Structural selectivity of aromatic diamidines

Article abstract of DOI:10.1021/jm049491e

Competition dialysis was used to study the interactions of 13 substituted aromatic diamidine compounds with 13 nucleic acid structures and sequences. The results show a striking selectivity of these compounds for the triplex structure poly dA:(poly dT)

Full text of DOI:10.1021/jm049491e

J. Med. Chem. 2004, 47, 5729-5742
5729
Structural Selectivity of Aromatic Diamidines
Jonathan B. Chaires,*,† Jinsong Ren,‡ Donald Hamelberg,§ Arvind Kumar,§ Vandna Pandya,§
David W. Boykin,*,§ and W. David Wilson*,§
James Graham Brown Cancer Center, Department of Medicine, Health Sciences Center, University of Louisville,
529 S. Jackson St., Louisville, Kentucky 40202, Key Laboratory of Rare Earth Chemistry and Physics,
Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin 130022, China, and
Department of Chemistry, Georgia State University, Atlanta, Georgia 30303
Received June 25, 2004
Competition dialysis was used to study the interactions of 13 substituted aromatic diamidine
compounds with 13 nucleic acid structures and sequences. The results show a striking selectivity
of these compounds for the triplex structure poly dA:(poly dT)2, a novel aspect of their interaction
with nucleic acids not previously described. The triplex selectivity of selected compounds was
confirmed by thermal denaturation studies. Triplex selectivity was found to be modulated by
the location of amidine substiuents on the core phenyl-furan-phenyl ring scaffold. Molecular
models were constructed to rationalize the triplex selectivity of DB359, the most selective
compound in the series. Its triplex selectivity was found to arise from optimal ring stacking on
base triplets, along with proper positioning of its amidine substituents to occupy the minor
and the major-minor grooves of the triplex. New insights into the molecular recognition of
nucleic acid structures emerged from these studies, adding to the list of available design
principles for selectively targeting DNA and RNA.
Introduction
midine has shown very promising therapeutic activity
against a variety of parasitic microorganisms.5 Rapid
cellular uptake with nuclear accumulation of this com-
pound is clearly observed by fluorescence microscopy
with the intrinsic compound fluorescence.6 In a very
promising result for the use of these compounds in
human treatment, a bis-methoxamidine prodrug of
DB75, which is activated to DB75 in vivo, has success-
fully completed phase I clinical trials. Phase II clinical
trials for the treatment trypanosomal diseases and for
treatment of Pneumocystis carinii pneumonia are pro-
ceeding, and to date the results show promising activity
and low toxicity.5
Natural products and synthetic organic cations that
bind in the DNA minor groove have proven useful as
probes for evaluation of DNA molecular recognition as
well as clinically useful agents in chemotherapy. The
antibiotic netropsin, for example, is a dicationic ami-
dine-guanidine natural product that has been extremely
useful in characterizing the full molecular basis for DNA
minor groove recognition.1-4 Pentamidine is a synthetic
diamidine dication that has been successfully used
clinically for many years against parasitic microorgan-
isms 5 and has seen increased use due to three primary
factors: (i) the spread of AIDS throughout the world,
(ii) the dramatic increase in microbial parasitic diseases
in tropical environments, and (iii) the increase in
immune-compromised transplant patients in developed
countries. The biological activity of the family of aro-
matic diamidines, such as those in Figure 1, is cor-
related with their ability to interact strongly with the
DNA minor groove in AT sequences with subsequent
inhibition of one or more DNA-directed enzymes or
control proteins.5 Because of problems with toxicity and
the lack of oral bioavailability of pentamidine and
related diamidines, a major synthetic effort has gone
into the search for additional compounds in this general
family that have improved biological activity and bio-
availability.
DB75 binds strongly in the minor groove of AT DNA
sequences and forms hydrogen bonds with the edges of
AT base pairs at the floor of the groove.7,8 The crescent
shape of the compound matches the curvature of the
minor groove receptor site in AT sequences. The com-
pound fits snuggly into the narrow and deep minor
groove of AT sequences and covers four base pairs in
its DNA complex. The G-NH2 group in minor groove
sequences with GC base pairs sterically interferes with
strong minor groove complex formation. As a result,
diamidines such as DB75 bind significantly more weakly
to GC than to AT DNA sequences. Interestingly, DB75
can actually switch to an intercalation rather than a
minor groove binding mode in GC-containing se-
quences.9 Apparently the relatively flat surface of the
diphenylfuran aromatic system of DB75 can stack with
GC base pairs better by intercalation than fit into the
wide and sterically hindered minor groove of GC se-
quences.
A furan-based diamidine called furamidine (DB75,
Figure 1) is related in structure to pentamidine. Fura-
* Address correspondence regarding nucleic acids binding analysis
to J.B.C., chemical synthesis to D.W.B., and molecular modeling to
W.D.W. E-mail: J.B.C., jbchai01@gwise.louisville.edu; D.W.B.,
chedwb@panther.gsu.edu; W.D.W., chewdw@panther.gsu.edu.
† University of Louisville.
The amidine terminal groups of DB75 contribute
significantly to the DNA binding interactions through
stacking with the walls of the minor groove, hydrogen
bonding, and polyelectrolyte interactions.8 Their charges
‡ Chinese Academy of Sciences.
§ Georgia State University.
10.1021/jm049491e CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/02/2004

5730 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
Figure 1. Structures of the aromatic diamidines are grouped into related sets. The top row contains the DB75 parent compound
and para derivatives. The second and third rows contain meta-meta and meta-para derivatives, respectively. The left column
contains isomers of DB75 that allow a rigorous test of the importance of the different structural features of the parent compound.
Other isomer sets are composed of the alkyl amidines DB181, DB581, and DB480, and the thiophenes DB351, DB495 DB560.
DB471, DB361, and DB443 provide a group of related cyclic amidine analogues. All compounds are substituted at the 2- and
5-positions of the central heterocycle, except for DB516 and DB480, which are substituted at the 2- and 4-positions.
drive coupled ion and (possible) water release, providing
a favorable free energy contribution.10 Replacement of
the amidines with a cyclic imidazoline group increases
the planarity of the overall structure and causes an
enhancement of the intercalation binding mode of the
compound, particularly with GC base pair containing
sequences.9 Alkylation of the amidine substituents can
affect the compound-DNA interaction strength as well
as the sequence preference in a manner that depends
on the size and shape of the alkyl group. Introduction
of a cyclopentyl or an isopropyl group on the amidine,
for example, enhances DNA interactions, while more
bulky groups can inhibit binding.10
The clinical activity of this class of compounds along
with the structural information on their DNA minor
groove binding mode as well as their varied and novel
DNA interactions makes the compound set an ideal
starting point for evaluation of interactions with a
variety of DNA and RNA structures and sequences.
Because of the different DNA sequence and compound
structure dependent binding modes of this compound
family, it is clear that within the group there could be
a variety of nucleic acid interaction types and strengths
that are not yet discovered. It is essential to develop a
broader knowledge of the possible nucleic acid interac-
tions of these compounds if we are to better understand
their biological activity and develop additional agents
against resistant organisms and diseases. An excellent
method for rapid investigation of compound interactions
with different nucleic acid sequences and structures is
the recently developed competition dialysis method.11-13
In the competition dialysis method, an array of nucleic
acid sequences and structures is dialyzed against a
common free ligand solution. More ligand accumulates
in the nucleic acid sample with the preferred sequence
or structure, providing a simple and direct quantitative
measure of selectivity. A variety of compounds with
systematic structural changes (Figure 1) were designed
and synthesized to determine how modifying different
parts of the DB75 structure would affect the ability of
the compound to interact with a validated panel of test
DNA and RNA structures and sequences.
The cationic groups of furamidine, for example, are
in the para-para positions of the phenyl component,
and minor groove binding is preferred in DNA com-
plexes of this compound. Isomers with para-meta and
meta-meta cationic groups (Figure 1), however, shift
their binding preference to an intercalation or mixed
binding mode with DNA duplexes, even with pure AT
DNA sequences.14 We have selected several different
types of compounds with modified positions of the
cationic groups for use in the dialysis assay. As de-
scribed above, imidazoline substituents as well as
alkylation of the amidine groups can also cause varia-
tions in DNA binding strength and binding mode, and
we selected four types of terminal side chains, amidine,
isopropyl-amidine, imidazoline, and dimethyl-imidazo-
line, to evaluate in detail the influence of this very
important function on the nucleic acid interactions of
the compounds. To probe the limits of furan substitu-
tion, the furan was converted to the thiophene hetero-
cycle, and 2,4-substituted furan isomers were prepared
to compare with the usual 2,5-substitution positions of
DB75 and analogues. The competition dialysis results
reported here provide a rationale to explain the rela-
tionships between sequence selectivity and mode of
binding to DNA for unfused aromatic dications related
to furamidine (Figure 1).

Structural Selectivity of Aromatic Diamidines
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5731
75 mL of ether and the resultant solid was filtered, washed
with ether, and dried under vacumm at 35 °C to yield 3.72 g
(92%) of imidate ester hydrochloride, which was used directly
in the next step. A suspension of imidate ester HCl (0.87 g,
0.002 mol) in 30 mL of absolute ethanol was saturated with
ammonia at 0-5 °C and the mixture was stirred at room
temperature for 12 h. The solvent was removed and the solid
was dissolved in ice water, filtered, and basified with cold 2
M NaOH, and the white precipitate which formed was filtered,
washed with water, and dried to yield 0.49 g (80%) of free base.
The freebase (0.3 g, 0.001mol) was converted into the HCl salt
by treating with ethanolic HCl to yield 0.35 g (90%). Mp: 277-
279 °C. 1H NMR (DMSO-d6/D2O): 8.18 (brs, 2H), 8.10 (d,
2H, J ) 7.6), 7.71-7.63 (m, 4H), 7.14 (s, 2H). 13C NMR
(DMSO-d6): 166.1, 152.5, 131.3, 130.6, 129.3, 129.0, 127.4,
123.3,110.5.FABMS: m/e305(M+ +1).Anal.(C18H16N4O‚2HCl‚
0.25H2O) C, H, N.
2,5-Bis[3-(2-imidazolino)phenyl]furan Dihydrochlo-
ride (DB361). A mixture of the above imidate ester HCl (0.87
g, 0.002 mol) and ethylenediamine (0.24 g, 0.004 mol) in 15
mL of dry ethanol was heated under reflux for 12 h. The
solvent was removed and the residue was treated with water,
filtered, and basified with 5 N NaOH. The precipitated solid
was filtered, washed with water, dried, and crystallized from
chloroform:ether, to yield an off-white solid (0.57 g, 80%). The
free base (0.45 g, 0.0013 mol) was converted into its HCl salt
with ethanolic HCl to yield 0.51 g (88%). Mp: 258-261 °C. 1H
NMR (DMSO-d6/D2O): 8.65 (s, 2H), 8.16 (d, 2H, J ) 8), 7.91
(dd, 2H, J ) 0.8, J ) 8), 7.67 (dd, 2H, J ) 8), 7.22 (s, 2H), 4.03
(s, 8H). 13C NMR (DMSO-d6): 164.5, 151.9, 130.8, 129.9, 129.4,
127.4, 123.6, 122.8, 110.1, 44.4. FABMS: m/e 357 (M+ + 1).
Anal. (C22H20N4O‚2HCl‚H2O) C, H, N.
Materials and Methods
Materials. The synthetic polydeoxyribonucteotides poly dT
and poly dA:poly dT used in melting sudies were purchased
from Amersham-Pharmacia Biotech, Inc. (Piscataway, NJ) and
were used without further purification. All studies were
done in BPES buffer, consisting of 6 mM Na2HPO4, 2 mM
NaH2PO4, 1 mM Na2EDTA, 0.185 M NaCl, pH ) 7.0.
Synthesis. The syntheses of DB75,15 DB181,16 DB351,17
DB480,18 and DB51618 have been previously reported. Melting
points were recorded using a Thomas-Hoover (Uni-Melt)
capillary melting point apparatus and are uncorrected. TLC
analysis was carried out on silica gel 60 F254 precoated
aluminum sheets and detected under UV light. 1H and 13C
NMR spectra were recorded by employing a Varian GX400 or
Varian Unity Plus 300 spectrometer, chemical shifts (d) are
in ppm relative to TMS as internal standard, and coupling
constants (J) are reported in hertz. Mass spectra were recorded
on a VG analytical 70-SE spectrometer. Elemental analyses
were obtained from Atlantic Microlab Inc. (Norcross, GA) and
are within (0.4 of the theoretical values. The compounds
reported as salts frequently analyzed correctly for fractional
moles of water. In each case proton NMR showed the presence
of water. All chemicals and solvents were purchased from
Aldrich Chemical Co., Fisher Scientific, Frontier Scientific, or
Lancaster.
2,5-Bis(3-amidinophenyl)furan
Dihydrochloride
(DB359). Divinyl sulfone (5.9 g, 0.05 mol) in 20 mL of dry
ethanol was added to a boiling mixture of 3-bromobenzalde-
hyde (18.5 g, 0.01 mol), anhydrous sodium acetate (2.46 g, 0.03
mol), and 3-benzyl-5-(2-hydroxyethyl)thiazolium chloride (0.27
g, 0.001 mol) in 120 mL of dry ethanol. The mixture was
allowed to reflux for 16 h, cooled, filtered, washed with ether
and water, and dried. The solid was extracted with boiling
chloroform (4 × 200 mL), and the combined chloroform extracts
were washed with a 20% NaHCO3 solution and dried over
sodium sulfate. The solution was passed through a silica gel
column to yield 1,4-bis(3-bromophenyl)butane-1,4-dione as a
white crystalline solid (7.90 g, 40%). Mp: 203-204 °C. 1H NMR
(DMSO-d6): 8.09 (d, 2H, J ) 2); 7.89 (t of d, 2H, J ) 1.6, J )
7.6), 7.82 (ddd, 2H, J ) 0.8, J ) 2.0, J ) 8.0), 7.5 (dd, 2H, J )
8.0), 3.39 (s, 4H). 13C NMR (DMSO-d6): 197.1, 138.4, 135.2,
130.4, 129.9, 126.4, 121.7, 32.3. MS: m/e 396 (M+).
2,5-Bis[3-(2-(4,4-dimethyl)imidazolino)phenyl]furan Di-
hydrochloride (DB471). A mixture of the above imidate ester
HCl (0.87 g, 0.002 mol) and 1,2-diamino-2-methylpropane
(0.296 g, 0.004 mol) in 15 mL of dry ethanol was heated at
reflux for 12 h. The free base was obtained (0.66 g, 80%), as a
white crystalline solid. The free base (0.6 g, 0.0014 mol) was
converted, with ethanolic HCl, into the yellow hydrochloride
salt (0.68 g, 96%). Mp: 259-262 °C. 1H NMR (DMSO-d6/
D2O): 8.97 (s, 2H), 8.22 (d, 2H, J ) 8.4), 8.08 (d, 2H, J ) 8.4),
7.70 (dd, 2H, J ) 8.4), 7.63 (s, 2H), 3.80(s, 4H), 1.53 (s, 12H).
13C NMR (DMSO-d6): 161.2, 151.7, 130.5, 129.5, 129.1, 127.3,
To a solution of the above diketone (7.92 g, 0.02 mole) in 80
mL of boiling acetic anhydride were added 5-6 drops of
concentrated H2SO4, and the mixture was heated at reflux for
5 m. The solution was cooled, poured into ice water, and
stirred, and the light brown solid which formed was filtered.
The solid was washed with water, dissolved in chloroform (200
mL), washed with saturated NaHCO3 and water, and dried
over Na2SO4. The dried solution was passed through a bed of
silica gel and after removal of the solvent yielded 5.7 g (75%)
of 2,5-bis-(3-bromophenyl)furan as a white crystalline solid.
Mp: 130-134 °C. 1H NMR (DMSO-d6): 7.98 (d, 2H, J ) 1.2),
7.79 (d, 2H, J ) 8.0), 7.37 (dd, 2H, J ) 8.0), 7.14 (s, 2H). 13C
NMR (DMSO-d6): 151.3, 131.8, 130.6, 129.9, 125.7, 122.3,
122.0, 109.3; MS m/e 378 (M+)
A mixture of the above dibromofuran (5.0 g, 0.013 mol) and
CuCN (4.15 g, 0.046 mol) in 10 mL of dry N-methyl-2-
pyrrolidone was heated at reflux under N2 for 2 h (monitored
by TLC). The mixture was cooled, stirred with 150 mL of 10%
aqueous NaCN for 5 h., filtered, and washed with water and
dried. The solid was dissolved in warm chloroform and passed
through a bed of neutral alumina and after evaporation of the
solvent yielded 2,5-bis(3-cyanophenyl)furan as white crystal-
line needles (2.65 g, 75%). Mp: 230-233 °C. 1H NMR (DMSO-
d6): 8.32 (s, 2H), 8.14 (d, 2H, J ) 8), 7.72 (d, 2H, J ) 8), 7.63
(dd, 2H, J ) 8), 7.25 (s, 2H). 13C NMR (DMSO-d6): 151.2,
130.7, 130.6, 129.8, 127.7, 126.8, 118.2, 112.1, 109.9. MS m/e
270 (M+). Anal. (C18H10N2O) C, H, N.
123.6, 122.8, 109.8, 61.2, 56.4, 27.0. FABMS: m/e 413 (M+
1). Anal. (C26H28N4O‚2HCl‚0.5H2O) C, H, N.
+
2,5-Bis[3-(2-imidazolinophenyl)]-3,4-dimethylfuran Di-
hydrochloride (DB443). A mixture of 2,5-bis(3-bromophenyl)
furan (3 g, 0.0079 mol), paraformaldehyde (4 g), and 80 mL of
48% HBr in acetic acid was stirred for 2 days at room
temperature. The mixture was carefully poured into 200 mL
of ice water, and the precipitated solid was filtered and washed
with water. The solid was dissolved in dichloromethane,
washed with 10% NaHCO3 and water, dried over Na2SO4, and
filtered, and the solvent was removed under reduced pressure.
The solid was recrystallized from CHCl3:ether (8:2) to yield
2,5-bis(3-bromophenyl)-3,4-dibromomethylfuran as off-white
solid (4.3 g, 76%). Mp: 162-163 °C. 1H NMR (DMSO-d6): 7.88
(dd, 2H, J ) 1.2), 7.68 (dd, 2H, J ) 1.2, J ) 8), 7.53 (dd, 2H,
J ) 1.2, J ) 8), 7.37 (dd, 2H, J ) 8), 4.64 (s, 4H). 13C NMR
(DMSO-d6): 150.1, 131.8, 131.2, 130.5, 129.4, 124.9, 123.2,
119.8, 22.6. MS: m/e 564 (M+). Anal. (C18H12Br4O) C, H.
A solution of the above dibromomethyl analogue (4 g, 0.0071
mol) in 80 mL of ether:THF (2:1) was added to a well-stirred
suspension of LAH (1.3 g, 0.0038 mol) in 30 mL of THF over
10 min. It was stirred at room temperature for 1.5 h, cooled,
and carefully decomposed with wet ether, diluted with 200 mL
of water, extracted with 2 × 100 mL of CH2Cl2, and dried over
Na2SO4. The solvent volume was reduced under partial
vacuum and the resultant solid was filtered and crystallized
from ether:hexane (9:1) to yield 2,5-bis(3-bromophenyl)-3,4-
dimethylfuran as a white solid (2.5 g, 87%). Mp: 135-136 °C.
1H NMR (DMSO-d6): 7.69 (d, 2H, J ) 1.2), 7.67 (d, 2H, J )
1.2), 7.41 (dd, 2H, J ) 8), 7.39 (d, 2H, J ) 8), 7.24 (dd, 2H, J
) 8), 2.25 (s, 6H). 13C NMR (DMSO-d6): 146.1, 132.5, 130.1,
A mixture of 2.5 g (0.0092 mol) of the above dicyano
compound in 35 mL of dry ethanol was saturated while stirring
and cooling (0-5 °C) with dry HCl gas. The mixture was
stirred at room temperature for 6 days (aliquots of the mixture
were monitored by TLC and IR). The mixture was diluted with

5732 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
7.85 (brd, 2H, J ) 7.6), 7.78 (s, 2H), 7.65 (dd, 2H, J ) 7.6). 13
NMR (DMSO-d6/D2O): 165.3, 141.7, 133.8, 129.9, 129.5, 128.5,
126.8, 126.4, 124.6. FABMS: m/e 321 (M + 1). Anal.
(C18H16N4S‚2HCl‚2.25H2O) C, H, N.
129.5, 129.3, 128.8, 121.3, 118.6, 113.6, 9.9. MS: m/e 406 (M+).
Anal. (C18H14Br2O) C, H.
C
A mixture of the above dibromo analogue (2.03 g, 0.005 mol)
and CuCN (1.78 g, 0.2 mol) in 15 mL of N-methyl-2-pyrrolidone
was heated at reflux under N2 for 1.5 h (TLC monitored),
cooled, diluted with water, and stirred with 50 mL of 10%
NaCN for 2 h, filtered, washed with water, and dried in air.
The light gray solid was chromatographed over neutral
alumina using ether, ether:chloroform (1:1) to yield 2,5-bis(3-
cyanophenyl)-3,4-dimethylfuran as a white crystalline solid
2-(3-Amidinophenyl)-5-(4-amidinophenyl)furan Dihy-
drochloride (DB555). To a stirred mixture of 1-(4′-bro-
mophenyl)-3-dimethylaminopropane-1-one hydrochloride (14.0
g, 0.05 mol), 3-benzyl-5-(2-hydroxyethyl)-4-methyl-1,3-thiazo-
lium chloride (0.68 g, 0.0025 mol), and triethylamine (15.15
g, 0.15 mol) in 200 mL of anhydrous dioxane was added
3-bromobenzaldehyde (9.25 g, 0.05 mol), and the mixture was
heated at reflux for 12 h (under nitrogen). The solvent was
removed and the residue stirred with 200 mL of water, and
the gummy mass was extracted with 3 × 20 mL of warm
chloroform, washed with 2 × 100 mL of 10% NaHCO3, dried
over Na2SO4, and filtered. After removing the solvent, the re-
sidual mass was stirred with 30 mL of ether:ethanol (1:2), and
filtration yielded 1-(4-bromophenyl)-4-(3-bromophenyl)bu-
tane-1,4-dione as a white crystalline solid (7.96 g, 40%). Mp:
1
(2.2 g, (74%). Mp: 189-191 °C. H NMR (DMSO-d6): 8.13 (d,
2H, J ) 0.8), 8.03 (dd, 2H, J ) 0.8, J ) 8), 7.74 (dd, 2H, J )
0.8, J ) 8), 7.66 (dd, 2H, J ) 8), 2.23 (s, 6H). 13C NMR (DMSO-
d6): 146.1, 132.5, 130.1, 129.5, 129.3, 128.8, 121.3, 118.6, 113.1,
9.1. MS: 298 (M+). Anal. (C20H14N2O). C, H, N.
A mixture of 1.49 g (0.005) of the above dicyano compound
in 35 mL dry ethanol was saturated while stirring and cooling
(0-5 °C) with dry HCl gas. The mixture was stirred at room
temperature for 6 days (aliquots were monitored by TLC and
IR). The mixture was diluted with 75 mL of ether, and the
resultant solid was filtered, washed with ether, and dried
under vacuum at 35°C to yield 1.97 g (85%) of the imidate ester
hydrochloride. A mixture of imidate ester HCl (0.92 g, 0.002
mol) and ethylenediamine (0.24 g, 0.004 mol) in 30 mL of dry
ethanol was heated at reflux for 12 h, and the solvent was
removed under reduced pressure. The resultant solid was
treated with water basified with 10% aq NaOH. The precipi-
tated solid was filtered, washed with water, dried, and
crystallized with a ethanol:ether mixture to yield 0.64 g (83%)
off-white free base. The free base (0.58 g, 0.0015 mol) was
converted to its yellow hydrochloride salt with ethanolic HCl
(0.63 g, 92%). Mp: 267-269 °C (dec). 1H NMR (DMSO-d6/
D2O): 8.22 (s, 2H), 8.05 (d, 2H, J ) 8), 7.79 (d, 2H, J ) 8), 7.7
(dd, 2H, J ) 8), 4.0 (s, 4H), 2.22 (s, 6H). 13C NMR (DMSO-d6/
D2O): 165.5, 146.3, 132.2, 131.3, 130.5, 127.2, 125.1, 123.2,
122.1,44.9,9.9.FABMS: m/e385(M++1).Anal.(C24H24N4O‚2HCl)
C, H, N.
2,5-Bis[3-amidinophenyl]thiophene Dihydrochloride
(DB495). To a mixture of 2,5-bis(trimethylstannyl)thiophene
(5.11 g, 0.0125 mol) and 3-bromobenzonitrile (4.55 g, 0.025 mol)
in 45 mL of dry dioxane was added Pd [(PPh3)4] (0.575 g, 2
mol %) under nitrogen. The mixture was heated at reflux for
8 h (TLC monitored), the solvent was distilled under vacuum
and the oily residue was dissolved in 200 mL of dichlo-
romethane, and 100 mL of 10% KF (aq) was added. The
mixture was stirred for 1 h, filtered through Celite, and
washed with water, and the organic layer was dried over
Na2SO4. The solvent was removed under vacuum and the re-
sulting oily residue was chromatographed over silica gel (elu-
tion with 3:7 hexane:benzene) to afford 2,5-bis(3-cyanophenyl)-
thiophene (2.95 g, 82%) as a white solid. Mp: 178-180 °C. 1H
NMR (DMSO-d6): 7.87 (dt, 2H, J ) 0.8, J ) 1.2), 7.82 (ddd,
2H, J ) 0.8, J ) 1.2 Hz, J ) 8), 7.57 (dt, 2H, J ) 1.2, J ) 8),
7.52 (dt, 2H, J ) 0.8, J ) 8), 7.35 (s, 2H). 13C NMR (DMSO-
d6): 142.3, 136.1, 131.0, 129.9, 129.7, 129.0, 125.6, 118.3, 113.5.
MS: m/e 286 (M+). Anal. (C18H11N2S) C, H, N.
1
118-119 °C. H (NMR (DMSO-d6): 8.10 (dd, 1H, J ) 1.6),
8.0 (brd, 1H, J ) 7.6), 7.93 (d, 2H, J ) 7.6), 7.83 (brd, 1H, J )
7.6), 7.74 (d, 2H, J ) 7.6), 7.51 (dd 1H, J ) 8), 3.16 (br, 2H),
3.17 (br, 2H). 13C NMR (DMSO-d6): 197.5, 197.3, 138.5, 135.4,
135.3, 131.5, 130.7, 130.1, 129.6, 126.9, 126.6, 121.8, 32.3, 32.2.
MS: m/e 396 (M+). Anal. (C16H12Br2O2) C, H.
A solution of the above diketone (7.92 g, 0.02 mol) in 150
mL of CHCl3:CH3OH (7:3) was saturated with dry HCl gas at
0-5°C and stirred at 25 °C for 2 h (TLC). The solvent was
removed under reduced pressure, the residue was stirred with
water, and the solid was filtered, washed with water, sus-
pended in 10% NaHCO3 (150 mL), filtered, washed with water,
dried, and recrystallized from CHCl3:ether (3:7) to yield 2-(4-
bromophenyl)-5-(3-bromophenyl)furan as a white solid (6.7 g,
89%). Mp: 120-121 °C. 1H NMR (DMSO-d6): 7.98 (dd, 1H, J
) 1.6), 7.80 (brd, 1H, J ) 8), 7.77 (d, 2H, J ) 8.8), 7.62 (d, 2H,
J ) 8.8), 7.48 (brd, 1H, J ) 8), 7.39 (t, 1H, J ) 8), 7.17 (d, 1H,
J ) 3.5), 7.11 (d, 1H, J ) 3.6). 13C NMR (CDCl3): 153.0, 152.1,
132.5, 131.9, 130.3, 130.2, 129.4, 126.7, 125.4, 123.0, 122.3,
121.0, 109.4, 107.8. MS: m/e 378 (M+). Anal. (C16H10Br2O)
C, H.
A mixture of the above dibromo analogue (3.78 g, 0.01 mol)
and CuCN (2.67 g, 0.03 mol) in 60 mL of N-methyl-2-
pyrrolidinone was heated at reflux (under nitrogen) for 2.5 h.
The mixture was cooled, diluted with 100 mL water, and
stirred with 100 mL of 10% NaCN (aq) for 3 h, filtered, washed
with water, and dried. The crude product was dissolved in
chloroform and chromatographed over neutral alumina to yield
2-(4-cyanophenyl)-5-(3-cyanophenyl)furan as a pale solid (2.3
g, 85%). Mp: 290-293 °C.
1H NMR (DMSO-d6): 8.30 (t, 1H, J ) 1.6), 8.14 (brd, 1H, J
) 8), 8.02 (d, 2H, J ) 8.5), 7.86 (d, 2H, J ) 8.5), 7.74 (brd, 1H,
J ) 8), 7.65 (m, 1H), 7.32 (d, 1H, J ) 4), 7.30 (d, 1H, J ) 4).
13C NMR (DMSO-d6): 151.8, 151.6, 133.4, 132.5, 130.9, 130.5,
128.8, 127.8, 126.9, 123.9, 118.4, 118.0, 112.1, 111.3, 110.2,
109.5. MS: m/e 270 (M+). Anal. C18H10N2O) C, H, N.
The above dicyano compound (2.70 g, 0.01 mol) in 60 mL of
ethanol was saturated with dry HCl gas at 0-5 °C. The
reaction mixture was stirred at room temperature for 7-8 days
(monitored by IR and TLC) and diluted with ether, and the
yellow imidate ester hydrochloride was filtered, washed with
ether, and dried under vacuum for 6 h (3.88 g, 89%). The solid
was used in next step without further purification. A suspen-
sion of the imidate ester hydrochloride (0.87 g, 0.002 mol), in
25 mL of ethanol, was saturated with ammonia gas at 0-5 °C
and stirred for 24 h at room temperature. The solvent was
removed under reduced pressure, diluted with water, and
basified with 2 N NaOH (aq) to pH 10. The solid was filtered,
washed with water, and dried. The solid was dissolved in 15
mL of ethanol, treated with 2 mL of saturated ethanolic HCl,
and stirred at 35 °C for 2 h. The solvent was removed under
reduced pressure, and the residue was triturated with ether,
filtered, washed with ether, and dried under vacuum at 45
The above dicyano compound (2.86 g, 0.01 mol) in 70 mL of
ethanol was saturated with dry HCl gas at 0-5 °C and then
stirred at room temperature for 7 days (monitored by IR and
TLC). Ether was added to the mixture and the yellow imidate
ester dihydrochloride was filtered and washed with ether. The
solid was dried at 40 °C for 5 h to yield 4.3 g (95%). The
hygroscopic solid was used in the next step without further
purification. A suspension of imidate ester dihydrochloride (0.9
g, 0.002 mol) in 35 mL of ethanol was saturated with ammonia
gas at 0-5°C, stirred at room temperature for 2 days, and the
solvent was removed under reduced pressure. The solid was
suspended in water, the pH was adjusted to 9, and the pale
solid was filtered, washed with water, and dried. The solid
was stirred in saturated ethanolic HCl (20 mL) at 50-55 °C
for 1 h, treated with ether, filtered, and dried in a vacuum
oven at 75 °C for 24 h to yield 0.61 g (77%) of a pale yellow
1
1
solid. Mp: 305-307 °C (dec). H NMR (DMSO-d6/D2O): 9.65
°C for 24 h to yield 0.51 g (67%). Mp: 358-360 °C (dec). H
(br, 4H), 9.47 (br, 4H), 8.26 (br, 2H), 7.97 (brd, 2H, J ) 7.6),
NMR (DMSO-d6/D2O): 8.28 (dd, 1H, J ) 1.6 J ) 1.2), 8.15

Structural Selectivity of Aromatic Diamidines
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5733
(ddd, 1H, J ) 1.6, J ) 1.2, J ) 8), 8.03 (d, 2H, J ) 8.8), 7.93
(d, 2H, J ) 8.8), 7.77 (ddd, 1H, J ) 1.6, J ) 1.2, J ) 8), 7.68,
anophenylboronic acid (1.62 g, 0.011 mol) in 30 mL of
n-propanol, which was followed by addition of tetrakis(tri-
phenylphosphine)palladium (0.23 g, 2 mol %). The reaction
mixture was heated at reflux for 12 h (under nitrogen). The
solvent was removed under reduced pressure. The residue was
diluted with 50 mL of water, filtered, and dried. The solid was
dissolved in 100 mL of chloroform and filtered through Celite.
The solvent was removed and the solid was crystallized from
ether:benzene to yield 2-[3-cyanophenyl]-5-(4-cyanophenyl)-
thiophene (2.1 g, 73%) as a bright yellow crystalline solid.
Mp: 169-170 °C. 1H NMR (DMSO-d6): 8.13 (dd, 1H, J ) 1.6),
7.96 (ddd, 1H, J ) 0.8, J ) 1.6, J ) 8), 7.83 (brs, 4H) 7.74
(md, 1H, J ) 8), 7.71 (d, 1H, J ) 4), 7.62 (d, 1H, J ) 4), 7.62
(dd, 1H, J ) 8). 13C NMR (DMSO-d6): 142.0, 141.5, 137.2,
134.1, 132.7, 131.1, 130.1, 129.6, 128.3, 127.2, 126.7, 125.5,
118.2, 118.0, 112.2, 109.8. MS: m/e 286 (M+). Anal. (C18H10N2S)
C, H, N.
(dd, 1H, J ) 8), 7.31 (d, 1H, J ) 3.6), 7.26 (d, 1H, J ) 3.6). 13
C
NMR (DMSO-d6/D2O): 165.4, 165.0, 152.7, 152.0, 134.6, 130.5,
129.9, 128.9, 128.8, 128.6, 127.2, 126.1, 123.9, 123.3, 111.5,
110.3. Anal. (C18H16N4O. 2HCl‚0.25H2O). C, H, N.
2-[4-(N-Isopropylamidino)phenyl-5-[3-(N-isopropyl-
amidino)phenyl]furan (DB581). Freshly distilled isopro-
pylamine (0.5 mL, 0.005 mol) was added to the suspension of
imidate ester hydrochloride (0.65 g, 0.0015 mol) in 10 mL of
dry ethanol, and the mixture was stirred at room temperature
for 6 d. The solvent was distilled under vacuum, the residue
was treated with ice-cold water, and the pale yellow solution
was adjusted to pH 9 with 1 M NaOH. The solid was extracted
with CHCl3 and the organic phase was collected and dried over
anhydrous Na2SO4. The solvent was evaporated to dryness and
the residue was washed with ether, filtered, dried, and
crystallized from CHCl3:ether (1:4) to yield a pale crystalline
The above dicyano compound (2.86 g, 0.01mol) in 70 mL of
ethanol was saturated with dry HCl gas at 0-5 °C and then
stirred at room temperature for 7 days (monitored by IR and
TLC). Ether was added to the mixture and the yellow imidate
ester dihydrochloride was filtered and washed with ether. The
solid was dried at 40 °C for 5 h, to yield 4.3 g (95%). The
hygroscopic solid was used in the next step without further
purification. A suspension of imidate ester dihydrochloride (0.9
g, 0.002 mol) in 35 mL of ethanol was saturated with ammonia
gas at 0-5 °C and stirred at room temperature for 2 days,
and the solvent was removed under reduced pressure. The
solid was suspended in water, the pH was adjusted to 9, and
the pale solid was filtered, washed with water, and dried. The
solid was stirred in saturated ethanolic HCl (20 mL) at 50 °C
for 1 h, diluted with ether, filtered, and dried in a vacuum
oven at 75 °C for 24 h to yield 0.61 g (78%) of a pale yellow
1
solid (0.5 g, 90%). Mp: >300 °C. H NMR (DMSO-d6): 9.31
(br, 4H), 8.22 (s, 1H), 8.13 (d, 1H, J ) 8), 8.02 (d, 2H, J ) 8),
7.85 (d, J ) 8, 2H), 7.64 (m, 2H), 7.35 (d, J ) 3.6, 1H), 7.31 (d,
J ) 3.6, 1H), 4.12 (m, J ) 6, 2H), 1.3 (t, J ) 6, 12H). 13C NMR
(DMSO-d6): 159.0, 157.2, 152.6, 152.2, 138.4, 136.4, 130.5,
129.6, 128.3, 126.8, 125.5, 124.0, 122.7, 121.3, 108.5, 108.3,
43.4, 22.7, 22.6.
The free base (0.4 g, 0.0008 mol) was dissolved in 5 mL of
dry ethanol and saturated with dry HCl gas. The resulting
mixture was stirred for 15-20 min and the solvent was
removed under vacuum. The resulting residue was triturated
with dry ether, and the resulting solid was filtered, washed
with ether, and dried in vacuo at 60 °C for 24 h to yield yellow
1
crystalline solid (0.45 g, 92%): mp 300 °C. H NMR (DMSO-
d6): 9.75 (d, J ) 8, 1H), 9.65 (d, J ) 3.6, 1H), 9.61 (s, 1H),
9.51 (s, 1H), 9.31 (s, 1H), 9.19 (s, 1H), 8.25 (s, 1H), 8.15 (s,
1H), 8.03 (d, J ) 8, 2H), 7.87 (d, J ) 8, 2H), 7.67 (m, J ) 8,
2H), 7.36 (d, J ) 3.6, 1H), 7.32 (d, J ) 3.6, 1H), 4.17 (m, J )
6, 2H), 1.31 (m, J ) 6, 12H). Anal. (C24H28N40‚2HCl‚0.2H2O)
C, H, N.
1
solid. Mp: 320-321 °C (dec). H NMR (DMSO-d6/D2O): 8.12
(dd, 1H, J ) 1.6, J ) 8), 8.01 (brd, 1H, J ) 8), 7.91(d, 2H, J )
8.8), 7.88 (d, 2H, J ) 8.8), 7.76 (d, 1H, J ) 4), 7.75 (brd, 1H,
J ) 8), 7.72 (d, 1H, J ) 4), 7.69 (dd, 1H, J ) 8). 13C NMR
(DMSO-d6/D2O): 165.4, 165.0, 143.0, 141.9, 138.5, 134.2, 130.6,
130.3, 129.3, 128.9, 127.5, 127.0, 126.6, 125.6, 125.0. FABMS:
m/e 321 (M+ + 1).
2-[3-(Amidinopheny)]-5-[4-(amidinopheny)]thio-
phene Dihydrochloride (DB560). To a mixture of 2-tribu-
tylstannylthiophene (19.6 g, 0.05 mol) and 4-bromobenzonitrile
(9.1 g, 0.05 mol) in 45 mL of dry dioxane was added Pd[(PPh3)4]
(1.15 g, 2 mol %) under nitrogen. The mixture was heated at
reflux for 8 h (TLC monitored), the solvent was distilled under
vacuum, and the oily residue was treated with 100 mL of
dichloromethane and 100 mL of 10% KF (aq). The mixture was
stirred for 1 h, filtered through Celite, and washed with water,
the organic layer was dried over sodium sulfate, and the
solution was concentrated under reduced pressure to yield an
oily residue which was chromatographed over silica gel (elution
with hexane followed by 3:7 hexane:benzene). 2-(4-Cyano-
phenyl)thiophene (7.7 g, 83%) was obtained as a white solid.
Mp: 85-86 °C. 1H NMR (DMSO-d6): 7.81 (brs, 4H), 7.67 (brm,
2H), 7.18 (brm, 1H). 13C NMR (DMSO-d6): 141.2, 138.0, 132.9,
128.8, 127.9, 126.0, 125.8, 118.6, 109.5. MS: m/e 185 (M+).
Competition Dialysis. Competition dialysis experiments
were done exactly as previously described.11-13 The nucleic
acids used in the test array are listed in Table 1. The
preparation of these nucleic acids is fully described in refs 12
and 13. Competition dialysis studies were done in BPES buffer.
Thermal Denaturation Studies. Thermal denaturation
studies were done using a Cary 3E spectrophotometer (Varian,
Inc., Moutain View, CA) equipped with a Peltier temperature
controller.11 Solutions of poly dA:(poly dT)2 (final concentration
of 20 µM expressed in triplets) in BPES buffer were prepared
by direct mixing with aliquots from ligand stock solution,
followed by incubation for 12 h at 24 °C to ensure equilibration.
Samples were heated at a rate of 1 °C min-1 while continuously
monitoring absorbance at 260 nm.
Computational Methods. The three different planar low-
energy conformations of DB359 were obtained by rotating the
phenyl furan bond, and each conformation was docked into a
central intercalation site of the triplex dT6‚dA6‚dT6 followed
by energy minimization. The coordinates of the triplex and
central intercalation site used in this study were generated
as previously described for the dT10‚dA10‚dT10 triplex 10mer.19,20
Each minimized complex was then immersed in a periodic
box of TIP3P21 water molecules. Fifteen Na+ ions were place
around the complex using the Leap module in AMBER to
obtain electrostatic neutrality. The atomic charges and force
field parameters of DB359 were determined using the RESP22
methodology at the HF/6-31G* level of theory in Gaussian 98
and comparable parameters of the Cornell et al. force field.23
All minimization and molecular dynamics simulations were
carried out using the Amber 524 suites of program. The
complexes were equilibrated by using a multistage equilibra-
tion protocol previously described.25,26 After equilibration, each
final structure was subjected to 1 ns of molecular dynamics
simulation in the isobaric-isothermal ensemble at a constant
Bromine (1.76 g, 0.011 mol) in 15 mL of dichloromethane
was added over 15 m to a well-stirred solution of 2-(4-
cyanophenyl)thiophene (1.85 g, 0.01 mol) in 50 mL of dichlo-
romethane at 5-10 °C. The mixture was stirred at room
temperature for 2.5 h. The mixture was diluted with 50 mL of
water and 50 mL of dichloromethane. The organic layer was
washed with 5% sodium thiosulfate, 5% sodium bicarbonate,
and water and dried over sodium sulfate. The solvent was
removed under reduced pressure and the residual solid was
triturated with ether:hexane (1:4) to yield 2-bromo-5-(4-
cyanophenyl)thiophene (2.14 g, 81%) as a off-white crystalline
1
solid. Mp: 98-99 °C. H NMR (DMSO-d6): 7.83 (d, 2H, J )
8.4), 7.78 (d, 2H, J ) 8.4), 7.53 (d, 1H, J ) 4), 7.30 (d, 1H,
J ) 4). 13C NMR (DMSO-d6): 142.7, 136.8, 132.8, 131.9,
126.6, 125.6, 118.3, 112.9, 109.9. MS: m/e 264 (M+). Anal.
(C11H6BrNS) C, H, N.
Potassium carbonate (3.03 g, 0.022 mol) in 15 mL of water
was added to a well-stirred solution of the above 2-(4-
cyanophenyl)-5-bromothiophene (2.64 g, 0.01mol) and 3-cy-

5734 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
Table 1. Nucleic Acid Samples Used in Competition Dialysis Experiments
conformation
nucleic acid
λ (nm)
ꢀ (M-1 cm-1
)
monomeric unit
single-strand pyrimidine
single-strand purine
duplex DNA
poly dT
poly dA
264
257
260
260
260
260
262
254
260
260
254
260
260
8 520
8 600
nucleotide
nucleotide
base pair
base pair
base pair
base pair
base pair
base pair
base pair
base pair
base pair
triplet
C. perfringens (31% GC)
calf thymus (42% GC)
M. lysodeikticus (72% GC)
poly dA:poly dT
12 476
12 824
13 846
12 000
13 200
16 800
14 280
12 460
16 060
17 200
39 267
[poly (dAdT)]2
[poly (dGdC)]2
duplex RNA
DNA:RNA hybrid
Z DNA
triplex DNA
quadruplex DNA
poly rA:poly rU
poly rA:poly dT
[Br-poly (dGdC)]2
poly dA:(poly dT)2
(5′T2G20T2)4
tetrad
a Key: λ, wavelength; ꢀ, molar extinction coefficient at the wavelength λ, expressed in terms of the monomeric unit specified.
Figure 2. Comparative binding of 13 aromatic diamidine derivatives to 13 nucleic acid structures.
temperature of 300 K and pressure of 1 bar. The SHAKE27
triplex poly dA:(poly dT)2, often better than to duplex
algorithm was applied to all bonds involving a hydrogen atom,
DNA forms. All compounds also bind well to the parallel
thus allowing a time step of 2.0 fs in solving the Newton’s
stranded qudruplex (T2G20T2)4, but not as well as to the
equations of motion. The particle mesh Ewald28 method was
triplex. There is little or no interaction with duplex
used to treat the long-range electrostatic interactions, and the
RNA, with a DNA:RNA hybrid, or with left-handed Z
DNA. Compound binding to duplex DNA forms is highly
variable and depends on base composition and sequence.
Figure 3 shows the contrasting binding pattern of the
aromatic diamidines to those of common intercalating
and groove-binding compounds. Data are presented as
Tukey box plots,29 a graphical tool that concisely sum-
marizes the properties of distributions. In a box plot,
the horizontal line within the box marks the 50th
percentile, while the square symbol marks the arith-
metic mean. The top edge of the box marks the 75th
percentile and the lower edge the 25th percentile. The
verticle lines extending from the box mark the 90th (top)
and 10th (bottom) percentiles. Crosses mark the 1st and
99th percentiles, and bars mark the complete range of
the data. The plots in Figure 3 show the distribution of
binding to each of the 13 nucleic acid structures
nonbonded van der Waals interactions were subjected to a 9
Å cutoff. Structures were collected at time steps of 1ps. The
last 100 ps of each simulation (100 structures) was averaged
and the average structure was energy minimized. The interac-
tion energy of each complex was calculated from the average
structures by subtracting the energies of the free triplex and
the free drug from that of the complex.
Results
Figure 2 shows the results of competition dialysis
experiments using the aromatic diamidines shown in
Figure 1. The interaction of 13 compounds with 13
different nucleic acid structures and sequences (Table
1) was studied. Several generalities may be seen at a
glance in Figure 2. None of the compounds bind to
single-stranded DNA as represented by poly dA and poly
dT. In contrast, all compounds bind tightly to the DNA

Structural Selectivity of Aromatic Diamidines
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5735
bind strongly to triplex and quadruplex structures.
Binding to duplex forms, however, is greatly diminished
for DB359 relative to DB75. The result is to render
DB359 more structurally selective than DB75, not by
enhancing triplex affinity but rather by reducing affinity
toward all other structural forms.
The structural selectivity of the aromatic diamidines
can be quantitatively compared by use of the specificity
sum, SS.30 To compute the specificity sum, binding
profiles such as shown in Figure 4 are first normalized
with respect to the maximal amount bound (Cmax) to any
of the structures. The normalized amounts bound are
then simply summed. The specificity sum is thus defined
as
13
Cb,i
SS )
∑
i ) 1Cmax
Since thirteen structures are used in this version of
the competitions dialysis assay, SS ranges from 1
(binding to only one structure; maximal structural
selectivity) to 13 (equal binding to all structures; no
structural selectivity). The data shown in Figure 4 yield
SS ) 4.0 for DB75 and SS ) 2.3 for DB359. Values for
the specificity sum for all compounds studied are shown
in Figure 5A and in Table 2. SS ranges from 2.3 to 4.9
for the various aromatic diamidines. In a study of 126
compounds from a wide variety of chemical classes, an
average value of SS ) 4.5 ( 2.0 was found (J. B.
Chaires, unpublished data). The aromatic diamidines
studied here are thus all more structurally selective
than average, in particular because of their general
preference for the triplex structure. (On the basis of the
average of multiple experiments, we estimate that the
amount of ligand bound to each nucleic acid structure
is determined with 5-10% error. Since SS is a sum of
normalized values, its error would be roughly equal to
the largest numerical error among the quantities
summed, based on the rules of error propagation. We
therefore believe that SS values are accurate to ap-
proximately 10%.)
While the metric SS is a measure of structural
selectivity, it does not contain information about com-
pound affinity. A compound might bind selectively to a
particular structure, but it may bind weakly. The ratio
Cmax/SS is therefore of interest, since it provides a
measure of both selectivity and affinity.30 Cmax is the
maximal amount of ligand bound to any structure and
is directly proportional to binding affinity. If Cmax is
large (high affinity) and SS is small (high selectivity),
the largest value for the ratio Cmax/SS will result. In
the study of 126 compounds, a range in Cmax/SS values
of 0.06 to 9.8 was found, with an average of Cmax/SS )
2.4 ( 2.2. Values for Cmax/SS for the aromatic di-
amidines are shown in Figure 5B and Table 2. Values
range from 1.6 to 6.2, indicating that the aromatic
diamidines possess greater combined affinity and se-
lectivity than average.
Figure 3. Comparison of the binding of aromatic diamidines
(A) to that of standard intercalators (B) and groove binders
(C). Data are presented as Tukey box plots, a graphical tool
that concisely summarizes the properties of distributions as
described in the text. The standard intercalators represented
in panel B include actinomycin D, ethdium bromide, propidium
iodide, 9-aminoacridine, quinacrine, daunorubicin, doxorubicin,
coralyne, thiazole orange, and ellipticine. The groove binders
represented in panel C include berenil, distamycin, DAPI,
DB75, DB351, DB181, and methyl green.
observed in competition dialysis experiments for each
class of compound.
The binding pattern for the aromatic diamidines
(Figure 3A) is clearly distinct from those observed for
common intercalators (Figure 3B) and groove binders
(Figure 3C). The aromatic diamidines show significant
binding to only AT-containing DNA duplexes, to the
TAT triplex, and to the parallel quadruplex structure.
Intercalators, as a class, show far less binding specificity
and bind to most structures in the assay, except for
single-stranded DNA and poly dA:poly dT. Intercalators
show a general preference for GC-rich duplex DNA, and
bind strongly to duplex RNA and to a DNA:RNA hybrid
structure. Groove binders strongly prefer binding to poly
dA:poly dT and [poly (dAdT)]2 over all other single- and
double-stranded structures. While groove-binders ap-
pear to bind strongly to triplex DNA, such binding is
illusory and was previously shown to arise from dis-
placement of the third strand and tight binding to the
resultant duplex.11 The most noticeable differences
between groove-binders and the aromatic diamidines
are the greater binding of the latter to the parallel-
stranded G-quadruplex structure, along with lesser
binding to the AT-rich natural DNAs.
Competition dialysis results for two compounds, DB75
and DB359, are shown in Figure 4 to facilitate detailed
comparison. DB75 is the parent compound of the
aromatic diamidines, while DB359 is a positional isomer
in which the amidine substituents were shifted from the
para-para orientation to the meta-meta orientation.
The change in geometry dramatically alters the struc-
tural preferences for binding. Both DB75 and DB359
Comparison of SS and Cmax/SS values identifies
compounds with the greatest combined selectivity and
affinity. In Figure 5A, for example, DB581 and DB359
show equal specificity sums and are the most selective
compounds of the group. However, inspection of Figure
5B shows that DB359 binds with higher affinity than

5736 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
Figure 4. Results of competition dialysis experiments using DB75 (A) and DB359 (B).
Figure 5. Values of the specificity sum, SS, and the ratio Cmax/SS (in µM) for aromatic diamidine compounds.
Cb
does DB581, as reflected by a larger Cmax/SS ratio. On
Kapp
)
the basis of SS and Cmax/SS values, DB359, DB555,
DB495, and DB560 are identified as the four compounds
with the most improved combined selectivity and affin-
ity relative to the parent compound DB75. All four
compounds show clear selectivity for triplex DNA and
to a lesser extent for the parallel quadruplex structure.
All strongly prefer multistranded structures over any
duplex or single-stranded form in the assay. These four
compounds are all either meta-meta or meta-para
isomers of DB75 and DB351, and it is clear that the
para to meta isomerization causes a significant enhance-
ment of triplex binding specificity.
Cf(CT - Cb)
where Cb is the amount bound (as shown in the
examples in Figures 2 and 4), Cf is the free ligand
concentration (maintained at 1 µM in the assay proto-
col), and CT is the total nucleic acid concentration (fixed
at 75 µM in the assay protocol). Binding free energies
may then be calculated by the standard relation
∆Gapp ) -RT ln Kapp
where R is the gas constant and T the temperature in
degrees kelvin. Values of ∆Gapp for selected structures
are listed in Table 2. (Binding constants obtained by
competition dialysis have been validated in many cases
by studies in which more laborious spectrophotometric
Competition dialysis data may be used to calculate
apparent free energies for binding interactions of any
of the structures in the assay.12 The apparent binding
constant for any structure is given by

Structural Selectivity of Aromatic Diamidines
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5737
Table 2. Selected Binding Properties of Aromatic Diamidines
Cmax/SS,
µM
∆GCT DNA
,
∆Gpoly (dAdT)
,
∆GTriplex
,
∆GQuadruplex,
compd
SS
kcal mol-1
kcal mol-1
kcal mol-1
kcal mol-1
DB581
DB359
DB555
DB495
DB471
DB480
DB560
DB181
DB516
DB361
DB075
DB351
DB443
2.32
2.32
2.81
2.92
2.98
3.57
3.60
3.64
3.71
3.98
4.01
4.50
4.90
3.74
6.23
5.94
5.88
2.36
2.71
4.67
3.23
2.78
2.68
4.23
4.08
1.62
-5.55
-5.74
-5.95
-5.91
-5.73
-6.21
-6.53
-6.53
-6.18
-6.10
-6.57
-6.82
-6.23
-6.22
-6.11
-6.62
-6.67
-6.07
-7.00
-7.02
-6.94
-6.86
-6.37
-7.15
-7.27
-6.34
-6.96
-7.31
-7.42
-7.44
-6.82
-7.03
-7.42
-7.16
-7.07
-7.09
-7.43
-7.49
-6.89
-6.15
-6.81
-6.92
-7.09
-6.32
-6.05
-6.96
-6.26
-6.63
-6.79
-6.97
-7.17
-6.75
Figure 6. Free energy difference plots (kcal mol-1) for the binding of aromatic diamidines to triplex (A, C) and quadruplex (B,
C) DNA. For panels A and B, duplex calf thymus DNA was taken as the reference. For panels C and D, duplex [poly (dAdT)]2 was
taken as the reference. Negative values indicate tighter binding to the triplex or quadruplex stuctures relative to the duplex
forms. Positive values indicate weaker binding.
titrations were done to obtain binding constants. Figure
S1 in Supporting Information shows the validation).
Selected free energy difference plots are shown in
Figure 6. Differences in apparent binding free energies
between multistranded triplex and quadruplex forms
and reference duplex structures are shown for all
aromatic diamidines. Calf thymus DNA and duplex poly
(dAdT) were used as reference duplexes. The notation
“∆∆GTriplex,CT DNA”, as an example, signifies the differ-
ence in binding free energies between triplex DNA and
calf thymus DNA, using the values listed in Table 2. In
these difference plots, negative values indicate more
favorable binding to the multistranded form relative to
the duplex form, while positive values indicate less
favorable interactions. Figure 6A,C shows that most
aromatic diamidines prefer the triplex poly dA:(poly dT)2
over both calf thymus DNA and duplex poly (dAdT). The
only exception is DB480, which binds about equally well
to the triplex and to poly (dAdT). Figure 6B,D shows
that the differential binding to the parallel stranded
quadruplex [T2G20T2]4 is more variable than to triplex
DNA. All compounds except DB181 and DB480 prefer
the quadruplex over calf thymus DNA. The relative
aversion of these two compounds for the quadruplex
structure is even greater when poly (dAdT) is taken as
the reference duplex.
From the data shown in Figures 5 and 6 and Table 2,
three compounds stand out as having improved struc-
tural selectivity than the parent compound DB75. These
compounds are DB359, DB495, and DB555. For these
compounds, binding to the triplex poly dA:(poly dT)2 is
greatly enhanced relative to any duplex form. To

5738 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
Figure 7. Thermal denaturation studies of selected aromatic diamidines. Differential melting curves for the denaturation of the
poly dA:(poly dT)2 triple helix is shown as a function of added DB75 (A), DB359 (B), DB495 (C), and DB55 (D). The molar ratios
(mol compound/mol triplet) of added compounds are indicated in Figure 8. Under the ionic conditions of these experiments, in the
absence of added ligand, the melting of the triplex occurs at 42.8 °C and melting of the duplex occurs at 76.0 °C.
confirm the preference for triplex, thermal denaturation
studies were done as shown in Figure 7. Differential
melting curves are shown in Figure 7. In the absence
edges of AT base pairs at the floor of the groove, and
the amidines form H-bonds with C2 keto groups of T
bases that project into the groove. The entire molecular
system makes excellent contacts with the walls of the
minor groove to create a complex that is well-optimized
for the structure of the para-para aromatic system and
the AT minor groove. An obvious initial difference
between DB75 and DB359 can be seen in Figure 9.
While rotation about the phenyl-amidine or phenyl-
furan bonds in DB75 does not change the molecular
conformation, similar rotations in DB359 do produce
different conformations.14 Two conformations of DB359
have the amidines pointed in the same direction, either
away from the furan O (linear form, upper left in Figure
9) or in the same direction as the O (curved form, lower
right in Figure 9). In the linear form, only one amidine
at a time can H-bond with AT bases at the floor of the
groove, because of steric hindrance from the hydrogen
atoms of the furan that also point into the groove. This
conformation does not have sufficient curvature to
match the DNA minor groove shape. In the curved form,
the amidines are very close, and although both could
form H-bonds, this complex is destabilized by charge
repulsion, and the entire phenylfuran aromatic system
is pushed out of the groove, so there is little favorable
contact with the bases and walls of the groove. This
conformation is too highly curved to match the groove
shape. Neither of these complexes nor any with the
amidines pointed in opposite directions are as favorable
as that for DB75, and as can be seen in Figure 4, DB359
of added ligand, the third strand dissociates at Tm
≈
43 °C, while the remaining duplex melts at Tm ≈ 76 °C.
Figure 7A shows that DB75 stabilizes both triplex and
duplex forms. The change in Tm as a function of the
molar ratio of added DB75 is shown in Figure 8A. DB75
clearly stabilizes the duplex to a greater extent than it
does the triplex form, an observation consistent with
the competition dialysis data shown in Figure 4A. In
contrast, as shown in Figures 7B and 8B, DB359 almost
exclusively stabilizes the triplex form and has little
affect on duplex melting. That observation is fully
consistent with the competition dialysis data shown in
Figure 4B. DB495 and DB555 (Figures 7C,D and 8C,D)
show behavior that is between that seen for DB75 and
DB359. DB495 and DB555 dramatically stabilize the
triplex form, but they also stabilize the duplex, although
to a far lesser extent than observed for DB75.
To construct plausible molecular models for the
DB359-DNA complexes to help understand the strong
and preferential binding of DB359 to triplex DNA,
several molecular mechanics simulations were con-
ducted with DB359 docked into duplex and triplex DNA
(Figure 9), and these were compared to results for DB75.
Crystal structures are available from the Neidle labora-
tory for complexes of DB75 and several derivatives with
an AATT minor groove binding site.3,8,31 In the com-
plexes the furan oxygen points into the minor groove,
edges of the two phenyls make close contact with the

Structural Selectivity of Aromatic Diamidines
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5739
Figure 8. Change in the melting temperature (∆Tm) in response to the addition of cationic furans. The change in the triplex Tm
is shown by solid squares, while the change in duplex Tm is shown by open squares. Data are shown for DB75 (A), DB359 (B),
DB495 (C), and DB555 (D). For multiphasic melting profiles, the Tm was defined as the temperature at the half-maximal absorbance
change.
binds more weakly to all AT-containing DNA duplexes
than DB75.
can be formed. The DB75 complex has the furan O
pointed in the opposite direction to the DB359 complex,
and the curvature of the molecular system places the
amidines in the same grooves as in the DB359 complex.
As can also be seen in Figure 4, the situation with
triplex binding is quite different, with both compounds
forming very favorable complexes. In docking experi-
ments with the triplex, no low-energy groove model
could be found with these compounds. Low-energy
triplex intercalated complexes are possible with both
DB359 and DB75, in agreement with experimental
results, and models with DB359 are shown as examples
in Figure 9. The lowest energy complex is formed with
the linear conformation (upper left in the Figure), and
it has excellent overlap of the DB359 aromatic system
with the TAT base triplet, one amidine in the minor
groove, and one amidine in the groove formed by the A
strand and the Hoogsteen T strand. The other groove,
formed by the two T strands, contains both T-CH3
groups and is more sterically constrained, more nonpo-
lar than the other two grooves, and is a less favorable
amidine docking site. Optimized complexes of other
conformations of DB359 have less favorable aromatic
overlap and/or less favorable amidine interactions to
give higher complex interaction energies. The curved
conformation of DB359 (lower right in Figure 9), for
example, has the optimum complex conformation with
the amidines in the same groove as with the linear
complex, but both the aromatic overlap and the amidine
interactions are less favorable than with the linear
complex. All other complexes have even less favorable
interactions energies. An intercalated complex of DB75
that is very similar to that in the upper left of Figure 9
Discussion
The competition dialysis studies described here show
that the aromatic diamidines display a unique pattern
of structural selectivity, a pattern that is distinct from
the patterns of classic intercalators or groove-binding
agents. Aromatic diamidines, in general, bind prefer-
entially to the triplex poly dA:(poly dT)2. Certain
compounds, such as DB359, exhibit a particularly high
selectivity for the triplex, derived not from greater
affinity for the triplex, but rather from reduced binding
affinity for all other structures. We note that the poly
dA:(poly dT)2 triplex is but one possible triple-helix form.
How the aromatic diamidines interact with triple helices
with different sequences and strand orientation requires
further study and would require competition dialysis
experiments using a more extensive array of triplex
structures.
DB359 is, in fact, one of the most triplex selective
compounds discovered to date. Table 3 compares the
selectivity and affinity of several compounds. The struc-
tures of these compounds are shown in the Supporting
Information (Figure S2). The pronounced triplex selec-
tivity of naphthylquinolines, exemplified by MHQ12,
was recently described in detail.30 MHQ12 emerged
from a rational design effort aimed at targeting triplex

5740 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
Figure 9. The five lowest energy molecular models for the interaction of DB359 with triplex DNA are shown. Molecular mechanics
energies (E) are shown below the models, and for comparison, ∆E values relative to the lowest energy structure in the upper left
are included. DB359 is shown in space-filling representation (carbon, white; hydrogen, light blue; nitrogen, dark blue; oxygen,
red) and the base triplets above and below the intercalated compound are shown as green tubes. Different conformations of
DB359, created by rotation about the phenyl-furan bonds, were docked into different orientations of a triplex intercalation site
and energy minimized as described in the text.
Table 3. Properties of Triplex-Binding Compounds
orientation. Molecular modeling studies (Figure 9) pro-
vide a rationale for the enhanced selectivity. The
energetically most favorable structure of the DB359-
triplex complex (Figure 9A) is one in which the com-
pound rings are optimally stacked over the base triplets,
with the amidines oriented to occupy the minor and
major-minor grooves of the TAT triplex. Placing the
amidines in the para-para orientations, as in DB75,
also allows favorable stacking over the base triplets and
proper positioning of the amidines in the triplex grooves.
DB75, however, also has a very favorable conformation
for interaction with the minor groove in duplex se-
quences of four or more AT base pairs. We cannot rule
out the possibility at this time that DB75 also binds to
the triplex grooves. It can clearly intercalate well with
the triplex, and for the usual wide minor groove of the
triplex, groove binding may be less favorable. Our
previous study of naphthylquinolines revealed the vital
importance of substituent groove binding to triplex
binding affinity.30 A QSAR was built that correlated the
triplex binding affinity of a series of naphthylquniolines
with the solvent accessible surface area of substituents
that would occupy triplex grooves. The model for the
DB359-triplex complex reinforces the importance of
groove interactions as a determinant of triplex selectiv-
ity and affinity.
compound
SS
Cmax/SS
∆GDuplex
∆GTriplex
MHQ1230
1.32
1.43
1.59
2.34
4.34
4.01
5.05
5.08
6.39
8.93
3.67
3.55
6.23
5.22
4.23
1.93
2.55
1.38
-4.2
0
-7.0
-6.5
-6.6
-6.95
-7.5
-7.4
-6.5
-6.7
-6.8
R-Naphthoflavone33
berberine11
DB359
-4.7
-5.7
-6.3
-6.6
-6.4
-6.1
-6.5
coralyne11
DB75
ethidium11
2,6-anthraquinone34
BePI35
structures and is by all measures the most selective
triplex binder designed and synthesized so far.
The selectivity and triplex affinity of DB359 compare
favorably to those of MHQ12, although its interactions
with triplex are serendipitous and were not achieved
by design. Several interesting points of comparison
emerge from the data of Table 3. Berberine and R-naph-
thoflavone, for example, are both more selective than
DB359, as judged by the SS parameter. Neither com-
pound, however, binds as tightly as does DB359 and
consequently have lower Cmax/SS values. Coralyne and
DB75 show the highest triplex binding affinity of the
compounds listed in Table 3, but their selectivity is
compromised by significant binding to other structures
and sequences. Optimizing selectivity for a particular
target is thus not simply a matter of maximizing its
affinity for that target but also of reducing compound
affinity for competing structures.
The vital role of geometry as a determinant of
selectivity for triplex is reinforced by the data obtained
for the compounds DB351 and DB495. For these com-
pounds, the furan core was replaced by a thiophene. In
DB351, the amidines are in the para-para orientation,
while in DB495 the orientation is shifted to meta-meta.
The enhanced triplex selectivity of DB359 relative to
the parent compound DB75 results from a geometry
change in which the amidine substiuents are shifted
from the para-para orientation to the meta-meta

Structural Selectivity of Aromatic Diamidines
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5741
(7) Neidle, S.; Nunn, C. M. Crystal structures of nucleic acids and
their drug complexes. Nat. Prod. Rep. 1998, 15, 1-15.
(8) Laughton, C. A.; Tanious, F.; Nunn, C. M.; Boykin, D. W.; Wilson,
W. D.; et al. A crystallographic and spectroscopic study of the
complex between d(CGCGAATTCGCG)2 and 2,5-bis(4-gua-
nylphenyl)furan, an analogue of berenil. Structural origins of
enhanced DNA-binding affinity. Biochemistry 1996, 35, 5655-
5661.
(9) Wilson, W. D.; Tanious, F. A.; Ding, D.; Kumar, A.; Boykin, D.
W.; et al. Nucleic Acid Interactions of Unfused Aromatic
Cations: Evaluation of Proposed Minor-Groove, Major-Groove
and Intercalation Binding Modes. J. Am. Chem. Soc. 1998, 120,
10310-.
(10) Mazur, S.; Tanious, F. A.; Ding, D.; Kumar, A.; Boykin, D. W.;
et al. A Thermodynamic and Structural Analysis of DNA
Minor-groove Complex Formation. J. Mol. Biol. 2000, 300, 321-
337.
(11) Ren, J.; Chaires, J. B. Sequence and structural selectivity of
nucleic acid binding ligands. Biochemistry 1999, 38, 16067-
16075.
(12) Ren, J.; Chaires, J. B. Rapid screening of structurally selective
ligand binding to nucleic acids. Methods Enzymol 2001, 340, 99-
108.
(13) Chaires, J. B. A competition dialysis assay for the study of
structure-selective ligand binding to nucleic acids. In Current
Protocols in Nucleic Acid Chemistry; John Wiley & Sons: New
York, 2002; p 8.3.1-8.3.8.
(14) Nguyen, B.; Tardy, C.; Bailly, C.; Colson, P.; Houssier, C.; et al.
Influence of compound structure on affinity, sequence selectivity,
and mode of binding to DNA for unfused aromatic dications
related to furamidine. Biopolymers 2002, 63, 281-297.
(15) Das, B. P.; Boykin, D. W. Synthesis and antiprotozoal activity
of 2,5-bis(4-guanylphenyl)furans. J. Med. Chem. 1977, 20, 531-
536.
(16) Boykin, D. W.; Kumar, A.; Xiao, G.; Wilson, W. D.; Bender, B.
C.; et al. 2,5-bis[4-(N-alkylamidino)phenyl]furans as anti-Pneu-
mocystis carinii agents. J. Med. Chem. 1998, 41, 124-129.
(17) Das, B. P.; Boykin, D. W. Synthesis and antiprotozoal activity
of 2,5-bis(4-guanylphenyl)thiophenes and -pyrroles. J. Med.
Chem. 1977, 20, 1219-1221.
The geometry change renders DB495 more selective
than DB351 (Figure 5). These compounds reveal an
additional effect. Substitution of sulfur for oxygen (in
the change of the furan to the thiophene) slightly
reduces selectivity. Thus, the SS value of DB351 is
larger than that for DB75, and the SS value of DB495
is greater than that for DB359. The reason for this
behavior is not entirely clear, but may result simply
from the larger radius of sulfur relative to oxygen, which
could perturb stacking interactions within the binding
site.
Compounds DB555 and DB560 are para-meta iso-
mers of the parent compounds DB75 and DB351,
respectively. The para-meta isomers are both more
selective than the parents. Interestingly, the difference
in selectivity between DB555 and DB560 is more
pronounced than observed between DB359 and DB495.
The effect of changing the furan to thiophene thus seems
to be more pronounced for the para-meta isomers than
for the para-para isomers.
The possible biological consequence of triplex binding
by the aromatic diamidines is unknown. Tracts contain-
ing runs of poly purines/poly pyrimidines can, under
superhelical stress, form intramolecular triple helix
structures, so-called “H-DNA”.32 The biological role of
H-DNA has not been definitively established, but the
nonrandom distribution of sequences that can form the
structure within genomes suggests that it may play a
role in the regulation of gene expression. If so, selective
binding to H-DNA by the aromatic diamidines may
contribute to their clinical mode of action. Alternatively,
triplex binding would represent a significant competitive
binding reaction that could effectively reduce compound
binding to its preferred duplex target.
(18) Francesconi, I.; Wilson, W. D.; Tanious, F. A.; Hall, J. E.; Bender,
B. C.; et al. 2,4-Diphenyl furan diamidines as novel anti-
Pneumocystis carinii pneumonia agents. J. Med. Chem. 1999,
42, 2260-2265.
(19) Wilson, W. D.; Tanious, F. A.; Mizan, S.; Yao, S.; Kiselyov, A.
S.; et al. DNA triple-helix specific intercalators as antigene
enhancers: Unfused aromatic cations. Biochemistry 1993, 32,
10614-10621.
(20) Wilson, W. D.; Mizan, S.; Tanious, F. A.; Yao, S.; Zon, G. The
interaction of intercalators and groove-binding agents with DNA
triple-helical structures: The influence of ligand structure, DNA
backbone modifications and sequence. J. Mol. Recognit. 1994,
7, 89-98.
(21) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R.
W.; Klein, M. L. Comparison of simple potential functions for
simulating liquid water. J. Chem. Phys. 1983, 79, 926-935.
(22) Bayly, C. I.; Cieplak, P.; Cornell, W.; Kollman, P. A. A well-
behaved electrostatic potential based method using charge
restraints for deriving atomic charges: The RESP model. J.
Phys. Chem. 1993, 97, 10269-10280.
(23) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K.
M.; et al. A second generation force field for the simulation of
proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc.
1995, 117, 5179-5201.
(24) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E.,
III.; Ross, W. S.; et al. AMBER, version 5.0; University of
California: San Francisco.
Acknowledgment. Support was provided by grant
CA35635 from the National Cancer Institute (J.B.C.)
and grant GM61587 from the National Institute of
General Medicine (W.D.W. and D.W.B.). J.B.C. holds
the James Graham Brown Endowed Chair in Biophys-
ics, and he thanks the James Graham Brown Founda-
tion for its support.
Supporting Information Available: Figures showing the
validation of binding constants obtained by competition dialy-
sis and structures of the compounds listed in Table 3 and a
table of analysis data for the compounds produced. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(25) Hamelberg, D.; Williams, L. D.; Wilson, W. D. Influence of the
dynamic positions of cations on the structure of the DNA minor
groove: Sequence-dependent effects. J. Am. Chem. Soc. 2001,
123, 7745-7755.
(26) Hamelberg, D.; Williams, L. D.; Wilson, W. D. Effect of a neu-
tralized phosphate backbone on the minor groove of B-DNA:
Molecular dynamics simulation studies. Nucleic Acids Res. 2002,
30, 3615-3623.
References
(1) Wemmer, D. E. Designed sequence-specific minor groove ligands.
Annu. Rev. Biophys. Biomol. Struct. 2000, 29, 439-461.
(2) Wemmer, D. E. Ligands recognizing the minor groove of DNA:
Development and applications. Biopolymers 1999, 52, 197-211.
(3) Neidle, S. DNA minor-groove recognition by small molecules.
Nat. Prod. Rep. 2001, 18, 291-309.
(27) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical
integration of the Cartesian equations of motion of a system with
comstraints: Molecular dynamics of n-alkanes. J. Comput. Phys.
1977, 23.
(28) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T. A.; Lee,
H.; et al. A Smooth Particle Mesh Ewald Methodology. J. Chem.
Phys. 1995, 103, 8577-8593.
(4) Bailly, C.; Chaires, J. B. Sequence-specific DNA minor groove
binders. Design and synthesis of netropsin and distamycin
analogues. Bioconjugate Chem. 1998, 9, 513-538.
(5) Tidwell, R. R.; Boykin, D. W. Dicationic DNA Minor Groove
Binders as Antimicrobial Agents. Small Molecule DNA and RNA
Binders: From Synthesis to Nucleic Acid Complexes; Wiley-
VCH: New York, 2003; pp 416-460.
(6) Lansiaux, A.; Tanious, F.; Mishal, Z.; Dassonneville, L.; Kumar,
A.; et al. Distribution of furamidine analogues in tumor cells:
Targeting of the nucleus or mitochondria depending on the
amidine substitution. Cancer Res. 2002, 62, 7219-7229.
(29) Cleveland, W. S. The Elements of Graphing Data; Wadsworth
Advanced Books and Software: Monterey, CA, 1985.
(30) Chaires, J. B.; Ren, J.; Henary, M.; Zegrocka, O.; Bishop, G. R.;
et al. Triplex selective 2-(2-naphthyl)quinoline compounds: Ori-

5742 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
Chaires et al.
gins of affinity and new design principles. J. Am. Chem. Soc.
2003, 125, 7272-7283.
(33) Trent, J. O. Advances in DNA Molecular Modeling. An Update.
Methods Enzymol. 2001, 340, 290-326.
(31) Trent, J. O.; Clark, G. R.; Kumar, A.; Wilson, W. D.; Boykin, D.
W.; et al. Targeting the minor groove of DNA: Crystal structures
of two complexes between furan derivatives of berenil and the
DNA dodecamer d(CGCGAATTCGCG)2. J. Med. Chem. 1996,
39, 4554-4562.
(34) Jenkins, T. C. Targeting multi-stranded DNA structures. Curr.
Med. Chem. 2000, 7, 99-115.
(35) Mergny, J. L.; Duval-Valentin, G.; Nguyen, C. H.; Perrouault,
L.; Faucon, B.; et al. Triple helix-specific ligands. Science 1992,
256, 1681-1684.
(32) Sinden, R. R. DNA Structure and Function; Academic Press: San
Diego, 1994; 398.
JM049491E