Hydroxybenzamide/pyrrole pair distinguishes T?A from A?T base pairs in the minor groove of DNA

Article abstract of DOI:10.1021/ja001692u

A new aromatic pair, 2-hydroxy-6-methoxybenzamide/1-methylpyrrole at the terminal position of hairpin polyamides has been designed for distinguishing T?A from A?T base pairs and both from G?C/C?G in the minor groove of DNA. Four eight-ring hairpin polyami

Full text of DOI:10.1021/ja001692u

9354
J. Am. Chem. Soc. 2000, 122, 9354-9360
Articles
Hydroxybenzamide/Pyrrole Pair Distinguishes T‚A from A‚T Base
Pairs in the Minor Groove of DNA
Ulf Ellervik, Clay C. C. Wang, and Peter B. Dervan*
Contribution from the DiVision of Chemistry and Chemical Engineering,
California Institute of Technology, Pasadena, California, 91125
ReceiVed May 16, 2000
Abstract: A new aromatic pair, 2-hydroxy-6-methoxybenzamide/1-methylpyrrole at the terminal position of
hairpin polyamides has been designed for distinguishing T‚A from A‚T base pairs and both from G‚C/C‚G in
the minor groove of DNA. Four eight-ring hairpin polyamides with benzamide (Bz), 2-hydroxybenzamide
(Hb-1), 2-hydroxy-6-methylbenzamide (Hb-2), and 2-hydroxy-6-methoxybenzamide (Hb-3) at the N-terminal
position were synthesized. The equilibrium association constants (Ka) were determined at four DNA sites
which differ at a single common position, 5′-TNTACA-3′ (N ) T, A, G, C). Quantitative DNase I footprint
titration experiments reveal that (Hb-3)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (4) bound the sequences 5′-TTTACA-
3′ and 5′-TATACA-3′ with high affinity; Ka ) 2.6 × 1010 M-1 and Ka ) 8.4 × 109 M-1, respectively, a 3-fold
specificity for T vs A was found. Importantly, the sequences 5′-TGTACA-3′ and 5′-TCTACA-3′ are bound
with 50 and 200 times lower affinity, revealing an overall specificity of Hb-3/Py of T > A . G > C. These
results expand the repertoire of sequences targetable by hairpin polyamides.
Introduction
selectivity (see Supporting Information). One could imagine that
unfavorable conformations at the terminal position due to
rotation of the terminal Hp to form an intramolecular hydrogen
bond between the 3-hydroxyl hydrogen and the carboxamide
oxygen would orient the key hydroxyl recognition element away
from the minor groove (Figure 2). This suggests that Hp/Py
pairs may be limited within the hairpin motif to internal
positions for T‚A discrimination and there is a need for the
design of new aromatic pairs at the terminal position for
distinguishing T‚A from A‚T.
During the early development of minor groove binders, it
was shown that six-membered aromatic rings such as 2-pyridine,
as well as the five-membered ring imidazole Im, when paired
with Py could target G‚C base pairs.4 The question arises
whether 2-hydroxybenzene (Hb) at the N-terminal position of
hairpin polyamides could act as a mimic of the 2-hydroxypyrrole
(Hp) with improved affinity and specificity for T. We describe
here a series of eight-ring polyamide hairpins with the general
sequence XPyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp and the N-ter-
minal position X ) benzamide (Bz), 2-hydroxybenzamide (Hb-
1), 2-hydroxy-6-methylbenzamide (Hb-2), and 2-hydroxy-6-
methoxybenzamide (Hb-3). Since Py is in the eighth position,
the terminal pairs in the hairpin conformation are Hb-1/Py, Hb-
2/Py, and Hb-3/Py (Figure 3). For controls, the parent benzamide
without the hydroxy is included. Four eight-ring hairpin
polyamides differing in the terminal position were synthesized
by solid-phase methods (Figure 4). The plasmid pCW15 was
designed to contain four six-base-pair recognition sites 5-T-
NTACA-3′ (where N ) T,A,G,C) which differ at a single
common position allowing for comparison of the affinities
The ability to distinguish T‚A from A‚T base pairs by
synthetic ligands which bind in the minor groove of DNA is an
essential milestone for targeting specific A/T rich sequences
within promoters for gene regulation studies.1,2 We have shown
that hydroxypyrrole/pyrrole (Hp/Py) pairs when placed internal
in a hairpin distinguish T‚A from A‚T.2 However, this has not
been demonstrated for the terminal position in hairpin poly-
amides where “end effects” would potentially alter molecular
interactions necessary for specificity.3 For example, the sub-
stituent at the 4-position of the N-terminal five-membered ring
is flanked in the minor groove by â-alanine at the C-terminus.
To evaluate hairpin polyamides capable of distinguishing T‚A/
A‚T base pairs at the terminal position, a series of hairpins with
3-hydroxy-1-methylpyrrole substituted at the 4-position by H,
formamide, acetamide (Hp-1, Hp-2, Hp-3) were synthesized
(Figure 1). From quantitative footprint titrations we find that
polyamides with a terminal pyrrole containing a 3-hydroxyl and
(i) a hydrogen in the 4-position showed no selectivity for T‚A
over A‚T, (ii) an acetamide at the 4-position showed 3-fold
specificity for T‚A over A‚T but diminished affinity, and (iii)
a formamide at the 4-position showed higher affinity but no
(1) (a) Dervan, P. B.; Bu¨rli, R. W. Curr. Opin. Chem. Biol. 1999, 3,
688-693. (b) Praseuth, D.; Guieysse, A. L.; Helene, C. Biochim. Biophys.
Acta 1999, 1489, 181-206. (c) Bailly, C.; Chaires, J. B. Bioconjugate Chem.
1998, 9, 513-538.
(2) (a) Kielkopf, C. L.; White, S.; Szewczyk, J. W.; Turner, J. M.; Baird,
E. E.; Dervan, P. B.; Rees, D. C. Science (Washington, D.C.) 1998, 282,
111-115. (b) Kielkopf, C. L.; Bremer, R. B.; White, S.; Szewczyk, J. W.;
Turner, J. M.; Baird, E. E.; Dervan, P. B.; Rees, D. C. J. Mol. Biol. 2000,
295, 557-567.
(3) Hawkins, C. A.; de Clairac, R. P.; Dominey, R. N.; Baird, E. E.;
White, S.; Dervan, P. B.; Wemmer, D. E. J. Am. Chem. Soc. 2000, 122,
5235-5243.
(4) Wade, W. S.; Mrksich, M.; Dervan, P. B. J. Am. Chem. Soc. 1992,
114, 8783-8794.
10.1021/ja001692u CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/15/2000

Aromatic Pair Distinguishes Base Pairs
J. Am. Chem. Soc., Vol. 122, No. 39, 2000 9355
Figure 2. Binding model for the complex formed between polyamide
(Hp-1)PyPyPy-(R)H Nγ-ImPyPyPy-â-Dp and a 5′-TTTACA-3′ se-
2
quence.
Quantitative DNase I Footprinting Titrations. Quantitative
DNase I footprint titrations on the 3′-32P-end-labeled 284 bp
pCW15 EcoRI/PVuII restriction fragment (10 mM Tris‚HCl, 10
mM KCl, 10 mM MgCl2, and 5 mM CaCl2, pH 7.0, 22 °C)
(Figure 7) reveal that BzPyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (1)
binds the 5′-TTTACA-3′and 5′-TATACA-3′ sites with com-
parable equilibration association constants, Ka ) 1.9 × 1011
M-1 and Ka ) 1.2 × 1011 M-1, respectively. Polyamide 1 binds
the two sites 5′-TGTACA-3′ and 5′-TCTACA-3′ with 5- and
10-fold lower affinity, respectively, revealing that the Bz/Py
pair is specific for both T‚A and A‚T over G‚C and C‚G.
Three polyamides containing 2-hydroxybenzamides at the
N-terminal position, (Hb-1)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp
(2), (Hb-2)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (3), (Hb-3)Py-
PyPy-(R)H2Nγ-ImPyPyPy-â-Dp (4), were tested for their ability
to discriminate 5′-TTTACA-3′ vs 5′-TATACA-3′. (Hb-1)-
PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (2), containing a single hy-
droxyl, bound both 5′-TTTACA-3′ and 5′-TATACA-3′ with the
same affinity (Ka ) 1.9 × 1010 M-1). (Table 1)
Figure 1. Hydrogen-bonding model of the polyamide/DNA complex
between eight-ring hairpin polyamides XPyPyPy-(R)H Nγ-ImPyPyPy-
2
â-Dp (X ) substituted Py or Hp) with 5′TTTACA-3′ sites. A circle
with two dots represents the lone pairs of N3 of purines and O2 of
cytosine. Two touching circles with dots represent the two lone pairs
of the O2 of thymine. Circles containing an H represent the exocyclic
amino hydrogen of guanine. Putative hydrogen bonds are illustrated
by dotted lines.
between different terminal pairs and the four Watson Crick base
pairs in the minor groove of DNA (Figure 5).
(Hb-2)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (3), with a 2-hy-
droxyl and a 6-methyl substitution also bound the two sites with
equal affinity. Discrimination of 5′-TTTACA-3′ from 5′-
TATACA-3′ was achieved by replacing the 6-methyl group in
polyamide 5 with 6-methoxy. (Hb-3)PyPyPy-(R)H2Nγ-ImPy-
PyPy-â-Dp (4) bound 5′-TTTACA-3′ with an affinity of Ka )
1.9 × 1010 M-1 and 5′-TATACA-3′ with 3-fold lower affinity.
Importantly, polyamide 4 bound to 5′-TGTACA-3′ and to 5′-
TCTACA-3′ with 55 and 200 times lower affinity, respectively.
Results
Polyamide Synthesis. Boc-PyPyPy-(R)Fmocγ-ImPyPyPy-â-
PAM-resin was synthesized in a stepwise manner from Boc-
â-alanine-PAM resin (0.55 mmol/g) using solid-phase meth-
odology5 in 16 steps (Figure 6). Sterically hindered acids such
as Hb-2 and Hb-3 were coupled as acid chlorides. A sample of
the resin was then cleaved by aminolysis with ((dimethyl)-
amino)propylamine (55 °C, 16 h) and purified by reversed-phase
HPLC to provide BzPyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (1).
Methyl-protected polyamides 2-4 were subsequently depro-
tected by treatment with sodium ethanethiolate in DMF (100
°C, 1 h) and purified by reversed-phase HPLC to afford (Hb-
1)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (2), (Hb-2)PyPyPy-(R)H2Nγ-
ImPyPyPy-â-Dp (3), and (Hb-3)PyPyPy-(R)H2Nγ-ImPyPyPy-
â-Dp (4) (Figure 4).
Discussion
BzPyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (1) demonstrates that a
simple benzene ring is able to pair with a 1-methylpyrrole ring
(Bz/Py) to distinguish A‚T/T‚A base pairs from G‚C/C‚G base
pairs in the minor groove of DNA. The Bz/Py pair is a higher
affinity binder than the Py/Py pair at the terminal position.
Interestingly, the Hb-1/Py pair with a single hydroxyl group
did not discriminate T from A. The hydroxyl group is likely in
(5) Baird, E. E.; Dervan, P. B. J. Am. Chem. Soc. 1996, 118, 6141-
6146.

9356 J. Am. Chem. Soc., Vol. 122, No. 39, 2000
EllerVik et al.
Figure 4. Structures of polyamides BzPyPyPy-(R)H Nγ-ImPyPyPy-
2
â-Dp (1), (Hb-1)PyPyPy-(R)H Nγ-ImPyPyPy-â-Dp (2), (Hb-2)PyPyPy-
2
(R)H Nγ-ImPyPyPy-â-Dp (3), (Hb-3)PyPyPy-(R)H Nγ-ImPyPyPy-â-Dp
2
2
(4).
Experimental Section
Figure 3. Hydrogen-bonding model of the hairpin polyamide/DNA
containing substituted benzamides with the 5′TTTACA-site. A circle
with two dots represents the lone pairs of N3 of purines and O2 of
cytosine. Two touching circles with dots represent the two lone pairs
of the O2 of thymine. Circles containing an H represent the exocyclic
amino hydrogen of guanine. Putative hydrogen bonds are illustrated
by dotted lines.
Boc-â-alanine-(-4-carboxamidomethyl)-benzyl-ester-copoly(styrene-
divinylbenzene) resin (Boc-â-PAM-Resin), N,N′-Dicyclohexylcarbo-
diimide (DCC), N-Hydroxybenzotriazole (HOBt), 2-(1H-Benzotriazole-
1-yl)-1,1,3,3-tetramethyluronium hexa-fluorophosphate (HBTU), and
Boc-â-alanine were purchased from Peptides International. N,N-
diisopropylethylamine (DIEA) and N,N-dimethylformamide (DMF)
were purchased from Applied Biosystems. DMF was distilled under
reduced pressure prior to synthesis. (R)-2-Fmoc-4-Boc-diaminobutyric
acid was from Bachem, dichloromethane (DCM) was reagent grade
from EM, thiophenol (PhSH) and dimethylaminopropylamine were from
Aldrich, and trifluoroacetic acid (TFA) was from Halocarbon. 1H NMR
were recorded on a Varian Mercury 300 instrument. Chemical shifts
are reported in parts-per-million downfield from the signal for Me4Si,
with reference to the solvent residual signal. UV spectra were measured
on a Hewlett-Packard model 8452A diode array spectrophotometer.
Matrix-assisted, laser desorption/ionization time-of-flight mass spec-
trometry was carried out at the Protein and Peptide Micro-
analytical Facility at the California Institute of Technology. HPLC
analysis was performed on a Beckman Gold system using a RAINEN
C18, Microsorb MV, 5-µm, 300 × 4.6 mm reversed-phase column in
0.1% (w/v) TFA with acetonitrile as eluent and a flow rate of 1.0 mL/
min, gradient elution 1.25% acetonitrile/min. Preparatory HPLC was
carried out on a Beckman HPLC using a Waters DeltaPak 25 × 100
mm, 100-µm C18 column, 0.1% (w/v) TFA, 0.25% acetonitrile/min.
18MΩ water was obtained from a Millipore MilliQ water purification
system, and all buffers were 0.2-µm filtered. Reagent-grade chemicals
were used unless otherwise stated.
the wrong conformation, oriented away from the floor of the
minor groove and forming a hydrogen bond with the proximal
carbonyl group. To force the hydroxyl group toward the minor
groove of the DNA by an unfavorable intramolecular steric
interaction, two 2-hydroxybenzamides substituted at the 6
position with methyl and methoxy were examined. Quantitative
footprint titration reveals that T versus A discrimination can be
achieved with the 6-methoxy but not the 6-methyl derivative.
These results suggest that the 6-methyl substituent projecting
to the floor of the helix is not large enough to prevent the
hydroxyl group from hydrogen bonding to the carbonyl group.
Presumably, the steric bulk of the 6-methoxy group favors the
rotamer with the hydroxyl group oriented toward the minor
groove (Figure 8). Polyamide 4 with the Hb-3/Py pair binds
5′-TTTACA-3′ and 5′-TATACA-3′ with binding affinities of
Ka ) 2.6 × 1010 M-1 and Ka ) 8.4 × 109 M-1, respectively.
Although the 3-fold T versus A selectivity is similar to the Hp/
Py pair, the overall affinity is significantly higher, and most
importantly, the selectivity of A/T vs G/C improved. The new
hydroxybenzamide/pyrrole pair (Hb/Py) is a step toward hairpin
polyamides capable of differentiating pure A‚T tract sequences
important in gene regulation studies
Boc-PyPyPy-(R)Fmocγ-ImPyPyPy-â-PAM-resin. Boc-PyPyPy-(R)F-
mocγ-ImPyPyPy-â-PAM-resin (0.335 mmol/g) was synthesized in a
stepwise fashion from 0.55 mmol/g Boc-â-PAM-resin by manual solid-
phase methods.5 (R)-2-Fmoc-4-Boc-diaminobutyric acid was incorpo-
rated as previously described for Boc-γ-aminobutyric acid.5,6
(6) Herman, D. M.; Baird, E. E.; Dervan, P. B. J. Am. Chem. Soc. 1998,
120, 1382-1391.

Aromatic Pair Distinguishes Base Pairs
J. Am. Chem. Soc., Vol. 122, No. 39, 2000 9357
Figure 5. Two hundred eighty-four base pair EcoRI/PVuII restriction fragment derived from plasmid pCW15. The targeted six-base-pair recognition
sites are shown in boxes.
Figure 6. Solid-phase synthetic scheme for (Hb-3)PyPyPy-(R)H Nγ-ImPyPyPy-â-Dp (4) starting from commercially available Boc-â-PAM-resin:
2
(i) 80% TFA/DCM, 0.4 M PhSH; (ii) Boc-Py-OBt, DIEA, DMF; (iii) 80% TFA/DCM, 0.4 M PhSH; (iv) Boc-Py-OBt, DIEA, DMF; (v) 80%
TFA/DCM, 0.4 M PhSH; (vi) Boc-Py-OBt, DIEA, DMF; (vii) 80% TFA/DCM, 0.4 M PhSH; (viii) Boc-Im-acid, DCC, HOBt, DIEA, DMF; (ix)
80% TFA/DCM, 0.4 M PhSH; (x) Fmoc-R-Boc-γ-diaminobutyric acid, HBTU, DIEA, DMF; (xi) 80% TFA/DCM, 0.4 M PhSH; (xii) Boc-Py-OBt,
DIEA, DMF; (xiii) 80% TFA/DCM, 0.4 M PhSH; (xiv) Boc-Py-OBt, DIEA, DMF; (xv) 80% TFA/DCM, 0.4 M PhSH; (xvi) Boc-Py-OBt, DIEA,
DMF; (xvii) 80% TFA/DCM, 0.4 M PhSH; (xviii) 2,6-Dimethoxybenzoic acid, DMF, DCM, oxalyl chloride, DIEA, r.t. 3 h; (xix) Piperidine/DMF
3:1; (xx) (Dimethylamino)propylamine, 55 °C, 16 h;(xxi) NaH, EtSH, DMF, 100 °C, 1h.
Procedure for Cleavage from the Resin. After the coupling was
completed, the resin was filtered off the reaction and washed with DMF
(2 × 30 s). DMF (1 mL) and piperidine (3 mL) were added and the
mixture shaken for 30 min at 22 °C. The resin was filtered off and
washed with DMF (2 × 30 s), DCM (3 × 30 s), MeOH (1 × 30 s),
Et2O (1 × 30 s) and dried in vacuo. The resin was treated with
(dimethylamino)propylamine (1 mL) with periodic agitation at 55 °C
for 16 h. The reaction mixture was then filtered to remove the resin,
TFA (0.1% (w/v), 7 mL) added, and the resulting solution purified by
reversed-phase HPLC. The pure compounds were recovered as white
powders upon lyophilization of the appropriate fractions.
to 0 °C. The mixture was added to HOAc (3 mL), and all volatiles
were removed (high vacuum, 40 °C). The residue was dissolved in
CH3CN (1 mL) and TFA (7 mL, 0.1% (w/v)) and purified by
preparative HPLC.
BzPyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (1). Benzoic acid (70 mg,
0.57 mmol) was dissolved in DMF (1.0 mL), and HBTU (75 mg, 0.20
mmol) and DIEA (0.5 mL) were added. The mixture was allowed to
stir for 10 min and added to the deprotected Boc-PyPyPy-(R)Fmocγ-
ImPyPyPy-â-PAM-resin (100 mg). The mixture was shaken for 5.0 h
at 22 °C, and the polyamide was cleaved from the resin according to
the general procedure. Preparative HPLC gave 1 (16 mg, 39% isolated
yield): UV λmax (H2O) 240, 312 (69500); 1H NMR (DMSO-d6) δ 11.10,
10.42, 10.20, 10.07, 10.04, 10.02, 9.97 (s, 1H each), 9.41 (bs, 1 H),
8.42 (bs, 2 H), 8.26 (t, 1 H, J ) 5.7 Hz), 8.12 (q, 2 H, J ) 5.6 Hz),
8.00 (dd, 2 H, J ) 6.3, 1.5 Hz), 7.56-7.70 (m, 3 H), 7.40 (d, 1 H, J
) 1.5 Hz), 7.33 (d, 1 H, J ) 1.5 Hz), 7.30 (d, 1 H, J ) 1.8 Hz), 7.28
(d, 1 H, J ) 1.5 Hz), 7.26 (d, 1 H, J ) 2.1 Hz), 7.25 (d, 1 H, J ) 2.1
Deprotection of Methyl-Protected Polyamides. To a slurry of NaH
(80 mg, 60% in mineral oil) in DMF (0.50 mL) was added a solution
of ethanethiol (0.32 mL) in DMF (0.50 mL), and the mixture was heated
to 100 °C for 5 min in a sealed tube. The polyamide, dissolved in
DMF (1.00 mL), was added to the ethanethiolate solution, and the
mixture was heated to 100 °C for 1 h in a sealed tube and then cooled

9358 J. Am. Chem. Soc., Vol. 122, No. 39, 2000
EllerVik et al.
Figure 7. (a) Quantitative DNase I footprint titration experiment with BzPyPyPy-(R)H Nγ-ImPyPyPy-â-Dp (1) on the 284 bp EcoRI/PVuII restriction
2
fragment from plasmid pCW15: lane 1, intact DNA; lane 2, A-specific reaction; lane 3, G-specific reaction; lane 4, DNase I standard; lanes 5-21,
0.1 pM, 0.2 pM, 0.5 pM, 1 pM, 2 pM, 5 pM, 10 pM, 15 pM, 25 pM, 40 pM, 65 pM, 100 pM, 150 pM, 250 pM, 450 pM, 650 pM, 1 nM
BzPyPyPy-(R)H Nγ-ImPyPyPy-â-Dp (1). The four sites 5′-TTTACA-3′, 5′-TATACA-3′, 5′-TGTACA-3′, and 5′-TCTACA-3′ that were analyzed
2
are shown on the right side of the gel. (b) Quantitative DNase I footprint titration experiment with (Hb-3)PyPyPy-(R)H Nγ-ImPyPyPy-â-Dp (4) on
2
the 284 bp EcoRI/PVuII restriction fragment from plasmid pCW15: lane 1, intact DNA; lane 2, A-specific reaction; lane-3, G-specific reaction;
lane 4, DNase I standard; lanes 5-21, 10 pM, 20 pM, 50 pM, 100 pM, 150 pM, 250 pM, 400 pM, 650 pM, 1 nM, 1.5 nM, 2.5 nM, 4 nM, 6.5 nM,
10 nM, 20 nM, 50 nM, 100 nM (Hb-3)PyPyPy-(R)H Nγ-ImPyPyPy-â-Dp (7). The four sites 5′-TTTACA-3′, 5′-TATACA-3′, 5′-TGTACA-3′, and
2
5′-TCTACA-3′ that were analyzed are shown on the right side of the gel.
a
Table 1. Equilibration Association Constants (M-1
)
polyamide
5′-TTTACA-3′
5′-TATACA-3′
5′-TGTACA-3′
5′-TCTACA-3′
1
2
3
4
1.9 ((0.3) × 1011
1.9 ((0.1) × 1010
1.0 ((0.2) × 1010
2.6 ((0.9) × 1010
1.2 ((0.1) × 1011
1.9 ((0.3) × 1010
1.0 ((0.1) × 1010
8.4 ((1.0) × 109
3.5 ((1.0) × 1010
6.3 ((1.7) × 109
1.5 ((0.6) × 109
4.7 ((0.5) × 108
1.7 ((0.7) × 1010
3.7 ((0.6) × 109
e1.0 × 108
e1.0 × 108
a Values reported are mean values from at least three DNase I footprint titration experiments, with the standard deviation for each data set
indicated in parentheses. The assays were perfomed at 22 °C at pH 7.0 in the presence of 10 mM Tris‚HCl, 10 mM KCl, 10 mM MgCl2, and 5 mM
CaCl2.
Hz), 7.21 (d, 1 H, J ) 2.1 Hz), 7.18 (d, 1 H, J ) 1.8 Hz), 7.15 (d, 2
H, J ) 1.5 Hz), 7.05 (d, 1 H, J ) 2.1 Hz), 6.95 (d, 1 H, J ) 1.5 Hz),
4.05, 3.94, 3.93, 3.92, 3.91, 3.88, 3.87 (s, 3 H each), 3.30-3.50 (m, 4
H), 3.17 (q, 2 H, J ) 6.3 Hz), 3.06 (p, 2 H, J ) 4.8 Hz), 2.81 (d, 6 H,
J ) 4.8 Hz), 2.42 (t, 2 H, J ) 7.2 Hz), 2.00-2.10 (m, 2H), 1.81 (p, 2
H, J ) 7.5 Hz); MALDI-TOF-MS calcd. for C60H72N20O10 (M + H):
1233.6. Found: 1233.7.
resin according to the general procedure. Preparative HPLC gave
methyl-protected (Hb-1)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (15 mg, 35%
1
isolated yield): UV λmax (H2O) 240, 312 (69500); H NMR (DMSO-
d6) δ 11.10, 10.19, 10.07 (s, 1 H each), 10.03 (s, 2 H), 10.01, 9.97 (s,
1 H each), 8.38-8.50 (bs, 2 H), 8.26 (t, 1 H, J ) 6.0 Hz), 8.12 (q, 2
H, J ) 6.0 Hz), 7.72 (d, 1 H, J ) 1.8 Hz), 7.69 (d, 1 H, J ) 1.8 Hz),
7.61 (s, 1 H), 7.58 (d, 1 H, J ) 2.1 Hz), 7.55 (bs, 1 H), 7.52 (d, 1 H,
J ) 1.8 Hz), 7.38 (d, 1 H, J ) 1.8 Hz), 7.33 (d, 1 H, J ) 1.5 Hz), 7.29
(d, 1 H, J ) 1.5 Hz), 7.28 (d, 1 H, J ) 1.5 Hz), 7.25 (s, 2 H), 7.22 (s,
1 H), 7.10-7.20 (m, 4 H), 7.05 (d, 1 H, J ) 1.5 Hz), 6.95 (d, 1 H, J
) 1.5 Hz), 4.05, 3.96, 3.93, 3.92, 3.91,3.90, 3.88, 3.87 (s, 3 H each),
3.30-3.50 (m, 4 H), 3.17 (q, 2 H, J ) 6.0 Hz), 3.07 (p, 2 H, J ) 5.7
Hz), 2.80 (d, 6 H, J ) 4.8 Hz), 2.42 (t, 2 H, J ) 7.2 Hz), 2.06 (bs, 2
(Hb-1)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (2). 2-Methoxybenzoic
acid (70 mg, 0.46 mmol) was dissolved in DMF (1.0 mL), and HBTU
(75 mg, 0.20 mmol) and DIEA (0.5 mL) were added. The mixture was
allowed to stir for 10 min and then was added to the deprotected Boc-
PyPyPy-(R)Fmocγ-ImPyPyPy-â-PAM-resin (100 mg). The mixture was
shaken for 4.5 h at 22 °C, and the polyamide was cleaved from the

Aromatic Pair Distinguishes Base Pairs
J. Am. Chem. Soc., Vol. 122, No. 39, 2000 9359
1.67 (p, 2 H, J ) 7.8 Hz); MALDI-TOF-MS calcd. for C60H72N20O11
(M + H): 1249.6. Found: 1249.7.
(Hb-2)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (3). 2-Methoxy-6-meth-
ylbenzoic acid (26 mg, 0.16 mmol) was dissolved in DMF/DCM (1:1,
0.2 mL) and added to a mixture of DMF (0.1 mL) and oxalyl chloride
(2M in DCM, 0.072 mL). The mixture was allowed to stir for 20 min,
and DMF (0.3 mL) and DIEA (0.2 mL) were added. The mixture was
then added to the deprotected Boc-PyPyPy-(R)Fmocγ-ImPyPyPy-â-PAM-
resin (50 mg) and shaken for 4.0 h at 22 °C. The polyamide was cleaved
from the resin according to the general procedure. Preparative HPLC
afforded methyl-protected (Hb-2)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (9
mg, 41% isolated yield): UV λmax (H2O) 240, 312 (69 500); 1H NMR
(DMSO-d6) δ 10.27, 10.19 (s, 1 H each), 10.03 (s, 2 H), 10.00, 9.97
(s, 1 H each), 8.26 (t, 1 H, J ) 6.0 Hz), 8.12 (q, 2 H, J ) 5.1 Hz), 7.61
(s, 1 H), 7.30-7.40 (m, 3 H), 7.29 (d, 1 H, J ) 1.5 Hz), 7.28 (d, 1 H,
J ) 2.4 Hz), 7.25 (bs, 2 H), 7.22 (d, 1 H, J ) 1.8 Hz), 7.15 (bs, 2 H),
7.04, 7.03, 7.00, 6.96 (s, 1 H each), 6.95 (d, 1 H, J ) 1.5 Hz), 6.93,
6.91 (s, 1 H each), 4.05 (2, 3 H), 3.92, 3.91 (s, 6 H each), 3.88, 3.87,
3.81 (s, 3 H each), 3.30-3.50 (m, 4 H), 3.17 (q, 2 H, J ) 6.0 Hz),
3.05 (bt, 2 H, J ) 7.6 Hz), 2.79 (s, 6H), 2.42 (t, 2 H, J ) 7.2 Hz), 2.27
(s, 3 H), 2.06 (bs, 2H), 1.80 (p, 2 H, J ) 8.1 Hz); MALDI-TOF-MS
calcd. for C62H76N20O11 (M+H): 1277.6. Found: 1277.7. The methyl-
protected (Hb-1)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (3.1 mg, 2.4 µmol)
was deprotected according to the general procedure to give 3 (1.4 mg,
48% isolated yield): UV λmax (H2O) 240, 312 (69500); 1H NMR
(DMSO-d6) δ 10.19, 10.18, 10.03, 10.02, 10.00, 9.96, 9.63 (s, 1 H
each), 8.25 (bt, 1 H, J ) 5.8 Hz), 8.12 (q, 2 H, J ) 3.4 Hz), 7.61 (s,
1 H), 7.32 (d, 1 H, J ) 1.7 Hz), 7.31 (d, 1 H, J ) 1.8 Hz), 7.29 (d, 1
H, J ) 2.1 Hz), 7.28 (d, 1 H, J ) 2.1 Hz), 7.24 (d, 2 H, J ) 1.8 Hz),
7.22 (d, 1 H, J ) 1.5 Hz), 7.10-7.20 (m, 3 H), 7.06 (d, 1 H, J ) 2.1
Hz), 7.04 (d, 1 H, J ) 1.8 Hz), 6.95 (d, 1 H, J ) 1.5 Hz), 6.78 (d, 1
H, J ) 8.1 Hz), 6.73 (d, 1 H, J ) 7.2 Hz), 4.05 (s, 3 H), 3.92, 3.91 (s,
6 H each), 3.88, 3.87 (s, 3 H each), 3.17 (q, 2 H, J ) 6.0 Hz), 3.04
(bs, 2 H), 2.78 (s, 6 H), 2.41 (t, 2 H, J ) 6.9 Hz), 2.25 (s, 3 H), 2.05
(bs, 2 H), 1.79 (p, 2 H, J ) 6.6 Hz); MALDI-TOF-MS calcd. for
C61H74N20O11 (M+H): 1263.6. Found: 1263.7.
(Hb-3)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (4). 2,6-Dimethoxyben-
zoic acid (28 mg, 0.16 mmol) was dissolved in DMF/DCM (1:1, 0.2
mL) and added to a mixture of DMF (0.1 mL) and oxalyl chloride
(2M in DCM, 0.072 mL). The mixture was allowed to stir for 20 min,
and DMF (0.3 mL) and DIEA (0.2 mL) were added. The mixture was
added to the deprotected Boc-PyPyPy-(R)Fmocγ-ImPyPyPy-â-PAM-resin
(50 mg) and shaken for 3.0 h at 22 °C. The polyamide was then cleaved
from the resin according to the general procedure. Preparative HPLC
gave methyl-protected (Hb-3)PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp (10 mg,
48% isolated yield): UV λmax (H2O) 240, 308 (69500); 1H NMR
(DMSO-d6) δ 10.96, 10.06, 10.03, 9.90, 9.88, 9.87, 9.83 (s, 1 H each),
8.26 (bs, 3 H), 8.12 (t, 1 H, J ) 6.2 Hz), 7.99 (q, 2 H, J ) 5.7 Hz),
7.48 (s, 1 H), 7.27 (t, 2 H, J ) 8.4 Hz), 7.20 (d, 1 H, J ) 1.6 Hz), 7.16
(d, 1 H, J ) 1.8 Hz), 7.14 (d, 1 H, J ) 2.1 Hz), 7.13 (d, 1 H; J ) 1.8
Hz), 7.11 (d, 1 H, J ) 1.5 Hz), 7.08 (d, 1 H, J ) 1.5 Hz), 7.02 (d, 1
H, J ) 1.9 Hz), 7.01 (d, 1 H, J ) 1.6 Hz), 6.90 (d, 1 H, J ) 1.5 Hz),
6.89 (d, 1 H, J ) 2.1 Hz), 6.82 (d, 1 H, J ) 1.8 Hz), 6.66, 6.63 (s, 1
H each), 3.92, 3.79, 3.78 (s, 3 H each), 3.78 (s, 6 H), 3.75, 3.74 (s, 3
H each), 3.67 (s, 6 H), 3.04 (q, 2 H, J ) 6.0 Hz), 2.94 (p, 2 H, J ) 4.8
Hz), 2.68 (d, 6 H, J ) 4.5 Hz), 2.28 (t, 2 H, J ) 6.9 Hz), 1.94 (bs, 2
H), 1.67 (p, 2 H, J ) 7.6 Hz); MALDI-TOF-MS calcd. for C62H76N20O12
(M+H): 1293.6; found: 1293.7. The methyl-protected (Hb-3)PyPyPy-
(R)H2Nγ-ImPyPyPy-â-Dp (2.7 mg, 2.1 µmol) was deprotected according
to the general procedure to give 4 (1.9 mg, 42% isolated yield) together
with (2,6-dihydroxybenzoyl)HN-PyPyPy-(R)H2Nγ-ImPyPyPy-â-Dp
Figure 8. Binding models for two rotamers of (Hb-3)PyPyPy-(R)H Nγ-
ImPyPyPy-â-Dp and a 5′-TTTACA-3′ sequence.
2
H), 1.80 (p, 2H, J ) 8.7 Hz); MALDI-TOF-MS calcd. for C61H74N20O11
(M+H): 1263.6. Found: 1263.7. The methyl-protected (Hb-1)PyPyPy-
(R)H2Nγ-ImPyPyPy-â-Dp (3.1 mg, 2.5 µmol) was deprotected according
to the general procedure to give 2 (0.5 mg, 16% isolated yield): UV
1
λmax (H2O) 240, 312 (69500); H NMR (DMSO-d6) δ 12.25, 10.96,
1
(1.3 mg, 28% isolated yield): UV λmax (H2O) 240, 310 (69500); H
10.43, 10.06, 9.92, 9.90, 9.88, 9.83 (s, 1 H each), 8.25 (bs, 3 H), 8.13
(t, 1 H, J ) 5.9 Hz), 7.98 (q, 2 H, J ) 6.0 Hz), 7.92 (dd, 1 H, J ) 8.4,
1.8 Hz), 7.48 (s, 1 H), 7.37 (ddd, 1 H, J ) 8.7, 7.2, 1.5 Hz), 7.28 (d,
1 H, J ) 1.5 Hz), 7.20 (d, 1 H, J ) 1.8 Hz), 7.17 (d, 1 H, J ) 1.5 Hz),
7.14 (d, 1 H, J ) 1.5 Hz), 7.11 (bs, 2 H), 7.08 (d, 1 H, J ) 1.5 Hz),
7.07 (d, 1 H, J ) 2.1 Hz), 7.02 (d, 2 H, J ) 1.5 Hz), 6.85-6.95 (m,
3 H), 6.82 (d, 1 H, J ) 1.5 Hz), 3.92, 3.82, 3.79, 3.78, 3.77, 3.75, 3.73
(s, 3 H each), 3.04 (q, 2 H, J ) 6.6 Hz), 2.93 (bt, 2 H, J ) 7.2 Hz),
2.68 (bs, 6 H), 2.28 (t, 2 H, J ) 7.2 Hz), 1.92 (bq, 2 H, J ) 7.7 Hz),
NMR (DMSO-d6) δ 11.84, 10.08, 10.06, 9.89, 9.88, 9.83 (s, 1 H each),
8.25 (bs, 3 H), 8.12 (t, 1 H, J ) 5.7 Hz), 7.98 (q, 2 H, J ) 4.8 Hz),
7.48 (s, 1 H), 7.18-7.28 (m, 3 H), 7.16 (d, 1 H, J ) 1.5 Hz), 7.14 (d,
1 H, J ) 1.5 Hz), 7.11 (bs, 2 H), 7.08 (d, 1 H, J ) 1.5 Hz), 7.02 (d,
2 H, J ) 1.5 Hz), 7.01 (d, 1H, J ) 1.8 Hz), 6.91 (d, 1 H, J ) 1.5 Hz),
6.83 (d, 1 H, J ) 1.8 Hz), 6.52 (d, 1H, J ) 7.8 Hz), 6.47 (d, 1 H, J )
8.4 Hz), 3.92 (s, 3 H), 3.80 (s, 6 H), 3.79, 3.78, 3.77, 3.75, 3.73 (s, 3H
each), 3.04 (q, 2 H, J ) 6.0 Hz), 2.94 (p, 2 H, J ) 4.8 Hz), 2.68 (d,
6 H, J ) 4.2 Hz), 2.28 (t, 2 H, J ) 7.5 Hz), 1.94 (bq, 2H, J ) 8.4 Hz),

9360 J. Am. Chem. Soc., Vol. 122, No. 39, 2000
EllerVik et al.
1.67 (p, 2 H, J ) 8.7 Hz); MALDI-TOF-MS calcd. for C61H74N20O12
(M+H): 1279.6. Found: 1279.7.
phosphorimaging plates (Kodak Storage Phosphor Screen S0230
obtained from Molecular Dynamics) were pressed flat against gel
samples and exposed in the dark at 22 °C for 12-16 h. A Molecular
Dynamics 400S PhosphorImager was used to obtain all data from the
storage screens. The data were analyzed by performing volume
integration of all bands using ImageQuant v. 3.2 software. All DNA
manipulations were performed according to standard protocols.9
Construction of Plasmid DNA. The plasmid pCW15 was con-
structed by hybridization of the inserts 5′-CTAGCGGCTATGTAA-
ATGGATGGCGGCTATGTA TATGGATGGCGGCTATGTACAT-
GGATGGCGGCTATGTAGATGGATTGCA-3′ and 5′-ATCCATC-
TACATAGCCGCCATCCATGTACATAGCCGCCATCCATATA
CATAGCCGCCATCCATTTACATAGCC-3′. The hybridized insert
was ligated into linearized pUC19 XbaI /PstI plasmid using T4 DNA
ligase. The resultant constructs were used to transform Top10F′ OneShot
competent cells from Invitrogen. Ampicillin-resistant white colonies
were selected from 25 mL of Luria-Bertani medium agar plates
containing 50 µg/mL ampicillin and treated with XGAL and IPTG
solutions. Large-scale plasmid purification was performed with Qiagen
Maxi purification kits. Dideoxy sequencing was used to verify the
presence of the desired insert.
Acknowledgment. We are grateful to the National Institutes
of Health (GM-27681) for research support, the Knut and Alice
Wallenberg Foundation for a postdoctoral fellowship to U.E.,
and the National Institutes of Health for a research traineeship
award (GM-08501) to C.C.C.W. We thank G. M. Hathaway
for MALDI-TOF mass spectrometry.
Preparation of 3′- and 5′-End-Labeled Restriction Fragments.
The plasmid pCW15 was linearized with EcoRI and PVuII and then
treated with Sequenase (version 2.0 from United States Biochemical),
deoxyadenosine 5′-[R-32P]triphosphate, and thymidine 5′-[R-32P]-
triphosphate for 3′ labeling. The 3′ labeled fragment was loaded onto
a 7% nondenaturing polyacrylamide gel, and the desired 284 base pair
band was visualized by autoradiography and isolated. Chemical
sequencing reactions were performed according to published methods.7
Quantitative DNase I Footprinting. DNase I Footprinting reactions
were carried out as previously described.8 Photostimulable storage
Supporting Information Available: Details on the synthesis
and quantitative DNase I footprinting titrations of eight-ring
hairpin polyamides containing substituted pyrroles and hydroxy-
pyrroles at the terminal position (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
JA001692U
(8) Trauger, J. W.; Baird, E. E.; Dervan, P. B. Nature (London) 1996,
382, 559-561.
(9) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning; Cold
Spring Harbor Laboratory: Cold Spring Harbor, NY, 1989.
(7) (a) Iverson, B. L.; Dervan, P. B. Nucleic Acids Res. 1987, 15, 7823-
7830. (b) Maxam, A. M.; Gilbert, W. Methods Enzymol. 1980, 65, 499-
560.