Effect of PNA backbone modifications on cyanine dye binding to PNA-DNA duplexes investigated by optical spectroscopy and molecular dynamics simulations

Article abstract of DOI:10.1021/ja045145a

Optical spectroscopy and molecular dynamics simulations have been used to study the interaction between a cationic cyanine dye and peptide nucleic acid (PNA)-DNA duplexes. This recognition event is important because it leads to a visible color change, signaling successful hybridization of PNA with a complementary DNA strand. We previously proposed that the dye recognized the minor groove of the duplex, using it as a template for the assembly of a helical aggregate. Consistent with this, we now report that addition of isobutyl groups to the PNA backbone hinders aggregation of the dye when the substituents project into the minor groove but have a weaker effect if directed out of the groove. UV-Visible and circular dichroic spectroscopy were used to compare aggregation on the different PNA-DNA duplexes, while molecular dynamics simulations were used to confirm that the substituents block the minor groove to varying degrees, depending on the configuration of the starting amino acid. In addition to a simple steric blockage effect of the substituent, the simulations suggest that directing the isobutyl group into the minor groove causes the groove to narrow and the duplex to become more rigid, structural perturbations that are relevant to the growing interest in backbone-modified PNA for applications in the biological and materials sciences.

Full text of DOI:10.1021/ja045145a

Published on Web 02/18/2005
Effect of PNA Backbone Modifications on Cyanine Dye
Binding to PNA-DNA Duplexes Investigated by Optical
Spectroscopy and Molecular Dynamics Simulations
Isil Dilek,† Marcela Madrid,‡ Rojendra Singh,† Christian P. Urrea,† and
Bruce A. Armitage*,†
Contribution from the Pittsburgh Supercomputing Center and Department of Chemistry,
Carnegie Mellon UniVersity, 4400 Fifth AVenue, Pittsburgh, PennsylVania 15213
Received August 11, 2004; E-mail: army@andrew.cmu.edu
Abstract: Optical spectroscopy and molecular dynamics simulations have been used to study the interaction
between a cationic cyanine dye and peptide nucleic acid (PNA)-DNA duplexes. This recognition event is
important because it leads to a visible color change, signaling successful hybridization of PNA with a
complementary DNA strand. We previously proposed that the dye recognized the minor groove of the
duplex, using it as a template for the assembly of a helical aggregate. Consistent with this, we now report
that addition of isobutyl groups to the PNA backbone hinders aggregation of the dye when the substituents
project into the minor groove but have a weaker effect if directed out of the groove. UV-Visible and circular
dichroic spectroscopy were used to compare aggregation on the different PNA-DNA duplexes, while
molecular dynamics simulations were used to confirm that the substituents block the minor groove to varying
degrees, depending on the configuration of the starting amino acid. In addition to a simple steric blockage
effect of the substituent, the simulations suggest that directing the isobutyl group into the minor groove
causes the groove to narrow and the duplex to become more rigid, structural perturbations that are relevant
to the growing interest in backbone-modified PNA for applications in the biological and materials sciences.
Chart 1
Introduction
Peptide nucleic acid (PNA) is a synthetic DNA mimic in
which the hydrogen-bonding nucleobases are attached to a
polyamide backbone (Chart 1).1-3 The lack of negative charges
along the PNA backbone permits high-affinity hybridization to
complementary DNA and RNA sequences,4 even for targets that
have competing secondary or tertiary structure.5 This property
has led to numerous studies and applications for PNA as an
antisense agent, as a hybridization probe for diagnostics, and
as a capture strand for purification of nucleic acids.3,6
The chemical structure of PNA leads to distinct helical
secondary structures for its hybrids with DNA7,8 and RNA.9
The altered charge distribution along the backbone and distorted
spatial distribution of hydrogen-bond donor and acceptor groups
† Department of Chemistry.
‡ Pittsburgh Supercomputing Center.
in the grooves of the helices relative to canonical DNA or RNA
structures substantially impairs recognition by nucleic acid-
binding proteins and small molecules. For example, Wittung et
al.10 found that the DNA intercalator ethidium bromide and
minor-groove binder distamycin exhibited at least 100-fold lower
affinity for PNA-DNA hybrids compared with homologous
DNA-DNA duplexes. While several other methods are avail-
able for detecting PNA-DNA hybridization, a small-molecule
ligand for reporting on PNA hybridization in solution would
be useful for its simplicity.
(1) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science 1991, 254,
1498-1500.
(2) Beck, F.; Nielsen, P. E. In Artificial DNA: Methods and Applications;
Khudyakov, Y. E., Fields, H. A., Eds.; CRC Press: Boca Raton, FL, 2002;
pp 91-114.
(3) Nielsen, P. E.; Egholm, M. In Peptide Nucleic Acids: Protocols and
Applications, 2nd ed.; Nielsen, P. E., Ed.; Horizon Bioscience: Norfolk,
VA, 2004; pp 1-36.
(4) Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.; Freier, S. M.;
Driver, D. A.; Berg, R. H.; Kim, S. K.; Norde´n, B.; Nielsen, P. E. Nature
1993, 365, 566-568.
(5) Armitage, B. A. Drug DiscoVery Today 2003, 8, 222-228.
(6) Demidov, V. V. Expert ReV. Mol. Diagn. 2001, 1, 343-351.
(7) Leijon, M.; Graslund, A.; Nielsen, P. E.; Buchardt, O.; Norde´n, B.;
Kristensen, S. M.; Eriksson, M. Biochemistry 1994, 33, 9820-9825.
(8) Erikkson, M.; Nielsen, P. E. Nat. Struct. Biol. 1996, 3, 410-413.
(9) Brown, S. C.; Thomson, S. A.; Veal, J. M.; Davis, D. G. Science 1994,
265, 777-780.
In 1999, we reported the discovery of a group of cationic
cyanine dyes that exhibited high-affinity, sequence-independent
binding to PNA-containing hybrids such as PNA-DNA du-
9
10.1021/ja045145a CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 3339-3345
3339

A R T I C L E S
Dilek et al.
plexes, PNA-PNA duplexes, and PNA2-DNA triplexes.11 In
each case, the dye bound to the helix not as a monomer but
rather as an aggregate, even under conditions where the dye
was monomeric in solution. Thus, the PNA-containing double
or triple helix acted as a template for the assembly of a helical
cyanine dye aggregate. Of practical importance was the finding
that these aggregates exhibit strongly blue-shifted absorption
spectra, resulting in an instantaneous and visible color change
from blue to purple, in the case of DiSC2(5) (Chart 1). This
discovery has since been elaborated into assays for genetic
screening and single nucleotide polymorphism detection.12,13
While the relatively low sensitivity of this colorimetric response
will likely preclude widespread applications, we have found the
dye to be useful as a simple qualitative indicator for PNA
hybridization to folded DNA target strands.14,15
The molecular-level details of the aggregation by cyanine
dyes on PNA templates were not clearly determined in our
previous work due to the lack of high-resolution structural
information. The blue-shifted absorption band and exciton-
coupled induced circular dichroic bands indicated that the dyes
formed face-to-face H-aggregates with right-handed helicity. We
speculated that the dyes were using the helical twist of the minor
groove as the template for growth of the aggregate based on
the fact that even in triple-helical hybrids, where the major
groove is occupied by a second PNA strand, aggregation is
unaffected. Nevertheless, the fact that aggregation seems to
occur readily on any sequence indicates that if the minor groove
is in fact the recognition site, then the dye molecules must not
penetrate deeply into the groove, where sequence-dependent
differences in the structure would be most significant.16 To
further test the hypothesis that the minor groove is involved in
cyanine dye recognition of PNA hybrids, we synthesized PNA
analogues with substituents that sterically block the minor
groove. Circular dichroic spectropolarimetry, UV thermal melt-
ing curves, and molecular dynamics simulations demonstrate
that these modified PNAs still form stable duplexes with their
complementary DNA sequences, but UV-vis and CD spectra
reveal that aggregation of the cyanine dye on these sterically
blocked duplexes is significantly inhibited.
°C and equilibrated for 5 min. UV-vis absorbance at 260 nm was
recorded every 0.5 °C as samples were cooled and then heated at a
rate of 1.0 °C/min.
Dye Spectroscopic Experiments. PNA-DNA duplexes were
prepared by mixing equimolar amounts of the complementary strands
in buffer containing 100 mM NaCl, 10 mM sodium phosphate (pH
7.0), and 20% methanol, heating to 90 °C, and then cooling slowly (1
°C/min) to room temperature. Methanol was included in the buffer to
minimize the adsorption of the dye/PNA complexes to the walls of the
cuvettes. Samples were heated to 35 °C, and 1.0 µM aliquots of DiSC2-
(5) stock solution in methanol was added. The samples were cooled to
15 °C and allowed to equilibrate for 5 min before the UV-vis or CD
spectrum was recorded. The samples were then heated to 35 °C prior
to addition of the next aliquot.
Materials. Reagents and solvents were purchased from either Aldrich
or Fisher Scientific and used without further purification. Boc-Leu-
(D)‚H2O and Boc-Leu(L)‚H2O and di-tert-butyl dicarbonate were
purchased from Peptides International; unmodified Boc-protected PNA
monomers were from Applied Biosytems. DNA was purchased from
Integrated DNA Technologies and used as received. DNA and PNA
stock solutions were prepared in deionized water, and the concentrations
were determined spectrophotometrically by use of extinction coefficients
calculated from DNA nearest-neighbor values or PNA monomer values.
DiSC2(5) was purchased from Aldrich and used without further
purification. Stock solutions of the dye were prepared in methanol and
filtered through glass wool. The concentration of DiSC2(5) was
determined in methanol by use of the manufacturer’s extinction
coefficient (ꢀ651 ) 260 000).
PNA Oligomer Synthesis. PNA oligomers shown in Chart 1 were
synthesized manually on a lysine-substituted MBHA resin by standard
solid-phase peptide synthesis techniques.17,18 Monomers were activated
by HBTU for 1 min in the presence of MDCHA and pyridine to
minimize racemization of chiral monomers during coupling. Oligomers
were purified by reverse-phase high-performance liquid chromatography
(HPLC) and characterized by matrix-assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectrometry (P, calculated mass
2985.99, observed mass 2986.81; PL, calculated mass 3097.99, observed
mass 3101; PD, calculated mass 3097.99, observed mass 3096.05).
Molecular Dynamics Simulations. The PNA-DNA double helix
was simulated by means of molecular dynamics as follows. The starting
coordinates were obtained from the canonical B-DNA-DNA structure
of the same sequence, simulated with the module nucgen of AMBER.19
The DNA backbone and sugar atoms of one of the strands were
substituted for their corresponding PNA atoms or deleted, according
to the correspondence between the PNA and DNA backbone atoms.20,21
Since this mapping scheme does not include the carbonyl oxygen atoms
of the base linker, these atoms were added by geometrical calculation.
The C7′-O7′ bonds were oriented toward the C-terminus of the PNA
strand, as observed in NMR8 and crystallographic22 studies. The force
field used was AMBER7 parm94,23 complemented with previously
determined parameters for the PNA backbone.24 Hydrogen atoms and
sodium ions were added with the module LEaP of AMBER.19 The
system was immersed in a water bath, where the minimum distance
Experimental Section
Equipment. UV-vis measurements were performed on a Varian
Cary3 spectrometer equipped with a thermoelectrically controlled
multicell holder. CD measurements were recorded on a Jasco J715
spectropolarimeter equipped with a thermoelectrically controlled single
cell holder.
UV Melting Curves. Samples including the appropriate amounts
of complementary single strands of PNA and DNA were heated to 95
(10) Wittung, P.; Kim, S. K.; Buchardt, O.; Nielsen, P. E.; Norde´n, B. Nucleic
Acids Res. 1994, 22, 5371-5377.
(11) Smith, J. O.; Olson, D. A.; Armitage, B. A. J. Am. Chem. Soc. 1999, 121,
2686-2695.
(17) Christensen, L.; Fitzpatrick, R.; Gildea, B.; Petersen, K. H.; Hansen, H.
F.; Koch, T.; Egholm, M.; Buchardt, O.; Nielsen, P. E.; Coull, J.; Berg, R.
H. J. Pept. Sci. 1995, 3, 175-183.
(12) Wilhelmsson, L. M.; Norde´n, B.; Mukherjee, K.; Dulay, M. T.; Zare, R.
N. Nucleic Acids Res. 2002, 30.
(18) Koch, T. In Peptide Nucleic Acids: Protocols and Applications, 2nd ed.;
Nielsen, P. E., Ed.; Horizon Bioscience: Norfolk, VA, 2004; pp 37-60.
(19) Pearlman, D. A.; Case, D. A.; Caldwell, J. W.; Ross, W. S.; Cheatham, T.
E. I.; DeBolt, S.; Ferguson, D.; Seibel, G.; Kollman, P. Comput. Phys.
Commun. 1995, 91, 1-42.
(13) Komiyama, M.; Ye, S.; Liang, X.; Yamamoto, Y.; Tomita, T.; Zhou, J.-
M.; Aburatani, H. J. Am. Chem. Soc. 2003, 125, 3758-3762.
(14) Kushon, S. A.; Jordan, J. P.; Seifert, J. L.; Nielsen, P. E.; Nielsen, H.;
Armitage, B. A. J. Am. Chem. Soc. 2001, 123, 10805-10813.
(15) Datta, B.; Armitage, B. A. J. Am. Chem. Soc. 2001, 123, 9612-9619.
(16) The same dye aggregates on certain DNA-DNA duplex templates by
binding in the minor groove, but there is a strong sequence dependence
and the aggregate structure is clearly different from the PNA-templated
structures described here since the absorbance shifts to 590 nm for DNA-
DNA compared with 530-540 nm for PNA-DNA. See Hannah, K. C.;
Armitage, B. A. Acc. Chem. Res. 2004, 37, 845-853 for a review of DNA-
templated cyanine dye aggregates.
(20) Sen, S.; Nilsson, L. J. Am. Chem. Soc. 1998, 120, 619-631.
(21) Soliva, R.; Sherer, E.; Luque, F. J.; Laughton, C. A.; Orozco, M. J. Am.
Chem. Soc. 2000, 122, 5997-6008.
(22) Rasmussen, H.; Kastrup, J. S.; Nielsen, J. N.; Nielsen, J. M.; Nielsen, P.
E. Nat. Struct. Biol. 1997, 4, 98-101.
(23) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M. J.;
Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P.
A. J. Am. Chem. Soc. 1995, 117, 5179-5197.
9
3340 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Cyanine Dye Binding to Modified PNA−DNA Duplexes
A R T I C L E S
from the surface of the double helix to the water box was 9 Å. Periodic
boundary conditions and constant pressure and temperature were used.
A 9 Å cutoff was used for the Lennard-Jones interactions. The particle
mesh Ewald (PME)25-27 method was used to treat the electrostatic
interactions. The integration step was 1 fs, and SHAKE28 was applied
to constrain bonds involving hydrogen atoms. The nonbonded pair list
was updated whenever an atom moved more than 0.5 Å since the last
list update. First, the water was equilibrated by performing 5000 steps
of energy minimization followed by 25 ps of molecular dynamics, while
all PNA-DNA atoms were restrained with a force constant of 500
kcal/(mol‚Å2). Next, the temperature was raised to 300 K and the
constraining force on the backbone atoms was slowly decreased to zero
during 250 ps, while the bases were still constrained. Finally, the force
constant on the bases was slowly decreased to zero during an additional
250 ps of molecular dynamics. All the atoms were then allowed to
move. The structure obtained after 1.5 ns of molecular dynamics
simulation was used to generate the two structures with leucine
substitutions. Leu side chains replaced one of the hydrogens of the
C5′ atoms of residues T14 and T17 of the PNA strand. The Leu side
chains were both oriented toward the minor groove for one of the
structures (PLD) and away from the minor groove for the other structure
(PDD). Simulations for the three structures were continued for a total
of 6.2 ns. The first 500 ps were not considered for the analysis, to
ensure equilibration. Since we are interested in the effect of the Leu
side chains in the minor groove, only the interval 0.5-4 ns was
considered for the analysis of this trajectory. Helical parameters were
measured with Curves,29 and VMD was used for visualization.30
Figure 1. Circular dichroic spectra recorded for 5.0 µM unmodified (PD)
and modified (PLD and PDD) PNA-DNA duplexes. Samples were prepared
in 10 mM sodium phosphate buffer (pH ) 7.0) with 100 mM NaCl and
20% methanol. Spectra were recorded at 25 °C, and eight scans collected
at 100 nm/min were averaged.
groove. On the other hand, the isobutyl side chains should point
away from the minor groove when leucine-based monomers
having the D-configuration are incorporated into the PNA
strands. Thus, L-leucine-modified PNA would be expected to
inhibit cyanine dye aggregation to a greater degree than would
a D-leucine-modified analogue.
Results
Design. To block the minor groove of a PNA-DNA duplex,
we decided to synthesize a backbone-modified PNA strand. The
canonical PNA backbone consists of N-(2-aminoethyl)glycine
units, but the R-carbon of the glycine component can be readily
substituted by using other R-amino acids to synthesize the
backbone. Numerous modifications of this type have been
successfully introduced into PNA with only relatively minor
decreases in affinity for complementary DNA strands.31-35
We chose to incorporate PNA monomers based on N-(2-
aminoethyl)leucine because (i) L-leucine-modified PNA has
already been shown to hybridize to complementary DNA with
only modest destabilizations relative to unmodified PNA33,34
and (ii) the uncharged isobutyl side chain should not directly
affect the electrostatic contribution to aggregation of the cationic
cyanine dye on the net anionic PNA-DNA duplex. On the basis
of the NMR structure of a PNA-DNA duplex,8 incorporation
of leucine-based monomers having the L-configuration should
result in projection of the isobutyl side chains toward the minor
Synthesis and Characterization of Modified PNA-DNA
Templates. Details of the synthesis of modified PNA monomers
constructed from L- and D-leucine with thymine as the nucleo-
base are given in the Supporting Information and follow the
approach used by Haaima et al.32 with only minor modification.
These monomers were then incorporated into PNA decamers
PL and PD at two positions (Chart 1; TL and TD represent the T
monomers constructed from L- and D-leucine, respectively). An
unmodified PNA strand (P) was also synthesized for compari-
son. Two L-lysine amino acids were included at the C-terminus
of the PNA strands to provide water solubility. PNA oligomers
were synthesized manually by standard solid-phase peptide
synthetic techniques,17,18 purified by HPLC, and characterized
by MALDI-TOF mass spectrometry.
Hybridization of PNA single strands with complementary
DNA strands yields double-helical structures, which can be
characterized by circular dichroic (CD) spectropolarimetry and
UV melting studies.4 Figure 1 presents CD spectra recorded
for 5 µM PD, PDD, and PLD duplexes. CD spectra for all three
duplexes are very similar and match previously reported spectra
for PNA-DNA double-helical structures. It is clear that
incorporation of L- and D-leucine-modified PNA monomers does
not alter the gross structural features of the hybrid, although
the variation in intensity may signal minor local changes in the
helical structure.
(24) Shields, G. C.; Laughton, C. A.; Orozco, M. J. Am. Chem. Soc. 1998, 120,
5895-5904.
(25) Darden, T.; York, D.; Pedersen, L. J. Chem. Phys. 1993, 98, 10089-10092.
(26) Darden, T. A.; Bartolotti, L.; Pedersen, L. G. EnViron. Health Perspect.
Suppl. 1996, 104, 69-74.
(27) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen,
L. G. J. Chem. Phys. 1995, 103, 8577-8593.
(28) van Gunsteren, W. F.; Berendsen, H. J. C. Mol. Phys. 1997, 34, 1311-
1327.
(29) Lavery, R.; Sklenar, H. Curves 5.1, Computer Program; Institut de Biologie
Physico-Chimique, CNRS, 1996.
(30) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics 1996, 14, 33-
38.
(31) Dueholm, K. L.; Petersen, K. H.; Jensen, D. K.; Egholm, M.; Nielsen, P.
E.; Buchardt, O. Bioorg. Med. Chem. Lett. 1994, 4, 1077-1080.
(32) Haaima, G.; Lohse, A.; Buchardt, O.; Nielsen, P. E. Angew. Chem., Int.
Ed. Engl. 1996, 35, 1939-1942.
Denaturation of a duplex to its component single strands
usually results in a substantial increase in the absorbance at 260
nm (hyperchromicity) due to unstacking of the bases. The
midpoint of this transition is referred to as the Tm (melting point)
of the duplex. Figure 2 shows the denaturation curves for the
unmodified PNA-DNA duplex (PD) as well as L-leucine (PLD)
(33) Pu¨schl, A.; Sforza, S.; Haaima, G.; Dahl, O.; Nielsen, P. E. Tetrahedron
Lett. 1998, 39, 4707-4710.
(34) Sforza, S.; Haaima, G.; Marchelli, R.; Nielsen, P. E. Eur. J. Org. Chem.
1999, 197-204.
(35) Sforza, S.; Corradini, R.; Ghirardi, S.; Dossena, A.; Marchelli, R. Eur. J.
Org. Chem. 2000, 2905-2913.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3341

A R T I C L E S
Dilek et al.
Table 2. Minor-Groove Widthsa of the Simulated Structures,
Averaged over the Trajectories
base pair
PD
PLD
PDD
A1-T20
G2-C19
T3-A18
A4-T17
T5-A16
C6-G15
A7-T14
C8-G13
T9-A12
7
6
6
7
8
8
8
7
5
8
6
8
9
8
8
7
6
7
6
4
4
4
4
5
5
6
b
G10-C11
a Determined at the C5′ atoms of each base pair with Curves29 and given
in angstroms. Positions of leucine modification are shown in boldface type.
b The minor-groove width was undefined for this base pair.
Figure 2. UV melting curves of 5.0 µM PNA-DNA duplexes. Data were
collected at 0.5 °C intervals during a heating ramp of 1.0 °C/min.
Table 1. Comparison of the Helicoidal Parameters of the Three
Averaged Structures, an NMR-Determined PNA-DNA,8 and
Canonical B-DNAa
global avg twist (deg)
diameter (Å)
pitch (Å)
PD
28
32
25
28
36
21
18
21
22
15
42
39
54
42
34
PLD
PDD
NMR
canonical B-DNA
a Determined at the C5′ atoms of each base pair with Curves.29
and D-leucine (PDD) modified analogues. Incorporation of
leucine side chains decreases the melting temperature of the
PNA-DNA duplex by ca. 4 °C/modified residue, regardless
of the chirality. Similar results for modified PNAs were reported
previously by Nielsen and co-workers.32 Importantly, the
modified duplexes are fully formed at the temperatures and
concentrations used for the spectroscopic experiments described
below.
Molecular Dynamics Simulations. Molecular dynamics
simulations were performed to further evaluate the structure of
the three duplexes. In each case, the PNA-DNA duplex was
constructed from a canonical B-form duplex as described in
the Experimental Section. The helical structure departed quickly
from the starting B-form geometry, with a root-mean-square
(RMS) deviation between the PNA-DNA structure and canoni-
cal B-DNA, averaged over the last 2 ns of the simulation, of
5.3 Å. This is in agreement with previous molecular dynamics
simulations, which obtained RMS deviations between PNA-
DNA double helices and B-DNA of 4-5 Å.21 The three
simulated structures remain double-helical and stable during the
trajectories, as indicated by stable RMS and energetic values.
Table 1 lists the helical parameters for the three duplexes,
along with those determined from a solution NMR structure of
an unmodified PNA-DNA duplex.8 The parameters for B-form
DNA are also included for comparison. The simulations
effectively reproduce the NMR parameters for the unmodified
duplex. For the L-leucine-modified PNA, the twist increases,
leading to a decrease in helical pitch, while the opposite is
observed for the D-leucine-modified analogue.
Figure 3. View into the minor groove of the three simulated duplex
structures, unmodified PD (green), PLD (blue), and PDD (red). Structures
are averaged over the trajectories. The leucine side chains are labeled and
colored orange (L-configuration, directed into the minor groove) and gray
(D-configuration, directed away from the minor groove).
measured at each base pair (Table 2). When the isobutyl
substituents are directed toward the minor groove, the effect is
to make the minor groove narrower than in the unmodified
duplex by up to 50%. The narrowing is most pronounced at
the center of the duplex, i.e., closest to the two modifications.
In contrast, the D-leucine modifications have virtually no effect
on the minor-groove width. No significant effect on the major-
groove widths was observed in the simulated structures (data
not shown).
Superposed views into the minor grooves for the three
structures are shown in Figure 3. The gross similarities in the
three helices are consistent with the lack of significant variation
in the CD spectra. As designed, the L-leucine side chains
(orange) project directly into the minor groove while the
Differences between the two leucine-modified structures are
also evident when the minor-groove widths of the duplexes are
9
3342 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Cyanine Dye Binding to Modified PNA−DNA Duplexes
A R T I C L E S
this feature appears only as a weak shoulder at even the highest
dye/duplex ratio when PLD is used as the template. DiSC2(5)
forms noncovalent dimers in aqueous solution and exhibits an
absorption band at 579 nm.37 The band at 579 nm in Figure 5B
indicates that as the DiSC2(5) concentration increases, the dye
dimerizes in solution rather than aggregating on the template.
Conversely, the absence of this band in Figure 5A illustrates
the preferential aggregation of the dye on the unmodified PD
template. Clearly, the presence of two TL monomers prevents
the aggregation of the dye on PLD.
As expected, DiSC2(5) shows markedly better aggregation
on the D-leucine-modified template (PDD). UV-Vis spectra for
the dye in the presence of the three templates are shown in
Figure 6. The stronger absorption band at 543 nm for PDD
compared to PLD is consistent with the MD simulations, which
indicated that the D-leucine substituents project away from the
minor groove. Nevertheless, the weaker absorption for PDD
relative to the unmodified duplex indicates that the D-leucine
modifications do impair binding of the dye.
Analogous dye binding experiments were performed by CD.
DiSC2(5) is achiral, so the unbound dye in solution does not
present any CD signal. When it binds to PNA-containing hybrids
and assembles into helical aggregates, it exhibits a strong
induced CD signal.11 This phenomenon is demonstrated in
Figure 7. Binding of DiSC2(5) to PD induces a strong positive
CD signal at 552 nm and a negative signal at 521 nm. The
bisignate bands are attributed to exciton coupling due to the
interaction of two or more neighboring chromophores,38 indicat-
ing the aggregation of multiple dye molecules on the PD
template. In addition, the positive splitting provides evidence
of a right-handed helical relationship for the dye chromophores,
consistent with templating of the aggregate by the PNA-DNA
duplex, which is itself a right-handed helix. In contrast, the CD
signal for PLD is 5-6-fold weaker, consistent with the UV-
vis results. In addition, the lack of a CD signal in the region of
579 nm verifies that the UV-vis maximum observed at this
wavelength for PLD is due to unbound, dimerized dye. As with
the UV-vis experiment, the induced CD signal for DiSC2(5)
in the presence of PDD is intermediate between the other two
templates.
Figure 8 illustrates the dye concentration dependence for
aggregation on the three PNA-DNA templates as measured
by the intensity of the negative CD band. The results further
illustrate the differential impact of the two leucine configurations
on dye aggregation: L-leucine modifications hinder dye ag-
gregation to a much greater extent than do D-leucine modifica-
tions.
Finally, we previously described how the cooperative nature
of dye aggregation on PNA-DNA leads to highly temperature-
dependent binding, reflected in “melting” transitions for the
aggregates that are independent of those for the templates.11,36
Figure 9 shows melting curves for DiSC2(5) on the three
templates recorded by the UV-vis absorbance at 543 nm. The
L-leucine-modified template reduces the stability of the ag-
gregate by 8.8 °C, compared with only 4.8 °C for the
D-analogue. The reductions in melting temperature reflect the
fact that less dye is bound to the two modified templates based
on the intensity of the dye monomer absorption band in Figure
Figure 4. Minor-groove widths measured between the C5′ atoms of base
pair A4-T17 as a function of time for (a) PD, (b) PLD, and (c) PDD.
D-leucine side chains (gray) are directed into solution. The
narrowed minor groove for the L-leucine-modified duplex is also
evident in the figure. Interestingly, the structural perturbation
involves not only the PNA strand but also the DNA strand
(compare the blue structure with the red and green ones).
An additional interesting feature of the simulations arose from
plotting the minor-groove widths at one of the modification sites
as a function of time (Figure 4). The trajectories are stable over
the last 3 ns of the simulation and the narrower groove for PLD
is evident, but what is striking about the data is the significantly
smaller fluctuations in the groove width for this duplex. This
suggests that the L-leucine modifications increase the rigidity
of the duplex while the D-leucine modification has no effect.
Assembly of Cyanine Dye Aggregates. Aggregation of
cyanine dyes on PNA-DNA templates is readily monitored by
optical spectroscopy.11,37 Both UV-vis and circular dichroic
(CD) spectra show blue-shifted bands not observed in the
absence of the PNA-DNA duplex. This effect is illustrated in
the UV-vis spectra of Figure 5A, where increasing amounts
of the dicarbocyanine dye DiSC2(5) are added into a solution
of 5 µM PD. The new absorption band at 543 nm is diagnostic
for successful aggregation of the dye on the template. At low
dye concentrations, the absorption spectrum is similar to the
absorption spectrum of the dye in methanol. At higher dye
concentrations, the new peak at 543 nm is prominent. We
previously attributed this concentration dependence to the high
stoichiometry (g6 dyes/duplex) and cooperative nature of the
aggregation process.11
Introduction of two L-leucine modifications into the PNA
strand significantly hinders aggregation of the dye (Figure 5B).
While the aggregate peak at 543 nm is quite strong for the PD,
(36) Wang, M.; Dilek, I.; Armitage, B. A. Langmuir 2003, 19, 6449-6455.
(37) West, W.; Pearce, S. J. Phys. Chem. 1965, 69, 1894-1903.
(38) Nakanishi, K.; Berova, N.; Woody, R. W. Circular Dichroism: Principles
and Applications; VCH Publishers: New York, 1994.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3343

A R T I C L E S
Dilek et al.
Figure 5. UV-Vis titration of DiSC2(5) into (A) PD or (B) PLD template duplexes. [PNA-DNA] ) 5.0 µM; dye was added in 1.0 µM aliquots. Spectra
were recorded at 15 °C.
Figure 6. Comparison of UV-vis spectra recorded for DiSC2(5) in the
presence of the three PNA-DNA duplexes. [PNA-DNA] ) 5.0 µM;
[DiSC2(5)] ) 10 µM. Spectra were recorded at 15 °C.
Figure 7. Comparison of CD spectra recorded for DiSC2(5) in the presence
of the three PNA-DNA duplexes. [PNA-DNA] ) 5.0 µM; [DiSC2(5)] )
10 µM. Four scans were collected at a scan rate of 200 nm/min and averaged.
Spectra were recorded at 15 °C.
6. Our previous work showed that the stability of the aggregate
is strongly dependent on the dye concentration.11
Interestingly, the molecular dynamics simulations suggest that
the effect of the isobutyl group is not only to block access to
the groove but also to induce changes in both the width and
the rigidity of the groove. In addition to our earlier report that
blockage of the major groove by use of a second PNA strand
to form a PNA2-DNA triplex has almost no effect on dye
aggregation, these results provide further evidence that the dye
molecules aggregate by using the minor groove of the PNA-
DNA double helix as a template.
The fact that aggregation on the D-leucine-derived template
is not restored to that of the unmodified duplex could be due to
several factors, including differential hydration of the unmodi-
fied and D-leucine-derived duplexes, residual steric blockage
by the isobutyl group even in the D-configuration, or a local
change in the helical geometry. The altered helical parameters
obtained from the MD simulations for PDD versus PD (Table
1) are consistent with the last option. An additional factor to
consider is the possibility of racemization of the chiral PNA
monomer during synthesis. Corradini, Nielsen, and co-work-
ers39,40 recently reported a systematic study of racemization in
which they varied the conditions for forming the active ester
Discussion
To test our earlier conjecture that DiSC2(5) uses the minor
groove of PNA-containing duplexes and triplexes as a template
for assembling a dye aggregate, we synthesized PNA oligomers
based on chiral monomers that project alkyl groups either into
or out of the minor groove. The leucine side chains we
incorporated introduced thermal destabilizations of 4.1-4.2 °C/
modification relative to a glycine-based PNA, independent of
the configuration. It is difficult to compare this with other results
in the literature, since different sequences, numbers, and
positions of modifications and terminal charges were used.
However, while the destabilizations we observe in this work
are larger than those reported previously, the PLD and PDD
duplexes are fully formed at the temperatures at which dye-
binding experiments were performed.
The UV-vis and CD results described above clearly dem-
onstrate that incorporation of L-leucine-derived PNA monomers
into a PNA-DNA duplex significantly hinders cyanine dye
aggregation, whereas the D-leucine analogue is less inhibitory.
9
3344 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Cyanine Dye Binding to Modified PNA−DNA Duplexes
A R T I C L E S
Figure 8. Plot of the induced CD intensity at 521 nm as a function of
DiSC2(5) concentration in the presence of PD, PDD, and PLD (5.0 µM).
Data points and error bars represent the mean and standard deviation of
three separate experiments.
Figure 9. Temperature-dependent absorbance curves recorded at 543 nm
for DiSC2(5) (10 µM) in the presence of PD, PDD, and PLD (5.0 µM).
Cooling curves were collected from 40 to 5 °C at a rate of 0.5 °C/min.
prior to coupling of the chiral monomer onto the growing PNA
oligomer. While leucine-modified monomers were not inves-
tigated, a phenylalanine-derived monomer underwent 5-6%
racemization under conditions similar to what we used for our
couplings. Thus, we estimate that our PNA strands are ap-
proximately 90% either L,L- or D,D-configuration. The presence
of a minor amount of L-leucine-containing PNA in the otherwise
D,D-configuration duplexes would account at least in part for
the inability to restore dye aggregation to the level observed
on the unmodified template.
In conclusion, the experiments reported here are consistent
with a molecular recognition mechanism in which cyanine dyes
assemble into helical aggregates by using the minor groove of
a PNA-DNA duplex as a template. This is distinct from minor-
groove recognition of duplex DNA in that the dyes most likely
do not penetrate to the floor of the groove, where sequence-
dependent structural features would be most evident. Rather,
in this model the dyes are bound rather loosely by the groove,
as evidenced by the relatively low temperatures required to
dissociate the aggregate.
Acknowledgment. This work was supported by grants from
Research Corporation and the National Science Foundation
(CHE-0315925) to B.A.A. M.M. thanks M. Orozco and R.
Soliva for sharing their PNA parameters. All simulations were
performed on the Terascale System at the Pittsburgh Super-
computing Center supported by NSF CHE-020022. Mass spectra
were measured in the Center for Molecular Analysis at Carnegie
Mellon University, supported by NSF CHE-9808188.
Supporting Information Available: Experimental details of
chiral PNA monomer synthesis (PDF). This information is
available free of charge via the Internet at http://pubs.acs.org.
(39) Tedeschi, T.; Corradini, R.; Marchelli, R.; Pushl, A.; Nielsen, P. E.
Tetrahedron: Asymmetry 2002, 13, 1629-1636.
(40) A recently reported submonomeric strategy for synthesizing PNAs with
chiral backbones affords significantly improved optical purity: Sforza, S.;
Tedeschi, T.; Corradini, R.; Ciavardelli, D.; Dossena, A.; Marchelli, R.
Eur. J. Org. Chem. 2003, 1056-1063.
JA045145A
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3345