The catalytic power of uracil DNA glycosylase in the opening of thymine base pairs

Article abstract of DOI:10.1021/ja062978n

Uracil DNA glycosylase (UNG) locates uracil and its structural congener thymine in the context of duplex DNA using a base flipping mechanism. NMR imino proton exchange measurements were performed on free and UNG-bound DNA duplexes in which a single thymin

Full text of DOI:10.1021/ja062978n

Published on Web 09/15/2006
The Catalytic Power of Uracil DNA Glycosylase in the Opening of Thymine
Base Pairs
Chunyang Cao, Yu Lin Jiang, Daniel J. Krosky, and James T. Stivers*
Department of Pharmacology and Molecular Sciences, The Johns Hopkins UniVersity School of Medicine,
Baltimore, Maryland 21205
Received April 28, 2006; E-mail: jstivers@jhmi.edu
Duplex DNA is a dynamic structure that shows both rapid
unpairing of individual base pairs and larger cooperative unfolding
events that occur on much slower time scales.1,2 Since DNA is a
ligand or substrate for a plethora of proteins and enzymes, it is of
significant interest to understand the role of such dynamic fluctua-
tions in site specific recognition. One paramount example is the
role of spontaneous DNA base pair opening in the process of
enzymatic recognition of damaged DNA bases by DNA gly-
cosylases.3 These DNA repair enzymes must locate extremely rare
oxidized, deaminated, or alkylated bases in a large background of
normal DNA bases in the genome that may be present in >107-
fold excess over the damaged base.3 In principle, two general
mechanisms may be envisioned for location of such damaged
sites: either the site itself posseses dynamic or structural features
that lead to passive enzymatic recognition, or the enzyme posseses
specialized search tools that allow for active inspection of the
damaged base while its distinguishing structural features are still
partially or completely obscured in the DNA base stack.
The DNA repair enzyme uracil DNA glycosylase (UNG) must
find uracil bases in the genome that arise from spontaneous
deamination of cytosine bases or the misincorporation of dUTP
opposite to adenine during DNA replication.3 It is well-established
that UNG uses an extrahelical recognition mechanism where the
uracil base and sugar are rotated 180° from the DNA base stack
into a highly specific recognition pocket that excludes other natural
DNA bases, including thymine and cytosine.4,5 Thus, one question
in uracil recognition is when along the base flipping reaction
coordinate UNG discriminates between the structural congeners T
and U. We have previously developed an NMR approach to address
the question of whether UNG can transiently open TA base pairs
and whether initial recognition involves spontaneous opening of
the base pair to dock thymine in a loose recognition pocket that
serves as a sieve to allow UNG to discriminate U from T1. These
studies clearly revealed that Escherichia coli UNG specifically
stabilizes thymine in an extrahelical state by slowing the closing
rate. Most interestingly, UNG does this without increasing the rate
of spontaneous TA base pair opening when it binds. The results
were consistent with a passive mechanism of TA base pair
inspection that is initiated by spontaneous, thermally induced
motions of the thymine base that serve to expose it to a transient
pyrimidine binding site on UNG.6 Additional evidence that UNG
can flip thymine is provided by the observation that the enzyme
can be converted into a thymine DNA glycosylase by simply
enlarging the active site such that the 5-methyl group is sterically
accommodated.7,8
Figure 1. Sequence of palindromic 10mer duplexes containing isostructural
T6X base pairs with one, two or three hydrogen bonds (D ) diaminopurine,
N ) 6-methylpurine).
Scheme 1. DNA Base Pair Opening and Imino Proton Exchangea
a The constant kint is the intrinsic rate for intramolecular catalysis of
proton exchange by weakly basic N1 and N3 groups on A and C bases (ref
9). The chemical exchange rate is kex ) kB[B] + kint
.
closing rates differ widely in the free DNA. This is expected since,
in a first approximation, the interaction of thymine with the UNG
recognition site should be independent of the nature of the TX base
pair. UNG is an ideal system for this analysis because it does not
directly interact with the base opposing the uracil or thymine.4,5
To provide a rigorous test of these predictions we have now
measured DNA base pair opening and closing rates in the absence
and presence of UNG for a series of increasingly destabilized TX
base pairs that contain one (T6N), two (T6A), or three (T6D)
hydrogen bonds (Figure 1). Because the imino proton is protected
from solvent while hydrogen bonded in a base pair, the base pair
opening and closing rates (kop and kcl), and the equilibrium constant
for opening (Kop) may be indirectly obtained by measuring the
individual imino proton exchange rates using NMR solvent
magnetization transfer methods (Scheme 1).9 A powerful aspect
of this approach is that the rate-limiting step for exchange may be
changed from the chemical step (kex) to base pair opening (kop) by
addition of increasing concentrations of a proton exchange catalyst
(kB[B]).
If a passive recognition model is generally applicable to UNG,
then (i) the rate of thymine base pair opening in free and UNG-
bound DNA should be equal for a series of thymine base pairs
(TX) regardless of their thermal stability (Figure 1), and (ii) the
closing rates in the presence of UNG should be similar even if the
Comparison of the measured opening rates (kop) in Table 1 for
each T6X base pair in the free and UNG-bound DNA shows that
9
13034
J. AM. CHEM. SOC. 2006, 128, 13034-13035
10.1021/ja062978n CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Base Pair Opening Parameters for the Thymine Base of
the T6X Base Pair in Free and UNG-Bound DNAa
b
R
Kop
kop
kcl
10-6 s-
1
1
TX base pair
(×
106)
(s-
)
(
×
)
T6D
free
bound
free
bound
free
bound
8.7 ( 0.5
120 ( 20
20 ( 0.4
1500 ( 280
500 ( 50
760 ( 100
61 ( 6
7 ( 2
40 ( 10
0.3 ( 0.1
1.8 ( 0.3
T6A
T6N
35 ( 6
138 ( 20
650 ( 200
700 ( 200
0.09 ( 0.02
1.3 ( 0.4
0.94 ( 0.2
a The parameters correspond to the mechanism in Scheme 1 and were
obtained at pH 8.0, T ) 10 °C using difluoroethylamine as a proton
exchange catalyst1. The parameters for the TA base pair have been
previously reported.1 b The constant R takes into account possible steric
effects of DNA that may hinder access of the general base catalyst to the
imino site as compared to the free nucleoside.9
Figure 2. Relative effects of UNG on the equilibrium and kinetic
parameters for T6:X base pair opening and closing. Negative numbers
indicate UNG-induced decreases in a rate or equilibrium constant.
opening (kop). These observations support a mechanism in which
the very earliest event in recognition by UNG involves spontaneous
expulsion of the base into a transient binding site that has been
detected in crystallographic studies of the herpesvirus UNG bound
to pTpTpT.6,14 Once this site is transiently occupied, the base has
an opportunity to partition forward along the base flipping reaction
coordinate (in the case of uracil), or alternatively, to re-enter the
DNA base stack (in the case of thymine). Since the lifetime of the
extrahelical base in the free DNA is so short (∼100 to 800 ns,
Table 1), and UNG exhibits diffusion controlled binding kinetics,
it is an improbable event that UNG encounters the extrahelical base
during a bimolecular collision1. Instead, UNG must be already
bound in the correct register to capture the pyrimidine base as it
emerges from the duplex, providing a mechanistic role for short-
range sliding on duplex DNA.3,5 The universality of such a sieve
mechanism is not yet clear, because crystallographic studies on
MutM, a glycosylase that removes oxidized guanine residues (G°),
suggested that it actively destabilizes GC and G°C base pairs using
a Phe side chain, but only flips G°.15 In contrast, a sieving site that
binds both extrahelical G and G° has been proposed with another
8-oxoguanine DNA glycosylase (hOgg1).16 In any event, solution
NMR has allowed detection of a high energy state important for
UNG recognition, and this state would be difficult or impossible
to uncover using any other method.
UNG does not actively accelerate base pair opening. This finding
extends the previous conclusion to a series of isostructural base
pairs with widely different thermodynamic stabilities1. Although
the enzymatic opening rates show no consistent trend (modest
decreases or a slight increase are observed, Table 1) the enzyme is
found to slow the closing rates (kcl) by factors of 23- and 20-fold
for the T6D (three H-bonds) and T6A (two H-bonds) duplexes, but
to have no effect on any exchange parameter for the destabilized
T6N base pair (one H-bond). In further experiments using a similar
palindromic 10 mer DNA in which the central TA base pair was
moved one position in the sequence (T7A),10 we have confirmed
the general conclusion that UNG selectively promotes imino
exchange of TA base pairs. However, as previously reported, imino
proton exchange of guanine bases adjacent to TA pairs is also
promoted by UNG, which is most reasonably attributed to a
proximity effect because isolated GC pairs show no imino exchange
enhancement with UNG.1,10
We note that the opening equilibrium (Kop) for the T6X series in
the free DNA increases by 55-fold in the order T6D < T6A < T6N,
and the trend is due mainly to increases in the opening rates. Two
unexpected findings are that the opening rates for free T6D and
T6A are similar despite the extra hydrogen bond in the T6D pair
and, furthermore, that free and bound T6D show a 3-fold faster
closing rate than the T6A. A possible explanation for the similar
opening rates is that opening of T6D through the major groove only
involves breaking of the hydrogen bonds involving the 6-NH2 and
N1 positions of D. Thus the third hydrogen bond between thymine
O2 and the 2-NH2 group of D serves as a pivot to partially hold
the rotated T6D pair in the stack and kinetically facilitate the closing
rate. A similar conclusion was previously proposed in a computa-
tional study of GC base pair opening,11 but more extensive
experiments will be required to a establish a pivot mechanism for
base pairs with three hydrogen bonds. UNG increases Kop for the
T6D and T6A duplexes by 13 and 75-fold, but has no effect on Kop
for T6N. Thus, UNG does little work to assist opening of base pairs
that are already kinetically and thermodynamically predisposed to
open. This supports previous studies where it was found that uracil
flipping by UNG is enhanced by weakly hydrogen bonded and
flexible base pairs.12,13
Acknowledgment. This work was supported by NIH Grant
GM56834.
Supporting Information Available: Experimental details, synthesis
of 6-methylpurine phosphoramidite, NMR spectra, exchange time
courses, buffer dependence of exchange rates, and table reporting the
exchange parameters for all imino protons in each TX duplex. This
material is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Cao, C.; Jiang, Y. L.; Stivers, J. T.; Song, F. Nat. Struct. Mol. Biol. 2004,
11, 1230.
(2) Parker, C. N.; Halford, S. E. Cell 1991, 66, 781.
(3) Stivers, J. T.; Jiang, Y. L. Chem. ReV. 2003, 103, 2729.
(4) Parikh, S. S.; Walcher, G.; Jones, G. D.; Slupphaug, G.; Krokan, H. E.;
Blackburn, G. M.; Tainer, J. A. Proc. Natl. Acad. Sci. U.S.A. 2000, 97,
5083.
(5) Stivers, J. T. Prog. Nucleic Acid Res. Mol. Biol. 2004, 77, 37.
(6) Savva, R.; McAuley-Hecht, K.; Brown, T.; Pearl, L. Nature 1995, 373,
487.
(7) Kwon, K.; Jiang, Y.; Stivers, J. Chem. Biol. 2003, 10, 1.
(8) Kavli, B.; Slupphaug, G.; Mol, C. D.; Arvai, A. S.; Peterson, S. B.; Tainer,
J. A.; Krokan, H. E. EMBO J. 1996, 15, 3442.
(9) Gueron, M.; Leroy, J.-L. Methods Enzymol. 1995, 261, 383.
(10) See Supporting Information.
(11) Giudice, E.; Varnai, P.; Lavery, R. Nucleic Acids Res. 2003, 31, 1434.
(12) Krosky, D. J.; Schwarz, F. P.; Stivers, J. T. Biochemistry 2004, 43, 4188.
(13) Krosky, D. J.; Song, F.; Stivers, J. T. Biochemistry 2005, 44, 5949.
(14) Jiang, Y. L.; Krosky, D. J.; Seiple, L.; Stivers, J. T. J. Am. Chem. Soc.
2005, 127, 17412.
(15) Banerjee, A.; Santos, W. L.; Verdine, G. L. Science 2006, 311, 1153.
(16) Banerjee, A.; Yang, W.; Karplus, M.; Verdine, G. L. Nature 2005, 434,
612.
The mechanistic implications of these data are best appreciated
by comparison of the fold changes in the opening and closing
parameters induced by UNG (Figure 2). The catalytic power of
UNG in promoting imino proton exchange may be defined as the
free
free
ratio kopUNG/kop or kclUNG/kcl because increasing the opening
rate (or decreasing the closing rate) promotes imino proton exchange
(Scheme 1). The salient finding of Figure 2 is that UNG’s catalytic
power is attributed mostly to decreases in kcl, whereas UNG has
little catalytic power with respect to accelerating TX base pair
JA062978N
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13035