A simple γ-backbone modification preorganizes peptide nucleic acid into a helical structure

Article abstract of DOI:10.1021/ja0625576

Peptide nucleic acid (PNA) is a synthetic analogue of DNA and RNA, developed more than a decade ago in which the naturally occurring sugar phosphate backbone has been replaced by the N-(2-aminoethyl) glycine units. Unlike DNA or RNA in the unhybridized st

Full text of DOI:10.1021/ja0625576

Published on Web 07/19/2006
A Simple γ-Backbone Modification Preorganizes
Peptide Nucleic Acid into a Helical Structure
Anca Dragulescu-Andrasi, Srinivas Rapireddy, Brian M. Frezza,†
Chakicherla Gayathri, Roberto R. Gil,* and Danith H. Ly*
Contribution from the Department of Chemistry, Carnegie Mellon UniVersity,
4400 Fifth AVenue, Pittsburgh, PennsylVania 15213
Received April 12, 2006; E-mail: dly@andrew.cmu.edu; rgil@andrew.cmu.edu
Abstract: Peptide nucleic acid (PNA) is a synthetic analogue of DNA and RNA, developed more than a
decade ago in which the naturally occurring sugar phosphate backbone has been replaced by the N-(2-
aminoethyl) glycine units. Unlike DNA or RNA in the unhybridized state (single strand) which can adopt a
helical structure through base-stacking, although highly flexible, PNA does not have a well-defined con-
formational folding in solution. Herein, we show that a simple backbone modification at the γ-position of
the N-(2-aminoethyl) glycine unit can transform a randomly folded PNA into a helical structure. Spectroscopic
studies showed that helical induction occurs in the C- to N-terminal direction and is sterically driven. This
finding has important implication not only on the future design of nucleic acid mimics but also on the design
of novel materials, where molecular organization and efficient electronic coupling are desired.
Introduction
PNA is a synthetic analogue of DNA and RNA, developed
more than a decade ago, in which the naturally occurring sugar
phosphate backbone has been replaced by the N-(2-aminoethyl)
glycine units, Scheme 1.10 PNA can hybridize to complementary
DNA or RNA strand through Watson-Crick base-pairing;
because of the unnatural backbone, PNA can neither be
recognized nor easily degraded by proteases or nucleasessall
of which make PNA an attractive reagent for biotechnology
applications.11 However, unlike DNA or RNA in the unhybrid-
ized state (single strand) whose structure, to a large degree, is
extended in solution due to the negatively charged phosphate
backbone, PNA tends to fold into complex globular struc-
tures,12,13 presumably due to the collapse of the hydrophobic
nucleobases. In fact, this conformational collapse has been
exploited in the development of stemless PNA molecular
beacons, taking advantage of the proximity between the two
termini in the unhybridized state.12,14-16 Several modifica-
tions have been made to the PNA N-(2-aminoethyl) glycine
backbone in attempts to increase its rigidity;17,18 however, only
a few such modifications showed improvements in the hybrid-
ization properties19-24smany of which require either elaborate
Biotechnology applications, for the most part, are developed
on the basis of the ability of specific substrates to bind to cellular
targets. Such binding would either interfere with the physio-
logical function of the targets and/or elicit the reporter signals.
Establishing high binding affinity and selectivity between the
substrates and the targets is crucial to the success of many of
these applications. One way to achieve this goal would be to
preorganize the substrates into the binding conformation prior
to complexation. This would minimize the entropic penalty asso-
ciated with most molecular recognition events and increase the
rigidity of the substratessthereby providing both stability and
selectivity.1 While this concept has been successfully applied
to the design of small molecules for guest-host recognition,2
it has been less successful with the design of larger and more
flexible oligomeric molecules such as peptides and nucleic acids
and their synthetic mimics, due to the large degree of freedoms
although some progress has been made.3-9 In this article we
show that a simple γ-backbone modification can transform a
randomly folded peptide nucleic acid (PNA) into a right-handed
helix. These conformationally preorganized helical PNAs bind
to DNA and RNA with exceptionally high affinity and sequence
selectivity.
(10) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science 1991, 254,
1497-1500.
(11) Nielsen, P. E. Acc. Chem. Res. 1999, 32, 624-630.
(12) Gildea, B. D.; Coull, J. M.; Hyldig-Nielsen, J. J.; Fiandaca, M. J. WO
A-9921881, 1999.
† Present address: The Scripps Research Institute, 10550 North Torrey
Pines Road, La Jolla, CA 92037.
(1) Cram, D. J. CHEMTECH 1987, 17, 120-125.
(2) Cram, D. J. Science 1988, 240, 760-767.
(3) Kool, E. T. Chem. ReV. 1997, 97, 1473-1487.
(4) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173-180.
(5) Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. Chem. ReV. 2001, 101,
3219-3232.
(13) Tackett, A. J.; Corey, D. R.; Raney, K. D. Nucleic Acids Res. 2002, 30,
950-957.
(14) Seitz, O. Angew. Chem., Int. Ed. 2000, 39, 3249-3252.
(15) Ranasinghe, R. T.; Brown, L. J.; Brown, T. Chem. Commun. 2001, 1480-
1481.
(16) Kuhn, H.; Demidov, V. V.; Coull, J. M.; Fiandaca, M. J.; Gildea, B. D.;
Frank-Kameneteskii, M. D. J. Am. Chem. Soc. 2002, 124, 1097-1103.
(17) Beck, F.; Nielsen, P. E. Artificial DNA: methods and applications; CRC
Press: Boca Raton, 2003.
(18) Kumar, V. A.; Ganesh, K. N. Acc. Chem. Res. 2005, 38, 404-412.
(19) Pokorski, J. K.; Witschi, M. A.; Purnell, B. L.; Appella, D. H. J. Am. Chem.
Soc. 2004, 126, 15067-15073.
(6) Dervan, P. B.; Edelson, B. S. Curr. Opin. Struct. Biol. 2003, 13, 284-
299.
(7) Barron, A. E.; Zuckermann, R. N. Curr. Opin. Chem. Biol. 1999, 3, 681-
687.
(8) Zondlo, N. J.; Schepartz, A. J. Am. Chem. Soc. 1999, 121, 6938-6939.
(9) Wengel, J. Acc. Chem. Res. 1999, 32, 301-310.
9
10258
J. AM. CHEM. SOC. 2006, 128, 10258-10267
10.1021/ja0625576 CCC: $33.50 © 2006 American Chemical Society

Transformation of PNA into a Helical Structure
A R T I C L E S
Scheme 1. Chemical Structure of DNA (RNA), PNA and L-serine
Derived γPNA (LS-γPNA)
synthesis19-22 or the use of relatively expensive D-amino acids
as starting materials.23,24 Of the various backbone modifications
that have been made, only a few were made at the γ-position.25-27
A systematic study correlating structure with function at this
position has not yet been established. To fill this void, we have
developed a research program to explore the structural effects
of γ-backbone modifications on the conformation and hybrid-
ization properties of PNA.
Figure 1. CD spectra of single-stranded PNA (P1) and LS-γPNAs (P2-
P5). Bold letters denote the LS-γPNA units shown in Scheme 1. All samples
were prepared in sodium phosphate buffer (10 mM sodium phosphate, 100
mM NaCl, 0.1 mM EDTA) at 2 µM strand concentration each. The CD
spectra were recorded at 22 ° C unless otherwise stated.
Results and Discussion
of right-handed helices.31 The CD profiles gradually shifted with
increasing number of modified backbone units, from ones that
resemble right-handed PNA-PNAs30 to ones that resemble
PNA-DNA helices32sas characterized by the negative band
at the 240 nm and a red-shift in the 215 nm maximum. The
most significant changes occurred in the 200-230 nm regions,
with a dramatic increase in the amplitude of the 210 nm band.
This spectral region has been previously assigned to the n-π*
transition of the amides in the backbone.33 The shift in the
spectrum could be attributed to the variation in the backbone
conformation, reflecting the helical twist of the oligomers. As
the number of modified backbone units increases, the helical
twist of the oligomer as a whole becomes more pronounced.
Rationale. We decided to incorporate L-serine side chain at
the γ-position of the PNA backbone because MD simulations
showed that this configuration is less disruptive to hybridization
than the D-configuration, and we reasoned that, since the serine
side chains (OH groups) can form hydrogen bonds (H-bond)
with water molecules, their inclusion could enhance the solu-
bility of PNA in aqueous solution. PNA monomers and oli-
gomers containing L-serine side chains at the γ-position (LS-
γPNAs) were synthesized (Scheme S1, Figures S1 and S2:
Supporting Information), and their conformation and hybridiza-
tion properties were characterized by CD, UV-vis, and multi-
dimensional NMR spectroscopy. The sequence shown in P1-
P5 (Figure 1) was chosen for this study because the hybridization
properties of this particular PNA (P1) oligomer have been well-
characterized.28
The Effects of γ-Backbone Modification on the Confor-
mation of PNA. A comparison of the CD spectra of PNA (P1)
and LS-γPNAs (P2-P5) is shown in Figure 1. No exciton
coupling pattern was observed for P1 in the nucleobase absorp-
tion regions (220-300 nm). This generally indicates the lack
of either helical base-stacking or excess helical sense (an equal
mixture of right-handed and left-handed helical structures)s
neither of which induces CD signal in this region. Although
possible,29 the second scenario is unlikely based on experimental
findings.30 On the other hand, we noticed distinct CD signals
for P2-P5, with biphasic exciton coupling patterns characteristic
The Effects of Temperature and Concentration on the
Exciton Coupling Patterns. To ensure that the observed exciton
coupling occurs as the result of molecular preorganization (not
intermolecular interaction), we performed temperature- and
concentration-dependent CD measurements on P5. Our results
showed that the intensities of the CD signals gradually decreased
with increasing temperature, Figure 2. The exciton coupling
pattern disappeared at around 60 °Csat which point the CD
profile of P5 resembles that of P1 recorded at 22 °C. However,
upon cooling, the CD signal returned to the original state,
indicating that the observed exciton coupling is the intrinsic
property of P5 and not the result of some artifact. We ruled out
the possibility of intermolecular interaction, since the concentra-
tion-dependent study showed that the amplitude of the CD
signals increased linearly with increasing concentration over a
100-fold range, Figure 3. Helical induction, inferred from the
exciton coupling patterns, appears to be general to all serine-
derived γPNAs since similar CD spectra were observed for other
oligomers with different nucleobase sequence composition
(Figure S3: Supporting Information).
(20) Lagriffoule, P.; Wittung, P.; Eriksson, M.; Jensen, K. K.; Norden, B.;
Buchardt, O.; Nielsen, P. E. Chem. Eur. J. 1997, 3, 912-919.
(21) Dueholm, K. L.; Petersen, K. H.; Jensen, D. K.; Egholm, M.; Nielsen, P.
E.; Buchardt, O. In Bioorg. Med. Chem. Lett. 1994, 4, 1077-1080.
(22) Govindaraju, T.; Kumar, V. A.; Ganesh, K. N. J. Am. Chem. Soc. 2005,
127, 4144-4145.
(23) Haaima, G.; Lohse, A.; Buchardt, O.; Nielsen, P. E. Angew. Chem., Int.
Ed. Engl. 1996, 35, 1939-1942.
(24) Sforza, S.; Corradini, R.; Ghirardi, S.; Dossena, A.; Marchelli, R. Eur. J.
Org. Chem. 2000, 2905-2913.
The Effects of r- and γ-Backbone Modification on the
Conformation of PNA. This is not the first time that exciton
(25) Englund, E. A.; Appella, D. H. Org. Lett. 2005, 7, 3465-3467.
(26) Tedeschi, T.; Sforza, S.; Corradini, R.; Marchelli, R. Tetrahedron Lett. 2005,
46, 8395-8399.
(27) Dose, C.; Seitz, O. Org. Lett. 2005, 7, 4365-4368.
(28) Demidov, V. V.; Protozanova, E.; Izvolsky, K. I.; Price, C.; Nielsen, P.
E.; Frank-Kamenetskii, M. D. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5953-
5958.
(31) Johnson, W. C. In Circular Dichroism: Principles and Applications; Wiley-
VCH: New York, 2000.
(32) Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.; Freier, S. M.;
Driver, D. A.; Berg, R. H.; Kim, S. K.; Norden, B.; Nielsen, P. E. Nature
1993, 365, 566-568.
(29) Sen, S.; Nilsson, L. J. Am. Chem. Soc. 2001, 123, 7414-7422.
(30) Wittung, P.; Nielsen, P. E.; Buchardt, O.; Egholm, M.; Norden, B. Nature
1994, 368, 561-563.
(33) Nielsen, E. B.; Schellman, J. A. J. Phys. Chem. 1967, 71, 2297-2304.
9
J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006 10259

A R T I C L E S
Dragulescu-Andrasi et al.
Figure 2. Temperature-dependent CD profiles of LS-γPNA (P5) at 2 µM
strand concentration, prepared in sodium phosphate buffer. Arrows show
the direction of increasing temperature.
Figure 4. CD spectra of PNA containing R- and γ-backbone modifications.
This particular oligomer, which comprised all four nucleobases, was chosen
for this study as a model system because of the simplicity in the synthesis.
The concentration of each strand was 5 µM in sodium phosphate buffer;
CD spectra were recorded at 22 °C.
Figure 3. CD spectra of P5 oligomer at various concentrations. The samples
were prepared in sodium phosphate buffer and recorded at 22 °C. (Inset)
Plot of CD amplitude (A) vs P5 concentration.
Figure 5. Spectroscopic evidence showing that helical induction occurs
in the C- to N-terminal direction. The concentration of each strand was
15 µM in sodium phosphate buffer; CD spectra were recorded at 22 °C.
coupling between neighboring nucleobases has been observed
with chiral PNAs. Sforza and co-workers34 showed that PNAs
containing R-backbone modifications exhibited distinct CD
patterns with the characteristic profile of a PNA-PNA double
helix. However, it should be noted that the CD signals of these
The Effects of γ-Backbone Modification on the Direc-
tionality of Helical Induction. Interestingly, we noticed the
absence of CD signals when the chiral unit was switched from
the C to the N terminus (compare the CD profile of P9 in Fig-
ure 5 to that of P6 in Figure 4). Notice that the concentration
of P9 in Figure 5 is 3 times higher than that of P6 in Figure 4,
but still no CD signal was observed for P9. Addition of an abasic
glycine unit (P10) had no effect on the CD profile (still no CD
signals). However, to our surprise we noticed that the CD signal
characteristic of a right-handed PNA-DNA helix reappeared
upon addition of an achiral PNA unit containing adenine
nucleobase (P11). The amplitude of the CD signal further
increased with additional achiral PNA units (compare the CD
profiles of P12, P13, and P14 to that of P11 and P9). The
difference in the signal intensities of P12, P13, and P14 at the
260-290 nm regions could be attributed to the difference in
the extinction coefficient of the nucleobases, since different
nucleobases were added to each oligomer. Together these results
suggest that helical induction is unidirectional, occurring in the
C to N terminus. Helical induction, however, does not appear
L
R-PNAs are weak compared to those of S-γPNAs. A direct
comparison can be seen in Figure 4, which shows that PNA
containing L-serine side chain at the R-position (P6) did not
show exciton coupling at 5 µM concentration as compared to
PNA with the same side chain placed at the γ-position (P7),
which showed pronounced Cotton effects characteristic of a
right-handed PNA-DNA helix. The fact that the CD profile of
P7, which contains one chiral unit, and that of P8, which con-
tains four chiral units, are similar to one another, in both shape
and amplitude in the nucleobase absorption regions, suggests
that a single chiral unit may be sufficient to preorganize the
entire oligomer into a helical structuresakin to the “sergeants
and soldiers” concept developed by Green and co-workers35 to
explain the effects of chiral induction in polymers.
(34) Sforza, S.; Haaima, G.; Marchelli, R.; Nielsen, P. E. Eur. J. Org. Chem.
1999, 197, 7-204.
(35) Green, M. M.; Cheon, K.-S.; Yang, S.-Y.; Park, J.-W.; Swansburg, S.; Liu,
W. Acc. Chem. Res. 2001, 34, 672-680.
9
10260 J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006

Transformation of PNA into a Helical Structure
A R T I C L E S
Scheme 2. Chemical Structurea of a Model (CT) PNA, L-Serine-Derived (LSer-γPNA) and L-Alanine-Derived (LAla-γPNA) γPNA System
a The sequence is written from N to C terminus.
Scheme 3. Chemical Structures of the Four (CT) PNA Rotamers
to be specific to the serine side chain since other amino acid
side chains incorporated at the γ-position, including cysteine
and alanine, showed similar exciton coupling patterns (data not
shown).
N4 bond (tertiary amide bond), Scheme 3.37-39 This character-
istic of PNA monomers leads to signal duplication in the H
1
NMR spectrum. The NMR spectrum gets more complicated as
the number of monomeric units increases because the number
of rotamers in solution is equal to 2n, where n is the number of
monomeric units in the PNA strand. For a dimer, 22 ) 4 possible
rotamers are expected to be present in solution. Scheme 3 shows
the four possible rotamers for the (CT) PNA dimer. In fact,
this is what was observed in the 1H NMR spectrum. The
spectrum showed four sets of signals with different intensities,
corresponding to the aromatic protons in the bases C-H5,
C-H6, TCH3, and T-H6 (Figure 6, insets A-D). To complete
the NMR assignment, the following strategy was used. The
cytosine and thymine H-6 aromatic protons in each rotamer were
used as the starting points, NOE cross-peaks in the NOESY
spectrum led to signals corresponding to the methylene protons
H-8 and H-8′. These signals in turn, led to another set of signals
corresponding to H-5 and H-5′ via a common long-range
proton-carbon correlation with the carbonyl carbon C-7 and
C-7′ in the HMBC spectrum, respectively. This correlation is
always present regardless of the geometry of the C7-N4 bond.
To determine the geometry of the tertiary amide bond, the
Solution Structure of a Model System. To gain further
insight into the conformation of γPNAs, we performed a series
of multinuclear and multidimensional NMR experiments on a
model (CT) PNA, L-serine-derived (LSer-γPNA) and L-alanine-
derived (LAla-γPNA)36 γPNA system (Scheme 2) using a
Bruker Avance DMX-500 instrument operating at 500.13 MHz
for 1H and 125.77 MHz for 13C. However, only PNA and
LAla-γPNA dimers were examined in detail because LSer-γPNA
showed extensive signal overlap and many of the characteristic
spectral features observed in LSer-γPNA were also observed in
LAla-γPNA, indicating similar conformational folding between
the two systems. The data were collected on samples dissolved
in D2O. The full NMR assignments were performed using the
following experiments: 1-D 1H NMR, 1H,1H correlation spec-
troscopy (COSY), 1H,1H nuclear Overhauser spectroscopy
(NOESY), 1H,13C heteronuclear single-quantum correlation
spectroscopy (HSQC), and 1H,13C selective heteronuclear multi-
ple bond correlation spectroscopy (SHMBC).
Solution Structure of (CT) PNA Dimer. It is well-known
that the restricted rotation around the tertiary amide bond in
PNA produces two possible rotamers in solution around the C7-
(37) Dueholm, K. L.; Egholm, M.; Behrens, C.; Christensen, L.; Hansen, H. F.;
Vulpius, T.; Petersen, K. H.; Berg, R. H.; Nielsen, P. E.; Buchardt, O. J.
Org. Chem. 1994, 59, 5767-5773.
(38) Krotz, A. H.; Buchardt, O.; Nielsen, P. E. Tetrahedron Lett. 1995, 36,
6937-6940.
(36) Kosynkina, L.; Wang, W.; Liang, T. C. Tetrahedron Lett. 1994, 35, 5173-
(39) Chen, S. M.; Mohan, V.; Kiely, J. S.; Griffith, M. C.; Griffey, R. H.
Tetrahedron Lett. 1994, 35, 5105-5108.
5176.
9
J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006 10261

A R T I C L E S
Dragulescu-Andrasi et al.
Figure 6. 1H NMR spectrum of the (CT) PNA dimer in D2O at 500 MHz.
Table 1. 1H NMR Assignments of (CT) PNA Dimer (500.13 MHz - D2O)a
H
A
B
C
D
H
A
B
C
D
2
3
5
8
3.29 t
3.80 s
4.00 s
4.88 s
6.11 d
7.70 d
3.16 t
3.68 s
5.16 s
4.60 s
6.11 d
7.64 d
3.30 t
3.81 s
4.11 s
4.85 s
6.04 d
7.62 d
3.14 t
3.67 s
4.35 s
4.56 s
6.12 d
7.61 d
2′
3′
5′
8′
T-H6
T-CH3
3.38 t
3.50 t
4.18 s
4.54 s
7.32 bq
1.80 d
3.37 t
3.61 t
4.24 s
4.57 s
7.27 bq
1.76 d
3.45 t
3.54 t
4.04 s
4.71 s
7.34 bq
1.77 d
3.43 t
3.56 t
4.08 s
4.76 s
7.30 bq
1.81 d
C-H5
C-H6
a J2,3 and J2′,3′ ≈ 6 Hz; JC-H5,C-H6 ) 8 Hz; JT-H6,T-CH3 ) 1 Hz; bq ) broad quartet.
Scheme 4. Newman Projection of a Right-Handed and
Left-Handed Helical (CT) PNA
NOESY spectrum was used to observe the NOE cross-
correlation peaks between H-8 and H-3 or H-8′ and H-3′ if the
carbonyl C-7 were to be pointing toward the C terminus, or
NOE cross-correlation peaks between H-8 and H-5 or H-8′ and
H-5′ if the C-7 carbonyl group were to be pointing toward the
N terminus. The full assignments for all four (CT) PNA rotamers
are tabulated in Table 1. In the 1H NMR spectrum, all the signals
corresponding to the methylene protons H-5, H-8, H-5′, and
H-8′ appeared as singlets (Figure 6, inset E), and all the signals
corresponding to the methylene protons H-2, H-3, H-2′, and
H-3′ appeared as triplets with a coupling constant J2-3 ≈ 6 Hz
(Figure 6, inset F). The degeneracy in chemical shift of each
pair of protons of the methylene groups clearly indicates that
the structure of the single-stranded (CT) PNA dimer is random
coil. Only free rotation around all the bonds can produce singlets
for the isolated methylene groups (H-5, H-8, H-5′, H-8′) and
triplets with average vicinal coupling constant of ∼6 Hz for
the vicinal methylene groups (H-2, H-3, H-2′, H-3′). The
degeneracy on the chemical shift of the methylene protons was
also observed in the one-bond proton-carbon heteronuclear
correlations in the 1H,13C HSQC experiments (not shown). Only
one signal per CH2 protons correlates with its corresponding
carbon signal.
Consistent with the X-ray crystal structure of PNA-DNA,40
the NMR solution structure of PNA-RNA,41 and computer
models of (CT) PNA dimer generated in our group, the torsion
angle formed by N1′-C2′-C3′-N4′ should adopt values in the
range of +50° to +60° for the structure to form a right-handed
(40) Brown, S. C.; Thomson, S. A.; Veal, J. M.; Davis, D. G. Science 1994,
265, 777-780.
(41) Menchise, V.; De Simone, G.; Tedeschi, T.; Corradini, R.; Sforza, S.;
Marchelli, R.; Capasso, D.; Saviano, M.; Pedone, C. Proc. Natl. Acad. Sci.
U.S.A. 2003, 100, 12021-12026.
9
10262 J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006

Transformation of PNA into a Helical Structure
A R T I C L E S
Figure 7. 1H NMR spectrum of (CLT) LAla-γPNA (top trace) and PNA (bottom trace) in D2O at 500 MHz. Expansion of the backbone CH2 proton signals.
Scheme 5. Chemical Structures of the Four (CLT) LAla-γPNA Rotamers
helix. An opposite sign of the dihedral angle is needed to form
a left-handed helix, Scheme 4. Although PNA is achiral, once
both the right-handed and left-handed helical structures are
formed, they are mirror images of one another, and the NMR
signals should be the same for both.
If both the right-handed and left-handed helices were to be
present in solution, the signals corresponding to H-2′ and H-3′
should not appear as simple triplets with J couplings of ∼6 Hz.
Each proton at each CH2 should not be chemically equivalent.
Instead, the NMR spectrum should show four complex multi-
plets for H-2′a, H-2′b, H-3′a and H-3′b. On the basis of the
dihedral angles involving these protons and a rough estimation
of J values using the Karplus equation, the NMR spectrum
should show two doublets of doublet of doublet (ddd) of 14,
10, and 3 Hz for a right-handed helix, and two doublets of
doublet of doublet (ddd) of 14, 3, and 3 Hz for a left-handed
helixswhere the 14 Hz corresponds to the geminal coupling
constant, 10 Hz to the trans diaxial vicinal coupling constant,
and 3 Hz for a coupling constant for a torsion angle of ∼+60°.
These signals, however, were not observedsruling out the possi-
9
J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006 10263

A R T I C L E S
Dragulescu-Andrasi et al.
Table 2. 1H NMR Assignments of (CLT) LAla-γPNA (500.13 MHz - D2O)a
H
A
B
C
D
H
A
B
C
D
2
3a
3b
5a
5b
8a
8b
C-H5
C-H6
2′
3.30 t
3.16 t
3.27 t
3.13 t
3′a
3′b
5′a
5′b
8′a
8′b
T-H6
T-CH3
Ala CH3
3.61 dd
3.18 dd
4.20 d
4.08 d
4.59 d
4.37 d
7.32 bq
1.79 bd
1.09 d
3.70 dd
3.30 dd
4.27 d
4.27 d
4.67 d
4.44 d
7.28 bq
1.77 bd
1.16 d
3.49 dd
3.45 dd
4.28 d
3.94 d
4.73 d
4.73 d
7.33 bq
1.79 bd
1.18 d
3.48 dd
3.40 dd
4.26 d
3.92 d
5.06 d
4.55 d
7.29 bq
1.80 bd
1.20 d
3.88 m
3.74 m
4.13 d
4.08 d
4.87 d
4.77 d
6.01 d
7.62 d
4.21 m
3.77 m
3.56 m
4.18 d
4.09 d
4.69 d
4.53 d
6.05 d
7.57 d
4.15 m
3.81 m
3.79 m
3.85 d
4.09 d
4.87 d
4.75 d
6.03 d
7.59 d
4.22 m
3.69 m
3.62 m
4.44 d
4.28 d
4.59 d
4.32 d
6.06 d
7.51 d
4.34 m
a J2,3 ≈ 6 Hz; J5a,5b ) J5′a,5′b ) J8a,8b ) J8′a,8′b ≈ 17 Hz; J2′, AlaCH3 ) 6.7 Hz; J2′,3′a ) 10.2 Hz; J2′,3′b ) 4.2 Hz; J3′a,3′b ) 15.2 Hz; JC-H5,C-H6
) 8 Hz; JT-H6,T-CH3 ) 1 Hz; bd ) broad doublet; bq ) broad quartet.
Figure 8. Double-quantum-filtered-phase-sensitive COSY of (CLT) LAla-γPNA in D2O at 500 MHz.
bility that these dimers existed in an equal population of right-
handed and left-handed helices.
H-5′, and H-8′ no longer appeared as singlets. Each methylene
proton is no longer chemically equivalent, and they appeared
as 17 Hz doublets (geminal coupling constant). The most inter-
esting part of the 1H NMR spectrum is the signals displayed by
the spin system composed of H-2′, H-3′a, H-3′b, and the CH3
group attached to C-2′. Due to the significant signal overlapping
in the 1D 1H NMR spectrum, a double-quantum-filtered-phase
sensitive COSY experiment was performed to resolve and
identify the coupling pattern of this spin system. Four sets of
signals were observed due to the presence of four rotamers,
Figure 8. Proton H-2′ appeared as a complex multiplet due to
coupling to the CH3 and to protons H-3′a and H-3′b. The signal
of H-3′a appeared as doublet of doublet with coupling constants
of 15.2 and 10.2 Hz, while H-3′b appeared as a doublet of
doublet with coupling constants of 15.2 and 4.2 Hz (Figure 9,
top left trace). This indicates that the presence of a preferred
conformation in solution that involves one trans diaxial inter-
1
Solution Structure of (CLT) LAla-γPNA Dimer. The H
L
NMR spectrum of the Ala-γPNA dimer, on the other hand,
showed an entirely different set of signals (Figure 7, top trace).
The structures of the four rotamers are shown in Scheme 5 and
the 1H NMR assignments in Table 2. The degeneracy in chem-
ical shift is broken for all of the methylene groups except for
H-2, which freely rotates at the N terminus and produced a triplet
with an average coupling constant of ∼6 Hz (signals at 3.1-
3.2 ppm in Figure 7, top trace). Figure 7 clearly shows the
1
L
differences between the H NMR spectra of Ala-γPNA and
PNA (Figure 7, top and bottom trace, respectively). Similar to
1
that of the PNA dimer, the H NMR spectrum of the LAla-
γPNA dimer showed signals corresponding to the four rotamers
L
in solution (Figure 7, top trace). The spectrum of Ala-γPNA,
however, is more complex. The methylene protons H-5, H-8,
9
10264 J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006

Transformation of PNA into a Helical Structure
A R T I C L E S
Figure 9. 1H NMR spectrum of (CT) LAla-γPNA in D2O at 500 MHz. Base aromatic signals. (Top left) H-3′a and H-3′b signals of the major rotameric
isomer. (Top right trace) signals (doublets) of the CH3 group at C-2′.
Scheme 6. Newman Projection of a Right-Handed and
in this case the CH3 group, prefers a trans diaxial relationship
with N4′ to minimize steric clash. In a left-handed helical con-
Left-Handed (CLT) LAla-γPNA
formation the gauche interaction between CH3 and N4′ increases
the energy of the system. AM1 semiempirical calculations of
the (CLT) LAla-γPNA dimer in both the left- and right-handed
conformations showed a difference in the heat of formation of
3.5 kcal/mol in favor of the right-handed helix. Assuming a zero
change in entropy between the two conformations, the ∆H is
equal to the ∆G. With a ∆∆G of 3.5 kcal/mol, more than 99%
of the dimer should be in the right-handed helical conformation.
The NMR data indicate that the four rotamers of (CLT)
action between H-2′ and H-3′a (J of 10.2 Hz) (Scheme 6, left
LAla-γPNA are in a right-handed helical conformation. The
structure) and an axial to equatorial interaction between H-2′
solution structure of LAla-γPNA obtained from NMR analysis
and H-3′b (J of 4.2 Hz). Given that the absolute configuration
is surprisingly similar to that of the unmodified PNA in a PNA-
DNA hybrid duplex, as determined by X-ray crystallography
(compare the two structures in Figure 10).41 (Note the similarity
in the backbone conformation.) These NMR results provide
strong evidence for the existence of γPNA exclusively in a right-
handed helical conformationsresembling that of PNA in PNA-
DNA and PNA-RNA double helix. On the basis of these results
we can anticipate that any substitution at C-2′ (γ-position) will
lead to a preferred, helical preorganization of the backbones
the helical sense of which will be determined by the stereo-
chemistry at this position. The S configuration will prefer a right-
handed helix, while the R configuration will prefer a left-handed
helix. The substituent C-2′ appears to be the determining factor
that directs the helical folding of γPNAs. We suggest that helical
induction occurs as the result of steric clash in the backbone,
between substituent at the C-2′ (γ-position) and N4′, followed
at C-2′ is S, the only conformation that produces this relative
spatial arrangement of protons H-2′, H-3′a, and H-3′b is the
one that leads to the formation of a right-handed helix. No
evidence for the formation of a left-handed helix was observed
in the NMR spectrum. If a left-handed helix were to be present
in solution, the coupling pattern of H3′a and H3′b would be
the same. Both should show a small vicinal coupling constant
of ∼4 Hz leading to a doublet of doublet of ∼15 and 4 Hz for
both protons. Furthermore, the spectrum should also show
methyl signals for a left-handed helical conformation. In this
case, a total of eight methyl signals should appear in the 1D 1H
NMR spectrum; however, only four methyl signals were
observed (Figure 9, top right trace).
The helical sense of γPNAs, in this case, is determined by
the stereochemistry at the γ-position (C-2′). The S configuration
prefers a right-handed helix because the amino acid side chain,
9
J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006 10265

A R T I C L E S
Dragulescu-Andrasi et al.
Figure 10. (A) X-ray structure of PNA in a PNA-DNA duplex41 and (B) solution structure of (CLT) LAla-γPNA determined by NMR.
Table 3. Melting Transitions (Tm’s)a of Hybrid Duplexes
average, the Tm’s of the hybrid duplexes increased by ∼3 °C
for DNA and ∼2 °C for RNA for every chiral unit incorporated.
PNA
sequence
Tm
(
°C)
∆Tm (°C)
The hybridization stability, in this case, is not affected by
backbone spacing as compared to the modification made at the
R-position (compare the Tm of P3 to that of P4).24 The extent
of this stabilization is significant, considering that no charge-
charge interaction is involved and that the modified unit can
be incorporated at every position without compromising the
overall binding affinity, unlike the observations made with
R-PNAs.24 In addition to the high binding affinity, these
modified LS-γPNAs bind to DNA and RNA with high sequence
selectivity, Figure 11 (and Figures S7 and S8: Supporting
Information). A single-base mismatch lowered the Tm by 16-
19 °C for DNA and 12-18 °C for RNA, depending on the
mismatched pair, as compared to 10-14 °C and 11-18 °C,
respectively, for PNA (P1).32 This improvement could be
attributed to structural preorganization.
P1
P2
P3
P4
P5
H-GCATGTTTGA-LLys-NH2
H-GCATGγTTTGA-LLys-NH2
H-GCAγTGγTTγTGA-LLys-NH2
H-GCATGγTγTγTGA-LLys-NH2
(44) 54
(48) 57
(53) 60
(53) 59
(+4) +3
(+9) +6
(+9) +5
H-γGγCγAγTγGγTγTγTγGγA-LLys-NH2 (63) 64 (+19) +10
a Tm’s of PNA-DNA (in parentheses) and PNA-RNA hybrid duplexes
in sodium phosphate buffer at 2 µM strand concentration each.
Conclusion
In summary, we have shown that a simple γ-backbone modi-
fication can transform a randomly folded PNA into a right-
handed helical structure. These helical PNAs bind to DNA and
RNA with exceptionally high affinity and sequence selectivity.
Helical induction, in this case, occurs in the C- to N-terminal
direction. The determining factor for helical induction appears
to be sterically driven, and stabilized by base-stacking. This
finding suggests that it may be possible to assemble other chro-
mophores into organized arrays. The results reported herein have
important ramification not only on the future design of nucleic
acid mimics with high binding affinity and sequence-selectivity
but also for the design of novel materials with more precise
spatial organization and efficient electronic coupling42-46sfor
possible applications in molecular electronics and photonic
devices.
Figure 11. Tm’s of P1-DNA, P5-DNA, P1-RNA and P5-RNA hybrid
duplexes containing perfectly matched and single-base mismatched se-
quences, where X ) A, C, G, T(U).
by base-stackingswhich provides the added stability to the
structure. This is clearly observed in the computer-generated
structures.
(42) Arai, T.; Inudo, M.; Ishimatsu, T.; Akamatsu, C.; Tokusaki, Y.; Sasaki,
T.; Nishino, N. J. Org. Chem. 2003, 68, 5540-5549.
The Effects of Structural Preorganization on the Binding
Affinity and Sequence-Specificity of PNA. To determine
whether structural preorganization contributes to the hybridiza-
tion stability of PNA, we measured the melting transitions (Tm’s)
(43) Solladie, N.; Hamel, A.; Gross, M. Tetrahedron Lett. 2000, 41, 6075-
6078.
(44) Dunetz, J. R.; Sandstrom, C.; Young, E. R.; Baker, P.; Van Name, S. A.;
Cathopolous, T.; Fairman, R.; de Paula, J. C.; Akerfeldt, K. S. Org. Lett.
2005, 7, 2559-2561.
L
of the various S-γPNAs (P2-P5) with complementary DNA
(45) Hippius, C.; Schlosser, F.; Vysotsky, M. O.; Bohmer, V.; Wurthner, F. J.
Am. Chem. Soc. 2006, 128, 3870-3871.
and RNA strands and compared them to that of PNA (P1),
Table 3 (and Figures S5 and S6, Supporting Information). On
(46) van der Boom, T.; Hayes, R. T.; Zhao, Y.; Bushard, P. J.; Weiss, E. A.;
Wasielewski, M. R. J. Am. Chem. Soc. 2002, 124, 9582-9590.
9
10266 J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006

Transformation of PNA into a Helical Structure
A R T I C L E S
L
Acknowledgment. We are grateful to Carnegie Mellon Uni-
versity and the Department of Chemistry for financial support.
The NMR spectrometers in the Department of Chemistry NMR
Facility at Carnegie Mellon University were purchased in part
with funds from the National Science Foundation (CHE-
0130903).
quantum-filtered-phase-sensitive COSY of (CLT) Ser-γPNA
in D2O at 500 MHz, UV-melting curves of PNA-DNA and
PNA-RNA hybrid duplexes containing perfectly matched and
single-base mismatched sequences, and CD spectra of PNA oli-
gomers containing different nucleobase sequence at the adjacent
N-terminal position with respect to the chiral backbone units.
This material is available free of charge via the Internet at
http://pubs.acs.org.
L
Supporting Information Available: Synthesis of S-γPNA
monomers and oligomers, HPLC and MALDI-TOF traces of
P5, CD spectra of other L-serine derived PNA oligomers, double-
JA0625576
9
J. AM. CHEM. SOC. VOL. 128, NO. 31, 2006 10267