biology and Genetics Program, Sloan-Kettering Institute, 1275 York Avenue, New York, NY 10021, USA
Table 1
DNA substrate
Exonuclease rate (fmol/min/[mu]g)
Relative activity normalized to *A18 substrate
Relative activity (Rrp1/ExoIII)
ExoIII
Rrp1
ExoIII
Rrp1
*A18
369 000 +- 180 000
26 800 +- 11 200
1.000
1.000
0.073
*G18
375 000 +- 168 000
8800 +- 2700
1.016
0.328
0.023
*C19
449 000 +- 112 000
1200 +- 750
1.217
0.045
0.003
*C21
245 000 +- 23 900
980 +- 350
0.664
0.037
0.004
*T18
47 100 +- 11 900
100 +- 30
0.128
0.004
0.002
*T20
14 400 +- 900
70 +- 30
0.039
0.003
0.005
REFERENCES
Return