Distorting duplex DNA by dimethylenesulfone substitution: A new class of 'transition state analog' inhibitors for restriction enzymes [13]

Full text of DOI:10.1021/ja972768y

2674
J. Am. Chem. Soc. 1998, 120, 2674-2675
Table 1. Distortion of DNA from Idealized Structure by EcoRV
and in the Linkage Modified by the Dimethylene Sulfone Group
Distorting Duplex DNA by Dimethylenesulfone
Substitution: A New Class of “Transition State
Analog” Inhibitors for Restriction Enzymes
Ki for sulfone
-substituted chimera
linkage
twist (deg)a
G-5-G-4
G-4-G-3
G-3-A-2
A-2-T-1
T-1-A1
32.5/40.1
33.3/32.9
35.4/36.6
19.3/19.8
23.0/16.0
19.9/19.8
37.4/36.6
40.4/32.9
29.6/40.1
32.4
Monika O. Bla¨ttler,§ Christian Wenz,‡ Alfred Pingoud,‡ and
Steven A. Benner*,§
ca. 20 nMb
25 pMc
Department of Chemistry, UniVersity of Florida
GainesVille, Florida 32611
A1-T2
T2-C3
Institute for Biochemistry
C3-C4
Justus-Liebig-UniVersity, Heinrich-Buff-Ring 58
C4-C5
D-35392 Giessen, Germany
ideal A-DNA18
ideal B-DNA14
r(GpC)7
36.0
34.7
20.8
ReceiVed August 8, 1997
r(GSO2C)5
After they bind (but before they cleave) duplex DNA, some
restriction enzymes (such as EcoRV1,2 and EcoRI3) distort the
duplex. The distorted duplex is not, of course, in its ground-
state conformation; it requires “binding energy” to bend DNA.4
Thus, an analogue of DNA that generates this distortion in the
unbound state (without altering other features of the substrate that
are recognized by the enzyme) should bind to these restriction
enzymes with a higher affinity than the DNA substrate itself. This
is, of course, the principle underlying transition-state analogues
generally, which approximate in structure the “distorted” transition
state (or a distorted high-energy intermediate) for an enzymatic
reaction.
Recently, we noted that duplex nucleic acid having a dimeth-
ylene sulfone unit replacing a phosphate has a distorted backbone
conformation5 reminiscent of the distortion produced by restriction
enzymes and other proteins that bend DNA when they bind. In
particular, the twist observed in a duplex built from the dinucle-
otide analogue r(GSO C) is a low 20.8° (instead of 34.7°), similar
to the twist observed2between the central four base pairs in DNA
carrying the recognition sequence bound to EcoRV (19.3-23.0°)
(Table 1). Thus, a DNA duplex having a dimethylene sulfone
substitution joining these base pairs should be “pre-distorted” in
its ground state and, therefore, bind more tightly to EcoRV than
the cognate DNA substrate itself.
a Twist is defined by the relative angle of the C1′-N bonds in
consecutive bases in the strand; the two values given refer to different
values in the two subunits of the noncrystallographic dimer (from
ref 1). b For duplex-incorporating 1 ACCAGAATTCGGATCCA-
GASO TATCGCCA; the recognition site is underlined. Cleavage occurs
2
between TA in the recognition site. c For duplex-incorporating 2
ACCAGAATTCGGATCCAGATASO TCGCCA; the recognition site
is underlined. Cleavage occurs between TA in the recognition site.
2
To learn whether a potent inhibitor of EcoRV could be obtained
by introducing a dimethylenesulfone linker into its restriction site
to replace a phosphate at a position where the twist is low, two
chimeric DNA analogues were synthesized, the first with a
dimethylene sulfone linker replacing a phosphodiester group
between the first AT unit in the EcoRV recognition site
(underlined) (ACCAGAATTCGGATCCAGASO TATCGCCA, 1),
the second with a dimethylene sulfone linker replacing a phos-
phodiester group between the second AT unit in the EcoRV
recognition site (ACCAGAATTCGGATCCAGATASO TCGCCA,
2
2
2). These are the linkages with the lowest twists in the EcoRV
restriction site when bound to the restriction enzyme (twists ≈
20°, Table 1).1,6
Iodide 7 of the 5′-homologated analogue of thymidine was
For such an oligonucleotide analogue to be an effective
inhibitor of the enzyme, the sulfone group must also mimic
interactions that the phosphate group has with the enzyme itself.
EcoRV makes many contacts to the phosphate groups in its
recognition sequence, both directly and via water molecules.1,2,6
As the sulfone SdO bond has a high dipole moment, the sulfone
should accept hydrogen bonds, although perhaps not as well as
phosphate, where each oxygen bears a charge of ca. -0.5.
synthesized by the method of Baeschlin et al. (Scheme 1).8
A
protected derivative of the 3′-homologated 2′-deoxyadenosine 8
was prepared by the procedure of Sanghvi et al.9 and then
converted by a four-step procedure to thiol 12. Thiol 12 and
iodohomothymidine 7 were then coupled, and the product was
converted to the 5′-tritylated derivative of the ASO T phosphora-
midite 16. This was incorporated into chimeric oligonucleotides
2
by standard solid-phase automated oligonucleotide synthesis.
Solvation of the sulfone group in the GSO C duplex5 is rather
2
similar to solvation of the phosphate group in the GPO C duplex,7
-
Neither the modified nor unmodified strand of duplexes
containing 1 and 2 were substrates for EcoRV. The duplexes
were, however, inhibitors of the cleavage of the 5′-radiolabeled
self-complementary EcoRV substrate GATCGACGATATCGTC-
GATC (20 mM Tris/HCl, pH 7.5, 50 mM NaCl, 10 mM MgCl2,
and 100 mg/mL BSA, at 21 ( 1 °C).10 Initial rates kobsd were
determined under steady-state conditions from the linear part of
the progress curves. Products were resolved by chromatography11
on DEAE-cellulose (Machery-Nagel) and quantitated using an
InstantImager (Canberra Packard). For duplexes containing 1,
2
providing experimental evidence suggesting that the sulfone group
might adequately serve as a neutral phosphate mimic.
‡ Justus-Liebig-University.
§ University of Florida.
(1) Winkler, F. K.; Banner, D. W.; Oefner, C.; Tsernoglou, D.; Brown, R.
S.; Heathman, S. P.; Bryan, R. K.; Martin, P. D.; Petratos, K.; Wilson, K. S.
EMBO J. 1993, 12, 1781-1795.
(2) Kostrewa, D.; Winkler, F. K. Biochemistry 1995, 34, 683-696.
(3) Kim, Y.; Choi, J.; Grable, J. C.; Greene, P.; Hager, P.; Rosenberg, J.
M. In Structural Biology: The State of the Art; Sarma, R. H., Sarma, M. H.,
Eds.; Adenine Press: New York, 1994; pp 225-246.
(4) Jencks, W. P. AdV. Enzymol. Relat. Areas Mol. Biol. 1975, 43, 219-
410.
(8) Baeschlin, D. K.; Daube, M.; Bla¨ttler, M. O.; Benner, S. A.; Richert,
C. Tetrahedron Lett. 1996, 37, 1591-1592.
(5) Roughton, A.; Portmann; S., Benner, S. A.; Egli, M. J. Am. Chem.
Soc. 1995, 117, 7249-7250.
(9) Sanghvi, Y. S.; Ross, B.; Bharadwaj, R.; Vasseur, J.-J. Tetrahedron
Lett. 1994, 35, 4697-4700.
(6) Thorogood, H.; Grasby, J. A.; Connolly, B. A. J. Biol. Chem. 1996,
271, 8855-8862.
(10) (a) Wenz, C.; Selent, U.; Wende, W.; Jeltsch, A.; Wolfes, H.; Pingoud,
A. Biochim. Biophys. Acta 1994, 1219, 73-80. (b) Wenz, C.; Jeltsch, A.;
Pingoud, A. J. Biol. Chem. 1996, 271, 5565-5573.
(7) (a) Day, R. O.; Seeman, N. C.; Rosenberg, J. M.; Rich, A. Proc. Natl.
Acad. Sci. U.S.A. 1973, 70, 849-853. (b) Rosenberg, J. M.; Seeman, N. C.;
Day, R. O.; Rich, A. J. Mol. Biol. 1976, 104, 145-167.
(11) Brownlee, G. G.; Saenger, F. Eur. J. Biochem. 1969, 11, 395-399.
S0002-7863(97)02768-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 03/07/1998

Communications to the Editor
J. Am. Chem. Soc., Vol. 120, No. 11, 1998 2675
Scheme 1a
precisely, a “high-energy intermediate analog”) for EcoRV, but
of a special sort, where the geometric distortion is distributed
over many atoms. This result might be compared with the
enhanced binding of nicked DNA,12 DNA duplexes containing
mismatches,13 or DNA duplexes distorted by cross-links14 for
proteins that bend (but do not cleave) DNA.
Furthermore, these results show that the sulfone unit is an
adequate mimic for phosphate in its hydrogen-bonding potential,
at least for the second AT linkage. The difference between the
Ki values for 1 and 2 might be explained by the differences
between the hydrogen-bonding network in which the correspond-
ing phosphates participate, or differences in the detailed geometry
in the linkages. In a DNA-EcoRV complex soaked with Mg2+
,
the phosphate replaced by a dimethylene sulfone unit in 1 forms
a hydrogen bond with the side chain of Ser 112 and three water
molecules, which interact directly or via water with hydrogen-
bonding groups from Asn 185, Thr 186, Thr 187, Asn 188, and
the next phosphate in the chain.1,6 The phosphate replaced by a
dimethylene sulfone in 26 is involved in a less extensive hydrogen-
bonding network. These differences may explain the greater
tolerance of the enzyme for sulfone substitution at the second
position. Alternatively, the dimethylene sulfone unit may mimic
the distorted geometry of the second AT linkage better than the
first. Interestingly, the fact that the chimera containing 2 is not
a substrate for the restriction endonuclease is consistent with the
proposal that the phosphate group 3′ to the phosphate being
cleaved helps activate the attacking water15 (however, see also
ref 2).
Many proteins that bind DNA distort it by bending and/or
unwinding. These include restriction enzymes, transcription
factors, and other regulatory elements. DNA analogues that are
“pre-distorted” by a backbone substitution should therefore be
general tools for blocking a variety of protein-nucleic acid
interactions. Furthermore, neutral analogues for phosphate are
frequently sought,16 and the dimethylene sulfone group should
be considered as a potential candidate. Finally, short oligonucle-
otide analogues, if they have the correct geometries and cor-
respondingly high affinities, could have value as biological
reagents, alternatives to “antisense” and “antigene” oligonucle-
otides and their analogues, which typically must be 16-30
nucleotides in length to exhibit useful selectivity (for a review,
see ref 17).
a Key: (a) PPh3/I2/pyridine; (b) triphenylsilyl (TPS) chloride/CH2Cl2/
imidazole (78% over 2 steps); (c) (iPr)OSO2CH2Li/THF/hexamethylphos-
phoric triamide (74%); (d) PPh3/I2/benzene/∆ (55%); (e) NaBH4/EtOH
(aqueous, 80%); (f) AcSH/DIAD/PPh3/THF (79%); (g) BzCl/4-(N,N-
dimethylamino)pyridine (DMAP)/pyridine; (h) NaOH/pyridine/EtOH
(49% over 2 steps); (i) Cs2CO3/THF/DMF; (j) oxone/THF/MeOH/NaOAc;
(k) pyridine/HF (69% over 3 steps); (l) DMTCl/DMAP/TEA/pyridine
(66%); (m) ((iPr)2N)2POCH2CH2CN/diisopropylammonium tetrazolide/
CH3CN/DMF (94%); TPS ) tert-butyldiphenylsilyl group; DMT ) 4,4′-
dimethoxytrityl; T ) thymine; A ) adenine; Ac ) acetyl; Bz ) benzoyl.
2, and a control oligodeoxynucleotide lacking an EcoRV site
(GATCGACGAGCTCGTCGATC), concentrations of 100, 1, and
1000 nM (respectively) were required to observe greater than 50%
reduction of the initial cleavage velocity.
Values for kcat and Km for the substrate were determined from
data collected without inhibitor. Kinetics simulations (Microsoft
EXCEL) were then performed for data collected with inhibitors
assuming competitive, noncompetitive, and uncompetitive inhibi-
tion. Under a competitive inhibition model, a best fit to the
experimental data was obtained with KEI (enzyme-inhibitor
disassociation constants) values of 21 nm and 190 nM for the
duplex incorporating 1 and the control. Further kinetic analysis,
at 0.1-30 nM [substrate] and 0, 0.1, and 0.3 nM [inhibitor], was
done for the best inhibitor, containing strand 2. Values for the
kinetic parameters kcat and Km and the inhibition constants KEI
and KEIS for competitive, noncompetitive, and uncompetitive
inhibition were obtained by fitting. Both noncompetitive and
uncompetitive mechanisms gave a mean residual roughly three
times larger than for competitive inhibition, which gave an
inhibition constant of 25 pM. A free fit gave no significant
improvement of the mean residual relative to the fit obtained if
competitive inhibition is assumed. Consequently, the data indicate
Acknowledgment. This paper is dedicated to Prof. Dieter Seebach
on the occasion of his 60th birthday.
Supporting Information Available: Experimental details, one table,
and three figures (9 pages, print/PDF). See any current masthead page
for ordering information and Web access instructions.
that the DNA duplex having the recognition sequence GATATSO
-
JA972768Y
2
TC (2) is a competitive inhibitor of EcoRV.
(12) Koudelka, G. B.; Harbury, P., Harrison, S. C.; Ptashne, M. Proc. Nat.
Acad. Sci. U.S.A. 1988, 85, 4633-4637.
The remarkably low inhibition constant of 2 (25 pM) might
be compared with the 2 nM KM for a standard 20-mer DNA
sequence lacking the dimethylene sulfone modification.10 The tight
(13) Klimasauskas, S.; Roberts, R. J. Nucleic Acids Res. 1995, 23, 1388-
1395.
(14) Erlanson, D. A.; Chen, L.; Verdine, G. L. J. Am. Chem. Soc. 1993,
115, 12583-12584.
binding of the duplex having the GATASO TC recognition site to
2
the EcoRV enzyme suggests that the distortion induced by the
dimethylene sulfone unit allows the chimeric duplex DNA to fit
in the active site without consuming “intrinsic” binding energy
to distort the structure from its ground-state conformation.4 This
makes the duplex a “transition-state analog” (or, perhaps more
(15) Jeltsch, A.; Alves, J.; Wolfes, H.; Maass, G.; Pingoud, A. Proc. Nat.
Acad. Sci. U.S.A. 1993, 90, 8499-8503.
(16) Nomizu, M.; Otaka, A.; Burke, T. R.; Roller, P. P. Tetrahedron 1994,
50, 2691-2702.
(17) Stein, C. A.; Chen, Y.-C. Science 1993, 261, 1004-1012.
(18) Kennard, O.; Hunter, W. N. Angew. Chem. 1991, 103, 1280-1304.