Formation and Stability of a Janus-Wedge Type of DNA Triplex

Article abstract of DOI:10.1021/ja038081x

A new type of DNA targeting with the formation of a Janus-Wedge (J-W) triple helix is described. The "wedge" residue (W) attached to a PNA backbone is designed to insert itself into double-stranded DNA and base pair with both Watson-Crick faces. To study

Full text of DOI:10.1021/ja038081x

Published on Web 12/13/2003
Formation and Stability of a Janus-Wedge Type of DNA Triplex
Dongli Chen, Meena, Sunil K. Sharma, and Larry W. McLaughlin*
Department of Chemistry, Merkert Chemistry Center, Boston College, 2609 Beacon Street,
Chestnut Hill, Massachusetts 02467
Received August 22, 2003; E-mail: larry.mclaughlin@bc.edu
We describe here the formation and stability of a Janus-Wedge
(J-W) type of triple helix: a complex formed from two DNA target
strands and a third probe strand capable of base pairing with the
Watson-Crick faces of the two target sequences.
Conventional DNA triplexes are formed by a third strand binding
to the functional groups on the Hoogsteen face of purine residues.1,2
These binding sites are present in the major groove of duplex DNA
and can be used to target specific polypurine sequences using a
polypyrimidine1,2 (forming T-A-T and C+-G-C base triplets) or
polypurine3 (forming A-A-T and G-G-C base triplets) probe
sequence. The disadvantage of the conventional approach is that
sequences of polypurines are the sole targets. The challenge to
generalized targeting is the intervening pyrimidines within a
polypurine sequence; pyrimidine targeting approaches include
derivatives that employ only a single hydrogen bond,4 recognition
of each base pair as a unit,5 or being able to bind a purines in
either strand.6 However, a variety of targeting approaches7-9 have
resulted in only moderate successes.
A Janus-Wedge (J-W) triple helix, a recognition motif first
suggested by Lehn10 in his work with heterocycles, involves the
ability of the incoming third strand to hydrogen bond with the
Watson-Crick (W-C) faces of the two target strands (see Figure
1a). Two central questions can be posed about this recognition
motif: (i) What is the stability of a triplex based upon the triplet
structure illustrated in Figure 1a? (ii) Can a J-W triplex be formed
by strand invasion of a native W-C duplex? Here we answer the
first of those questions.
To examine the formation and stability of a J-W triplex we pre-
pared an 8-mer T + C target site bracketed with 11 Watson-Crick
(W-C) base pairs on either side (Figure 1b). The two pyrimidine
bases should permit binding by the Janus-Wedge base (W) to both
W-C faces forming a series of base triplets such as that illustrated
in Figure 1a. A corresponding triplet involving a purine-pyrimidine
pair would exhibit a slightly different geometry. In this initial study,
formation of the J-W structure does not need to compete with W-C
base pairing. For the W residue we used 6-amino-pseudocytidine
attached to a peptide nucleic acid (PNA) backbone. The peptide
backbone was chosen for two reasons: (i) It is neutral and limits
charge-charge repulsion effects. (ii) PNA sequences can bind DNA
duplexes by strand invasion,11 a process that should be advantageous
for the targeting of native W-C duplexes.
Figure 1. (a) A Janus-Wedge base triplet: the third-strand residue W binds
to the W-C faces of both target residues, (b) the 11dC811-11T811 target
sequence.
To probe for complex formation, the W8K oligomer was added
to each of the single-stranded target strands as well as to the duplex
itself, and the mixtures were analyzed by nondenaturing PAGE
(Figure 2). The two DNA single strands are found in Lanes 5 and
6, while in Lanes 1 and 2 one equivalent of the W8K strand has
been added to each. The T8-target strand is more effectively shifted
to a lower mobility species (Lane 2) than is the dC8-target strand
(Lane 1) suggesting more effective binding to the former. Mixing
the 11T811 and 11dC811 strands resulted in a lower-mobility species
(Lane 4) interpreted to be the duplex of Figure 1b.
Addition of one equivalent of the W8K strand to the duplex target
resulted in complete shift of the duplex band to a lower-mobility
species that we interpret as being the J-W complex (Lane 3).
The gel shift experiment indicates that complex formation occurs,
but it cannot clarify the nature of binding. It appears that the W8K
strand binds more effectively to the T8 strand than the dC8 strand
(compare Lanes 1 and 2 of Figure 2), and in the triplex it could, in
principle, bind only to the T8 strand. We performed a chemical
probing experiment with bromide and monoperoxysulfite known12
to differentiate between single-stranded and double-stranded dC
To prepare the J-W strand, an acetoxy linker was attached to
the 5-position of the readily available 2,6-diaminopyrimidin-4-one.
The para-exocyclic amino group was protected as the CBZ
derivative; the corresponding ortho-amino group proved to be
largely unreactive. The 8-mer synthesis started with an L-lysine
(K) residue to assist the aqueous solubility of the PNA oligomer
as well as to provide some complementary charge-charge interac-
tions between the PNA and target sequence. The product W8K
oligomer was purified by HPLC and identified by MALDI-TOF
mass spectrometry.
Figure 2. Gel shift assay for W8K binding to the ss and ds 30-mers of
Figure 1b. (Lane 1) 11dC811 + W8K. (Lane 2) 11T811 + W8K. (Lane 3)
duplex + W8K. (Lane 4) 11dC811 + 11T811 duplex. (Lane 5) 11T811.
(Lane 6) 11dC811.
9
70
J. AM. CHEM. SOC. 2004, 126, 70-71
10.1021/ja038081x CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Thermodynamic Parameters for Complex Formationa
residues. Treatment of the duplex target with the bromide reagent
resulted in cleavage at essentially all of the dC residues of the dC8
target, but not those present in the duplex regions. After addition
of the W8K strand to the duplex, a second chemical probing
experiment indicated that none of the dC residues were susceptible
to cleavage (Supporting Information). This experiment indicates
that the W8K strand interacts with the dC8-target sequence as well
as with the T8-target sequence.
dC −W K
T −W K
WK triplex
8
8
8
8
8
∆H° (kcal/mol)
∆S° (cal/mol °K)
∆G°25 (kcal/mol)
-83.8 ( 0.5
-256 ( 2
-4.5 ( 0.4
-214 ( 2
-95.7 ( 1.9
-270 ( 6
-7.35 ( 0.07
-10.5 ( 0.1
-15.2 ( 0.3
19
KD
4.2 × 10-6
M
2.0 × 10-8
M
7.4 × 10-12
M
a Standard deviations reflect the reproducibilities of the van’t Hoff
analyses.
The orientation of the N- and C-termini of the W8K strand in
the three-stranded complex was determined by tethering a phenan-
throline residue to the N-terminus of the W8K strand. After addition
of CuSO4 and a thiol this reagent results in cleavage of DNA13 in
the vicinity of the Cu-phenanthroline. The results of this experiment
(Supporting Information) indicated that binding occurs with the
N-terminus oriented toward the 5′-terminus of the 11dC811 strand
(and toward the 3′-terminus of the 11T811 strand).
consistent with the gel shift data of Figure 2, which indicates that
formation of the dC8-W8K duplex is less favored than is the
formation of the T8-W8K duplex. The thermodynamic data indicate
differences in stabilities (∆∆G ∼ 3 kcal/mol). This difference may
reflect strand polarity. To form three interresidue H-bonds in the
dC8-W8K duplex, the dC8 strand must adopt the polarity that
appears to be less preferred (with the amino terminus of the W8K
strand oriented toward the 5′ terminus of the dC8 strand). The
formation of the triplex occurs with a ∆G°25 of -15.2 kcal/mol, a
value that is very similar to the sum of the ∆G°25 values for both
duplexes (-17.85 kcal/mol). This thermodynamic analysis confirms
that triplex formation by W8K occurs with binding to both of the
DNA target strands. Finally, the magnitude of the stabilization that
accompanies J-W triplex formation (absent competing W-C pairing)
is substantially greater than conventional DNA triplexes,16,17 or
hairpin triplexes18 of the same or in some cases longer-sequence
lengths.
The proposed J-W complex results from the W8K strand entering
the double-stranded complex through the major groove of the target
base pairs (see Figure 1a). An alternative binding mode would have
it enter through the minor groove. Since this initial study uses a
homopolymer and the acceptor/donor hydrogen bonding pattern for
the W-T interaction is symmetrical, the base pairing orientation
must be discerned on the basis of the interactions with the dC strand.
We prepared shorter (14-mer) target strands containing either a
central dC8 or d(C2isomC4C2) sequence. 5-Methyl-isocytidine
(isomC) should form three stabilizing hydrogen bonds with the W
residue if it enters from the minor groove, but should be destabilized
if oriented to enter from the major groove (Figure 3). We could
not prepare the fully substituted isomC8 due to acid depyrimidination
during assembly, although earlier work14 reported on the synthesis
of isomC10 sequences. While a mixture of the 8-mer dC8 the W8K
strand resulted in clearly definable A260 vs temperature transition,
and a Tm value of 23.3 °C,15 the d(C2isomC4C2) sequence gave no
discernible transition, suggesting destabilizing interactions between
the W and isomC residues. This experiment suggests that the type
of triplex formed is similar to that illustrated in Figure 1a with the
J-W strand entering from the major groove.
Experiments related to the second question regarding the ability
of related sequences to undergo strand invasion are presently under
study. The disruption of a Watson-Crick dA8‚dT8 target will require
an energetic cost of ∼7.1 kcal/mol20 with a remaining ∼8 kcal/
mol of stabilization energy. Effective strand invasion will likely
require sequences longer than 8-mers.
Acknowledgment. This work was supported by a grant from
the NIH (GM53201).
Supporting Information Available: Synthetic schemes and pro-
cedures, and a variety of gel and thermodynamic analyses (PDF). This
material is available free of charge via the Internet at http://pubs.acs.org.
Thermal denaturation of the J-W triplex formed from the duplex
target of Figure 1b and the W8K strand resulted in two transitions
(Supporting Information). The midpoint of the first transition
occurred near 43 °C, while the midpoint of the second occurred
near 70 °C. The latter transition was present in the absence of the
W8K strand and is interpreted and reflecting the helix-to-coil
transition for the duplex target. The early transition then reflects
the J-W triplex to DNA duplex + W8K transition. Since these
transitions were clearly separated, it was possible to perform a van’t
Hoff analysis on a series of transitions obtained at varying total
concentrations and obtain thermodynamic parameters characterizing
the J-W triplex (Table 1).
We examined three different transitions involving the two 8-mer
duplexes dC8-W8K, T8-W8K, and the 11dC811-11T811-W8K
triplex. For the three complexes, ∆G values of -7.35, -10.5, and
-15.2 kcal/mol, respectively, were obtained. The ∆G°25 values for
the two duplexes and the corresponding calculated KD values are
References
(1) Moser, H. E.; Dervan, P. B. Science 1987, 238, 645-650.
(2) Francois, J. C.; Saison-Behmoaras, T.; Helene, C. Nucleic Acids Res. 1988,
16, 11431-11440.
(3) Cooney, M.; Czernuszewicz, G.; Postel, E. H.; Flint, S. J.; Hogan, M. E.
Science 1988, 241, 456-459.
(4) Griffin, L. C.; Dervan, P. B. Science 1989, 245, 967-971.
(5) Gianolio, D. A.; McLaughlin, L. W. Nucleosides Nucleotides 1999, 18,
1751-1769.
(6) Li, J.-S.; Fan, Y.-H.; Zhang, Y.; Marky, L. A.; Gold, B. J. Am. Chem.
Soc. 2003, 125, 2084-2093.
(7) Horne, D. A.; Dervan, P. B. J. Am. Chem. Soc. 1990, 112, 2435-2437.
(8) Jayasena, S. D.; Johnston, B. H. Nucleic Acids Res. 1992, 20, 5279-
5288.
(9) Gowers, D. M.; Fox, K. R. Nucleic Acids Res. 1999, 27, 1569-1577.
(10) Branda, N.; Kurz, G.; Lehn, J. M. Chem. Commun. 1996, 2443-2444.
(11) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science 1991, 254,
1497-1500.
(12) Ross, S. A.; Burrows, C. J. Nucleic Acids Res. 1996, 24, 5062-5063.
(13) Chen, C. H. B.; Sigman, D. S. J. Am. Chem. Soc. 1988, 110, 6570-
6572.
(14) Jurczyk, S.; Kodra, J. R.; Rozzell, J. D.; Benner, S. A.; Battersby, T. R.
HelV. Chim. Acta 1998, 81, 793-811.
(15) The Tm for the corresponding T8 duplex was 34.6 °C.
(16) Pilch, D. S.; Brousseau, R.; Shafer, R. H. Nucleic Acids Res. 1990, 18,
5743-5750.
(17) Singleton, S. F.; Dervan, P. B. J. Am. Chem. Soc. 1992, 114, 6957-
6965.
(18) Bervers, S.; Schutte, S.; McLaughlin, L. W. J. Am. Chem. Soc. 2000,
122, 5905-5915.
(19) KD values estimated from the measured ∆G°25 values.
(20) SantaLucia, J., Jr.; Allawi, H. T.; Seneviratne, P. A. Biochemistry 1996,
35, 3555-3562.
Figure 3. Destabilizing/missing H-bond interactions between W and isomC.
JA038081X
9
J. AM. CHEM. SOC. VOL. 126, NO. 1, 2004 71