The α-Helical Peptide Nucleic Acid Concept: Merger of Peptide Secondary Structure and Codified Nucleic Acid Recognition

Article abstract of DOI:10.1021/ja038434s

A novel platform for nucleic acid recognition that integrates the α-helix secondary structure of peptides with the codified base-pairing capability of nucleic acids is reported. The resulting α-helical peptide nucleic acids (αPNAs) are composed of a repeating tetrapeptidyl unit, aa₁-aa₂-aa₃-Ser^B, where aa ₁ through aa₃ represent generic ancillary amino acids and B = nucleobases linked to Ser via a methylene bridge. Effective syntheses of constituent Fmoc-protected nucleoamino acids (Fmoc-Ser^B-OH, where B = thymine, cytosine, and uracil) are described along with a protocol for the solid-phase synthesis of 21mer αPNAs containing five such nucleobases. By varying the ancillary amino acids, two distinct classes of αPNAs were constructed, having a net charge of -1 or +6, respectively, at physiological pH. The modular nature of the αPNA platform was illustrated by the synthesis of symmetrical disulfide-bridged αPNA dimers containing 10 nucleobases. Hybridization of these αPNAs with ssDNA has been examined by thermal denaturation, gel electrophoresis, and circular dichroism (CD) and the data indicated that αPNA binds to ssDNA in a cooperative manner with high affinity and sequence specificity. In general, b2 αPNAs bind faster and more strongly with ssDNA than do the corresponding b1 αPNAs. Parallel αPNA-DNA complexes are more stable than their antiparallel counterparts. CD studies also revealed that the hybridization event involves the folding of both species into their helical conformations. Finally, NMR experiments provided conclusive evidence of Watson-Crick base pairing in αPNA-ssDNA hybrids.

Full text of DOI:10.1021/ja038434s

Published on Web 03/18/2004
The r-Helical Peptide Nucleic Acid Concept: Merger of
Peptide Secondary Structure and Codified Nucleic Acid
Recognition
Yumei Huang,† Subhakar Dey,† Xiao Zhang,† Frank So¨nnichsen,‡ and
Philip Garner*,†
Contribution from the Department of Chemistry and Department of Physiology and Biophysics,
Case Western ReserVe UniVersity, CleVeland, Ohio 44106-7078
Received September 10, 2003; E-mail: ppg@case.edu
Abstract: A novel platform for nucleic acid recognition that integrates the R-helix secondary structure of
peptides with the codified base-pairing capability of nucleic acids is reported. The resulting R-helical peptide
nucleic acids (RPNAs) are composed of a repeating tetrapeptidyl unit, aa1-aa2-aa3-SerB, where aa1 through
aa3 represent generic ancillary amino acids and B ) nucleobases linked to Ser via a methylene bridge.
Effective syntheses of constituent Fmoc-protected nucleoamino acids (Fmoc-SerB-OH, where B ) thymine,
cytosine, and uracil) are described along with a protocol for the solid-phase synthesis of 21mer RPNAs
containing five such nucleobases. By varying the ancillary amino acids, two distinct classes of RPNAs
were constructed, having a net charge of -1 or +6, respectively, at physiological pH. The modular nature
of the RPNA platform was illustrated by the synthesis of symmetrical disulfide-bridged RPNA dimers
containing 10 nucleobases. Hybridization of these RPNAs with ssDNA has been examined by thermal
denaturation, gel electrophoresis, and circular dichroism (CD) and the data indicated that RPNA binds to
ssDNA in a cooperative manner with high affinity and sequence specificity. In general, b2 RPNAs bind
faster and more strongly with ssDNA than do the corresponding b1 RPNAs. Parallel RPNA-DNA complexes
are more stable than their antiparallel counterparts. CD studies also revealed that the hybridization event
involves the folding of both species into their helical conformations. Finally, NMR experiments provided
conclusive evidence of Watson-Crick base pairing in RPNA-ssDNA hybrids.
Introduction
based mechanism of action if degradation of partially comple-
mentary nontarget mRNA sequences occurs. Furthermore, PS
The seminal discovery1 that translation can be inhibited by
an antisense oligonucleotide (oligo) suggested a general ap-
proach to therapeutic intervention.2 The related phenomenon
of RNA interference (RNAi)3 involves the generation of an
antisense RNA complexed to a protein with RNase activity.4
Currently, phosphorothioate oligodeoxynucleotides (PS ODNs)
and 2′-O-(2-methoxyethyl)-modified PS ODNs represent the
state-of-the-art in antisense technology.5 Although drugs based
on PS ODNs have reached the marketplace, there are drawbacks
associated with the current technology. For one thing, PS oligos
are not trivial to prepare in stereochemically homogeneous form
and are, therefore, formulated as mixtures of diastereomers. The
specificity of PS oligos may be compromised by their RNase-
oligos are known to interact with proteins,6 and this may lead
to unintended side effects. One solution to this collection of
problems involves replacing the repeating ribose phosphate with
a simpler and more versatile backbone. Morpholino oligos
(which are, in fact, derived from ribonucleotides) represent one
such backbone replacement.7 Despite their ability to inhibit
translation in cell-free experiments, morpholino oligos are not
taken up into cells efficiently.8
Yet another approach involves the use of a peptide or peptide-
like backbone. Even though peptide and peptide-like nucleic
acid surrogates were first proposed some 30 years ago,9,10
a
successful antisense drug based on this class of hybrid molecules
has yet to emerge. Of the many structural motifs examined,
Nielsen’s peptide nucleic acid (PNA) has come the closest to
† Department of Chemistry.
‡ Department of Physiology and Biophysics.
(1) Zamecnik, P. C.; Stephensen, M. L. Proc. Natl. Acad. Sci. U. S A. 1978,
75, 280.
(2) (a) Antisense Technology Part A: General Methods, Methods of DeliVery
and RNA Studies; Phillips, M. I., Ed.; Academic Press: San Diego, CA,
2000; Methods in Enzymology, Vol. 313. (b) Antisense Technology Part
B: Applications; Phillips, M. I., Ed.; Academic Press: San Diego, CA,
1999; Methods in Enzymology, Vol. 314;
(3) Fire, A.; Xu, S.; Montgomery, M. K.; Kostas, S. A.; Driver, S. E.; Mello,
C. C. Nature 1998, 391, 806.
(4) Hutva´gner, G.; Zamore, P. D. Science 2002, 297, 2056.
(5) Crooke, S. T., Ed. Antisense Drug Technology - Principles, Strategies
and Applications; M. Decker: New York, 2001.
(6) King, D. J.; Ventura, D. A.; Brasier, A. R.; Gorenstein, D. G. Biochemistry
1998, 37, 16489.
(7) Summerton, J.; Weller, D. Antisense Nucleic Acid Drug DeV. 1997, 2375,
187.
(8) Summerton, J.; Stein, D.; Huang, S.; Matthews, P.; Weller, D.; Partridge,
M. Antisense Nucleic Acid Drug DeV. 1997, 7, 63.
(9) (a) de Koning, H.; Pandit, U. K. Rec. TraV. Chim. 1971, 91, 1069. (b)
Buttrey, J. D.; Jones, A. S.; Walker, R. T. Tetrahedron 1975, 31, 73. (c)
Shvachkin, Yu. P. Zh. Obshch. Khim. 1979, 49, 1157.
(10) For a comprehensive compilation of peptide-based nucleic acid surrogates
up to 1999, see: Falkiewicz, B. Acta Biochim. Pol. 1999, 43, 509. (See
also ref 15).
9
4626
J. AM. CHEM. SOC. 2004, 126, 4626-4640
10.1021/ja038434s CCC: $27.50 © 2004 American Chemical Society

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
reaching this goal. This oligonucleotide surrogate stands out in
terms of its structural simplicity and ability to bind sequence
specifically to ssRNA and DNA.11 In these systems, the ribose
phosphate backbone is replaced by a repeating N-(2-amino-
ethyl)glycine unit,12 which simplifies their synthesis and imparts
biostability to the resulting oligomers. However, the first
generation Nielsen PNAs have drawbacks associated with them
(low solubility, insufficient orientation discrimination, poor
cellular uptake) that have hindered their therapeutic application.
The incorporation of chiral amino acids (D-Lys for example)13
in place of glycine represents a possible solution to such
problems, but the extent that one can actually functionalize the
PNA backbone appears to be limited. One can add property-
modifying functionality via conjugation (PNA-peptide or
PNA-DNA chimerae);14 however, this is done at the expense
of increasing molecular complexity.
The ability to alter a molecule’s chemical, biochemical, and/
or pharmacological properties while maintaining its receptor
binding affinity and specificity lies at the heart of successful
drug development. Thus, it can be argued that the ideal antisense
drug platform would consist of the simplest possible molecular
framework that allows one to introduce functionality indepen-
dent of the primary mRNA binding “domain”. This line of
reasoning led us to revisit the idea of replacing the ribose
phosphate backbone with a peptide that has nucleobases attached
to specific amino acid residues. It was surmised that the problem
with earlier peptide-based nucleic acid surrogates (including one
of our own design)15 was due to improper spacing of the nucleo-
bases. The peptide backbone in these systems would have to
adopt an energetically unfavorable conformation to form Wat-
son-Crick base-pairs with the target nucleic acid. We now pres-
ent a full account of our successful development of a new syn-
thetic construct that merges the R-helix secondary structure of
peptides with the codified base-pairing capability of nucleic
acids.16,17
Figure 1. The RPNA design concept.
pairing with a single-stranded nucleic acid target. These nucleo-
bases are attached to the peptide backbone via a flexible meth-
ylene linker to serine in order to preserve the N-glycoside
(O-C-N) substructure found in nucleic acids.19 We initially
considered the tetrad or (i, i + 4) peptide spacing motif. In tetrad
RPNAs made up of L-amino acids, the nucleobase-containing
residues which are three amino acids apart trace a right-handed
superhelix that can become aligned by tightening the R-helix
to a 310 helix. Assuming that the RPNA has a helical pitch of
5-6 Å, synchronized tilting of the nucleobases relative to the
R-helix axis would be necessary to approximate the nucleobase
spacing of a single-stranded nucleic acid in a standard B-helical
geometry. In our RPNA, the flexible -CH2-O-CH2- linker
acts as a hinge for this purpose. This, along with the known
deformability of both DNA and peptide R-helices,20 was ex-
pected to lead to an induced fit driven by Watson-Crick base-
pairing.
With the general RPNA design concept formulated, we next
turned to the question of what ancillary amino acids should be
incorporated into the peptide backbone. The goal was to
minimally edit a known peptide sequence that favors R-helix
formation. Our first RPNA (which we term backbone 1 or b1)
was based on an amphipathic helix sequence emanating from
Baltzer’s group.21 Specifically, this RPNA backbone included
hydrophobic amino acids (Ala and Aib), internal salt bridges
(Glu-(aa)3-Lys-(aa)3-Glu), a macrodipole (Asp-(aa)15-Lys), and
an N-acetyl cap to favor R-helix formation.22 The C-termini of
these RPNA modules end in a carboxamide function to preclude
any potential intramolecular end effects (Figure 2). The
incorporation of a Cys residue at position 21 allowed for
implementation of Goddard’s “disulfide stitchery” strategy23 to
connect two b1 RPNA modules at the N-terminus. Replacement
of Gly1 with Cys and Cys21 with Aib provided b1′ RPNAs for
connection at the C-terminus. We decided on a disulfide linkage
first to demonstrate the modular nature of our platform, even
though the resulting RPNA-dimers would probably not be
suitable drug candidates due to the ease of reductive disulfide
cleavage in the cytoplasm.
Results and Discussion
1. Design Concept. The basic R-helical peptide nucleic acid
concept is illustrated in Figure 1.18 Our prototype RPNA module
incorporated five nucleobases for Watson-Crick (W‚C) base
(11) (a) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchhardt, O. Science 1991,
254, 1497. (b) Egholm, M.; Buchhardt, O.; Nielsen, P. E.; Berg, R. H. J.
Am. Chem. Soc. 1992, 114, 1895. (c) For a comprehensive review, see:
Uhlmann, E.; Peyman, A.; Breipohl, G.; Will, D. W. Angew. Chem., Int
Ed. 1998, 37, 2796n.
(12) The idea of using N-(2-aminoethyl)glycine as a ribose phosphate replace-
ment may actually be traced back to an article by Westheimer. See:
Westheimer, F. H. Science 1987, 235, 1173.
(13) (a) Sforza, S.; Haaima, G.; Marchelli, R.; Nielsen, P. E. Eur. J. Org. Chem.
1999, 197. (b) Sforza, S.; Corradini, R.; Ghirardi, S.; Dossena, A.; Marchelli,
R. Ibid. 2000, 2905.
(14) Uhlmann, E.; Peyman, A.; Breipohl, G.; Will, D. W. Angew. Chem., Int.
Ed. 1998, 37, 2796.
(15) Garner. P.; Yoo, J. U. Tetrahedron Lett. 1993, 34, 1275.
(16) Preliminary accounts of this work: (a) Garner, P.; Dey, S.; Huang, Y.;
Zhang, X. Org. Lett. 1999, 1, 403. (b) Garner, P.; Dey, S.; Huang, Y. J.
Am. Chem. Soc. 2000, 122, 2405.
While the binding properties of backbone 1 RPNAs with
complementary ssDNAs were being studied, it was discovered
(17) After our original RPNA communication appeared (ref 16a), Mihara and
co-workers reported the incorporation of nucleobase-containing amino acids
into both coiled-coils and R-helical peptides designed to recognize HIV-1
RRE RNA: (a) Matsumura, S.; Ueno, A.; Mihara, H. Chem. Commun.
2000, 1615. (b) Takahashi, T.; Hamasaki, K.; Ueno, A.; Mihara, H. Bioorg.
Med. Chem. 2001, 9, 991.
(19) The synthesis of racemic Bz-SerU-OMe was reported in the Russian
chemical literature in another context: Timoshchuk, V. A.; Olimpieva, T.
I. Zhurnal Obshei Khimii 1988, 58, 2404.
(20) Curran, T.; Kerppola, T. K. Science 1991, 254, 1210.
(21) (a) Olofsson, S.; Johansson, G.; Baltzer, L. J. Chem. Soc., Perkin Trans.
2. 1995, 2047. (b) Broo, K. S.; Brive, L.; Ahlberg, P.; Baltzer, L. J. Am.
Chem. Soc. 1997, 119, 11362.
(22) Regan, L.; Degrado, W. F. Science 1988, 241, 976.
(23) Park, C., Campbell, J. L.; Goddard, W. A., III. J. Am. Chem. Soc. 1995,
117, 6287.
(18) Amino acid abbreviations: aa ) generic amino acid, Ala ) L-alanine, Aib
) 2-aminoisobutyric acid, Cys ) L-cysteine, Asp ) L-aspartic acid, Glu
) L-glutamic acid, Gly ) glycine, Lys ) L-lysine, Ser ) L-serine, SerT
)
(S)-2-amino-3-(5-methyl-2,4-dioxo-3,4-dihydro-2H-pyrimidin-1-ylmeth-
oxy)propionic acid, SerC ) (S)-2-amino-3-(4-amino-2-oxo-2H-pyrimidin-
1-ylmethoxy)propionic acid, SerU ) (S)-2-amino-3-(2,4-dioxo-3,4-dihydro-
2H-pyrimidin-1-ylmethoxy)propionic acid.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4627

A R T I C L E S
Huang et al.
acid group of 1 was protected as a benzyl ester using benzyl
bromide in the presence of tetrabutylammonium iodide and
KHCO3 in DMSO. Alcohol 2 was smoothly converted to the
MTM ether 3 in high yield via a Pummerer-like reaction using
benzoyl peroxide and dimethyl sulfide in acetonitrile.26 It is
worth mentioning here that the analogous MTM ether in the
Boc series was obtained as an oil and was difficult to purify by
flash chromatography. On the other hand, the Fmoc derivative
3 is a crystalline solid and is easy to purify using flash column
chromatography. This common intermediate 3, available in one
step from commercially available 2, was used to synthesize all
of the nucleoamino acids.
Figure 2. Helical-wheel representations of backbone 1 and 2 RPNAs. b1
sequence: Ac-CysAcm-Gly-SerB-Asp-Ala-Glu-SerB-Ala-Ala-Lys-SerB-Ala-
Ala-Glu-SerB-Ala-Aib-Ala-SerB-Lys-Gly-NH2. b2 sequence: Ac-CysAcm
Lys-SerB-(Ala-Ala-Lys-SerB)4-Gly-Lys-NH2.
-
Following Sugimura’s lead,27 N-bromosuccinimide (NBS)
was initially chosen as the promoter for the nucleosidation
reaction between bis-silylated thymine 428 and MTM ether 3.
A mixture of the desired product 5 along with its N3 regioisomer
(not shown) was obtained in about a 1.2:1 ratio. Both Ogilvie29
and Tsantrizos30 reported that only one regioisomer (N1) formed
when I2 was used as an activator with similar MTM ether
substrates. Using this activator, a single product was obtained
in 60% yield (95% based on recovered 3). The reaction profile
depends on various factors. One of them is the relative
concentration of I2 over 3, the optimal I2/3 ratio being 1.5. A
mixture of N1 and N3 regioisomers formed if over 2.0 equiv of
I2 was used or if the reaction time was extended beyond 48 h.
Also, when I2 was added to the reaction mixture before bis-
silylated thymine, 3 is converted to two (unidentified) side
products and 5 is not formed. Hydrogenolytic deprotection of
the benzyl ester using H2 and Pd-C (10%) in THF-MeOH
was clean, providing the thymine-containing nucleoamino acid
6 in quantitative yield.
that the rate of association was extremely slow even though
the resulting RPNA-ssDNA complexes were relatively stable.
We hypothesized that at neutral pH, b1 RPNAs were negatively
charged, thereby introducing a kinetic barrier to the process of
association with negatively charged ssDNA. It was surmised
that if one could make RPNAs that were positively charged at
neutral pH, this kinetic barrier could be reduced by virtue of
attractive Coulombic forces. In this context, it had been reported
that the attachment of a positively charged Lys-rich peptide to
DNA enhanced the rate of duplex formation 48 000-fold.24
A
new backbone was designed based on the simple logic of
replacing the negatively charged Asp and Glu residues in b1
with Ala and positively charged Lys residues, respectively.
Additional Lys residues were incorporated at Ala4 and Gly20,
and the Aib residue was replaced by Ala. Finally, the Gly1 and
Lys2 residues were interchanged in order to avoid proximal Lys
residues. The resulting backbone 2 or b2 RPNAs (Figure 2)
would now have a net charge of +6 at neutral pH. N-Terminal
dimerization of b2 RPNAs was possible via the Cys21 residue.
For dimerization at the C-terminus, the Lys1 and Cys21 residues
were interchanged to give the b2′ RPNA backbone motif.
2. rPNA Synthesis. 2a. Nucleoamino Acid Synthesis.
Thymine, cytosine, and uracil containing NR-9-fluorenylmethoxy-
carbonyl (Fmoc)-protected nucleoamino acids 6, 10, and 13 were
synthesized (Scheme 1) by a modification of the route we used
to prepare the analogous tert-butoxycarbonyl (Boc)-protected
species (see ref 15). The Fmoc group was to be incorporated at
the very beginning of the synthesis. The exocyclic amine group
of the cytosine was to be protected with the acid-labile Boc
group, which can be cleaved at the end of the RPNA synthesis
under mildly acidic (or even neutral)25 conditions, thus minimiz-
ing the exposure of the RPNAs to acid. The carboxylic acid
moiety was to be masked as its benzyl ester, which can be
cleaved under neutral hydrogenolysis conditions at the end of
the nucleoamino acid synthesis. Both the acid (Boc)- and base
(Fmoc)-sensitive functionality were to be retained in a way that
would not jeopardize the configurational stability of the nucleo-
amino acid building blocks.
For the incorporation of cytosine into our RPNA, it was
decided to protect the N4-amino group. The Boc group was
chosen for this purpose because it’s reactivity is orthogonal
to the Fmoc protecting group. Cytosine was treated with
Boc2O and a catalytic amount of DMAP, to give pure N4-Boc
cytosine 7 in 54% yield after a simple workup. The mono-
silylated 2-trimethylsilyl-N4-Boc-cytosine 8 undergoes reaction
with 3 in the presence of I2 to produce the desired compound
9 as single isomer in 46% isolated yield (96% based on
recovered 3).31 Clean removal of the benzyl group was observed
when 10 equiv of 1,4-cyclohexadiene was used along with 10%
Pd-C in THF-MeOH (1:1) as solvent.32 The cytosine-
containing nucleoamino acid 10 was thus obtained as a pure
solid in quantitative yield.
Following a similar procedure, the uracil nucleoamino acid
was synthesized. Only one major compound 12 was obtained
in 24% isolated yield (74% based on recovered 3) from the
iodine-mediated reaction of 3 and U‚2TMS33 11. 1,4-Cyclo-
hexadiene-mediated benzyl group cleavage gave 13 in quantita-
tive yield.
The first task was to synthesize the methylthiomethyl (MTM)
ether 3 on a large scale, using the minimum number of steps
from a commercially available starting material. Fmoc-Ser-OH
1 was chosen as the starting material and is a relatively
inexpensive commercially available chemical. The carboxylic
(26) Median, J. C.; Salomon, M.; Kyler, K. S Tetrahedron Lett. 1988, 29, 3773.
(27) Sugimura, H.; Osumi, K.; Kodaka, Y.; Sujino, K. J. Org. Chem. 1994, 59,
7653.
(28) Garner, P.; Park, J. M. J. Org. Chem. 1990, 55, 3772.
(29) Ogilvie, K. K.; Nguyen-Ba, N.; Hamilton, R. G. Can. J. Chem. 1984, 62,
1622.
(30) Tsantrizos, Y. S.; Lunetta, J. F.; Boyd, M.; Fader, L. D.; Wilson, M. C. J.
Org. Chem. 1997, 62, 5451.
(24) Corey, D. R. J. Am. Chem. Soc. 1995, 117, 9373.
(25) (a) Hwu, J. R.; Jain, M. L.; Tsay, S.-C.; Hakimelahi, H. Tetrahedron Lett.
1996, 37, 2035. (b) Siro, J. G.; Mart´ın, J.; Garc´ıa, J. L.; Remuin˜an, M. J.;
Vaquero, J. J. Synlett 1998, 147. (c) Ham, J.; Choi, K.; Ko, J.; Lee, H.;
Jung, M. Protein Pept. Lett. 1998, 5, 257.
(31) Huang, Y.; Dey, S.; Garner, P. Tetrahedron Lett. 2003, 44, 1441.
(32) Bajwa, J. S. Tetrahedron Lett. 1992, 33, 2299.
(33) Cheng, V.; Hughes, L.; Griffin, V. B.; Montserret, R.; Ollapally, A. P.
Nucleosides Nucleotides 1986, 5, 223.
9
4628 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
Scheme 1. Synthesis of Nucleoamino Acids. The Atom Designation Scheme Shown for Compound 5 Is General for All Nucleoamino Acids
The N1 and N3 isomers of the thymine derivatives can often
be distinguished by analyzing the peak-shape of the H-6′ signal
1
in H NMR spectrum. The N1 isomer is expected to give a
singlet while a doublet or broad singlet is expected for the N3
isomer due to adjacent NH proton coupling. However, this
simple analysis was not possible with the Fmoc-thymine
nucleoamino acid 6. Correlation of 13C NMR data (C-5′ and
C-6′) with that of similar pyrimidine derivatives can be used to
assign regiochemistry.34 However, such assignments are not
unambiguoussespecially when only one regioisomer is avail-
able. Since there was precedent for determining nucleobase
regiochemistry using two-dimensional long-range heteronuclear
multiple bond shift correlation (HMBC) NMR experiments, we
decided to apply this method to our system.35 The protecting
groups of the pyrimidine nucleoamino acids were removed by
treatment with 2% DBU in DCM followed by 95% TFA in
water. Assignments of the proton-attached carbons were made
via the heteronuclear multiple-quantum coherence (HMQC)
Figure 3. HMBC correlation of H-SerT-OH 14, H-SerC-OH 15, and
H-SerU-OH 16 regiochemistry.
experiments of these compounds. Subsequent HMBC experi-
ments showed that the thymine-, cytosine-, and uracil-containing
nucleoamino acids possess the HMBC patterns expected of the
N1 regioisomer (Figure 3).
2b. Solid-Phase rPNA Synthesis. Our general synthetic
route to RPNAs and their symmetrical disulfide dimers is
depicted in Scheme 2. The synthesis consisted of three distinct
stages: (1) solid-phase RPNA synthesis of the resin-bound
module 18, (2) cleavage of the peptides from the resin with
concomitant global deprotection of all amino acid residues
except Cys to give RPNA module 19, and (3) thiol deprotection
and disulfide bond formation to produce the symmetrical dimer
20. Because the nucleobase-containing serine residue is poten-
tially both acid- and base-labile, the strategic and tactical issues
associated with the solid-phase synthesis of these RPNAs
(34) Kalinowski, H. O.; Berger, S.; Braun, S. Carbon-13 NMR Spectroscopy;
Translated by Becconsall, J. K. Imprint Chichester; Wiley: New York,
1988; pp 441.
(35) (a) Bax, A.; Summers, M. F. J. Am. Chem. Soc. 1986, 108, 2093. (b) Huang,
J. J.; Ragouzeos, A.; Rideout, J. L. J. Heterocycl. Chem. 1994, 31, 1685.
(c) Tima´r, Z.; Kova´cs, L.; Kova´cs, G.; Schme´l, Z J. Chem. Soc., Perkin
Trans. 2000, 1, 19.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4629

A R T I C L E S
Huang et al.
Scheme 2. Solid-Phase Synthesis of RPNA.
was circumvented by adjusting the pH so as to convert all of
the free amines to their corresponding ammonium salts. Thus,
when the reaction was performed in 0.5 M HCl, a clean
conversion from monomer to dimer was observed. These
symmetrical B10 RPNA dimers were characterized by MALDI-
TOF or ESI mass spectrometry (See Table S2 in Supporting
Information).
3. Binding Studies. Our RPNA-DNA binding studies
addressed the following questions: (1) What is the stand alone
conformation of RPNAs? (2) What is RPNA’s capacity and
propensity to hybridize with ssDNA? (3) Is there specific
hydrogen bonding recognition between the nucleobases of DNA
and RPNA? (4) What are the orientational specificity, stoichi-
ometry, and conformation of such hybrids? Thermal denatur-
ation, gel electrophoresis, circular dichroism, and NMR spec-
troscopy were used to study RPNA and DNA interactions. Two
RPNA structural variants were studied, each differing in the
overall charge of the peptide backbone. The RPNAs constitute
both symmetrical (CCCCC, TTTTT, CCTCC) and asymmetric
(CTCCT) nucleobase sequences.
3a. Thermal Denaturation Studies on rPNA-ssDNA
Complexes. UV melting was used to evaluate the mutual affinity
of RPNA and ssDNA. This method is based on the hyperchro-
mic UV absorption shift of stacked versus unstacked base-pairs.
Cooperative binding is generally indicated by a sigmoidal
absorption versus temperature curve, with the transition midpoint
being defined as the melting temperature (Tm). Comparison of
the dissociation (heating) and association (cooling) curve can
also provide a qualitative indication of the binding kinetics. Tm
measurements on complexes with base-pair mismatches between
RPNA and DNA provide information about the sequence
specificity of binding.
DNA with dangling nucleobases (represented by italicized
letters) on the 5′ and 3′ ends was initially used for the RPNA
hybridization studies with d(G)5 and d(G)10 since DNAs made
up of contiguous tracts of guanosine nucleotides tend to form
aggregates in solution. The addition of the dangling bases at
both ends of DNA was also reported to increase the Tm of the
DNA duplex possibly through additional nonspecific interactions
(hydrophobic packing, neighboring base stacking) or the exclu-
sion of solvent intrusion into the terminal base-pairs.42 Subse-
quently, we incorporated dangling bases into all of our ssDNA
sequences.
The amino acid sequence of our first RPNA series (which
we termed as backbone 1 or b1) was designed based on an
amphipathic helix sequence (Ac-CysAcm-Gly-SerB-Asp-Ala-Glu-
SerB-Ala-Ala-Lys-SerB-Ala-Ala-Glu-SerB-Ala-Aib-Ala-SerB-
Lys-Gly-NH2; represented as BBBBB(b1), where B is generic
nucleobase). A characteristic sigmoidal curve was observed for
a CCCCC(b1) + d(TA3G5A3T) (Tm ) 34 °C) solution which
was annealed in water for 4 days at 4 °C (Table 1, entry 3).
This Tm value is indicative of the formation of a stable complex.
However, no reassociation of the complex was observed during
the cooling cycle. Both b1 (tail-to-tail Tm ) 39 °C) and b1′
(head-to-head Tm ) 51 °C) TTTTT dimers showed cooperative
binding with d(A10). Again no reassociation of the complex was
observed during the cooling cycle. A Tm difference of ap-
proximately 12 °C was observed for b1 (tail-to-tail) and b1′
resemble those associated with glycopeptides.36 Accordingly,
we chose the commercially available Fmoc-Rink amide MBHA
resin as the solid support. The base-labile Fmoc protecting group
(PG) for the N-terminal amines and acid-labile PGs (Boc and
tert-butyl ester) for all side chain functionality excepting Cys.
The I2-labile acetamidomethyl (Acm) PG was chosen for the
Cys residue to facilitate a separate disulfide bond formation
step.37 DBU38 was used for Fmoc deprotection, and Carpino’s
HATU39 reagent was used for all peptide couplings (except that
involving Fmoc-CysAcm-OH). Finally, the acid-labile Rink amide
linker40 permitted the release of a fully extended peptide amide
from the resin and removal of the acid-labile PGs at the same
time. Several homo and hetero pyrimidine base-containing
RPNAs were synthesized using optimized protocol in about 20%
overall average yield after preparative HPLC purification. An
abasic peptide corresponding to b2 was also synthesized by
replacing nucleoamino acids with Ser and was used as a control
sequence for binding studies. RPNAs were characterized by
either MALDI-TOF or ESI mass spectrometry (See Table S1
in Supporting Information).
2c. rPNA Dimer Formation. To assemble RPNA sequences
containing more than five nucleobases, we joined two modules
together using Goddard’s “disulfide stitchery” strategy. Our b1
and b2 RPNAs incorporated a Cys residue at the N-terminus
while the b1′ RPNA backbone had a Cys residue at the
C-terminus. The plan was to use postpurification I2-mediated
dimerization of S-Acm protected RPNAs to obtain the RPNA
dimers. In the event, exposure of the T5(b1) monomer to a
solution of I241 in MeOH-H2O provided the tail-to-tail T5(b1)-
dimer in 84% yield after HPLC purification. Similar treatment
of T5(b1) monomer resulted in a 49% purified yield of the head-
to-head T5(b1′)-dimer. However, when we applied these condi-
tions to cytosine-containing backbone 2 RPNAs, the result was
a complex reaction profile. It was hypothesized that the problem
might be the oxidation of Lys side chains (or other amino
groups) by I2 during the dimerization reaction. This problem
(36) Paulsen, H.; Schleyer, A.; Mathieux, N.; Meldal, M.; Bock, K. J. Chem.
Soc., Perkin Trans. 1 1997, 281.
(37) Veber, D. F.; Milkowski, J. D.; Varga, S. L.; Denkewalter, R. G.;
Hirschmann, R. J. Am. Chem. Soc. 1972, 94, 5456.
(38) Wade, J. D.; Bedford, J.; Sheppard, R. C. Tregear, G. W. Pept. Res. 1991,
4, 194. A control experiment with Fmoc-SerT-OMe indicated that exposure
to 20% piperidine led to the elimination of free thymine.
(39) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
(40) (Rink amide linker ) 4-(2′,4′-dimethoxyphenylaminomethyl)phenoxy-
acetamido) Rink, H. Tetrahedron Lett. 1987, 28, 3787.
(41) Kamber, B.; Hartmann, A.; Eisler, K.; Riniker, B.; Rink, H.; Sieber, P.;
Rittel, W. HelV. Chim. Acta 1980, 63, 899.
(42) Freier, S. M.; Burger, B. J.; Alkema, D.; Neilson, T.; Turner, D. H.
Biochemistry 1983, 22, 6198.
9
4630 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
Lys-SerB-Ala-Ala-Lys-SerB-Ala-Ala-Lys-SerB-Ala-Ala-Lys-SerB-
Gly-Lys-NH2; represented as BBBBB(b2), where B is a generic
nucleobase).
Table 1. UV Melting Data for RPNA-DNA Complexes
This change brings about a remarkable enhancement in the
kinetic and thermodynamic aspects of the complex formation.
Compared to the CCCCC(b1)-d(TA3G5A3T) complex (Tm ) 34
°C), a ∆Tm of +20 °C was observed for CCCCC(b2)-
d(TA3G5A3T) (Table 1, entry 2) in TE buffer (10 mM Tris-
HCl, 1 mM EDTA, pH 7) with no hysteresis observed between
the heating and cooling UV melting curve. Recalling that no
binding was observed between TTTTT(b1) and d(A10), a Tm of
17 °C was now observed for TTTTT(b2)-d(A10) (Table 2, entry
1). Binding affinity is proportional to the number of C‚G versus
T‚A base pairs in the complex (Table 1, compare entries 1, 2,
and 4). With one mismatch, a ∆Tm of -16 °C was observed
for CCTCC(b2)-d(TA3G5A3T) (Table 1, entry 5). Once again,
the incorporation of dangling bases at the 5′ and 3′ ends of the
DNA increases the Tm of RPNA and DNA complex. Compared
to CCCCC(b2)-d(AT3G5T3A), greater hysteresis, lower hyper-
chromicity and a lower melting temperature (Tm, 35 °C, ∆Tm
-14 °C) was observed for the complex formation between
CCCCC(b2) and d(G5) (Table 1, entry 6). Both b2 (tail-to-tail)
and b2′ (head-to-head) TTTTT dimers and d(A10) complexes
showed two-step melting (Table 2, entries 3 and 4) with nearly
identical Tm values. Detailed interpretation of the two-state
binding observed between homothymine RPNA-dimers and
ssDNA is complicated by the fact that both duplex and triplex
formation (with 1:1 or 2:1 RPNA/ssDNA binding stoichiom-
etries) is possible.
Table 2. UV Melting Data for RPNA-Dimer-d(A)10 Complexes
To see whether ionic interactions between positively charged
peptide backbone of b2 RPNA and negatively charged phosphate
backbone of ssDNA were contributing to the binding of RPNA
and ssDNA, an “abasic” peptide (Ac-CysAcm-Lys-(Ser-Ala-Ala-
Lys)4-Ser-Gly-Lys-NH2) with unmodified Ser residues was also
studied. No cooperative binding to d(TA3G5A3T) was observed
in the UV melting curve with this abasic peptide. This indicated
that the cooperative melting between RPNA and DNA was not
merely a reflection of nonspecific ionic interactions.
To probe the orientational preference of RPNA binding to
DNA, b2 RPNA having asymmetrical sequence (CTCCT) was
synthesized and its interaction with d(A3GAGGAA3) and
d(A3AGGAGA3) was studied. The sequence of the DNA was
designed to allow only one complex to form Watson-Crick
base pairs. UV melting experiments showed that the parallel
orientation is more favorable than antiparallel orientation. The
parallel (N-terminus of RPNA adjacent to the 5′-end of DNA
or N/5′) complex has a higher melting temperature (Tm ) 37
°C) and reforms during the cooling cycle more quickly than
the antiparallel (N/3′) complex (Tm ) 32 °C). On the UV melting
experiment time scale, no reassociation was observed in the case
of the antiparallel complex.
(head-to-head) dimers (Table 2, entries 1 and 2) indicating a
strong orientational preference for these complexes.
Even though the stability of the resulting RPNA-DNA
complexes for b1 was quite high (relative to that expected for
the corresponding DNA-DNA duplex) as evident from these
data, the rate of association between b1 (or b1′) RPNA and DNA
was slow. We hypothesized that, at neutral pH, the Asp and
Glu residues in b1 (or b1′) RPNA were negatively charged, and
may have unfavorable electrostatic interactions with the nega-
tively charged phosphate backbone of DNA, introducing a
kinetic barrier to the process of association with DNA.
Finally, the effect of ionic strength on b2 RPNA binding to
DNA was probed. Added salt can affect the properties of bio-
logical macromolecules, leading to modulation of their stability,
solubility, and biological activity.43 Increasing the ionic strength
has a stabilizing effect on DNA-DNA duplexes and a destabi-
lizing effect on Nielsen’s PNA-DNA duplexes.44 For b2 RPNA
and DNA binding, the RPNA and DNA are stabilized in part
by the ionic interactions (formation of ion pairs) and the
concomitant release of low molecular-weight ions (CF3COO-Na+)
A new backbone was designed based on the simple logic of
replacing the negatively charged Asp and Glu residues in b1
with Ala and positively charged Lys residues, respectively.
Additional Lys residues were incorporated at Ala4 and Gly20,
and the Aib residue was replaced by Ala. Finally, the Gly1 and
Lys2 residues were interchanged in order to avoid proximal Lys
residues. The resulting backbone 2 RPNAs have a net charge
of +6 at neutral pH (sequence: Ac-CysAcm-Lys-SerB-Ala-Ala-
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4631

A R T I C L E S
Huang et al.
Figure 4. Effect of ionic strength (0-150 mM NaCl in 10 mM Tris-HCl,
1 mM EDTA, pH 7.0) on the Tm of CCCCC(b2)-d(TA3G5A3T). Single strand
concentration ) 5 µM. Samples incubated overnight at 4 °C.
that were previously associated with the charged groups on the
biopolymer. Because ions are released in these biopolymer
charge neutralization reactions, the equilibrium shifts to favor
the complex formation when the salt concentration is reduced.45
Figure 4 shows the effect of the NaCl concentration on the Tm
of a b2 RPNA-DNA complex. As the salt concentration
increased, the Tm decreased. Experimentally, after 100 mM
NaCl, additional salt did not affect the stability of the complex
very much. A linear plot of Tm versus ln[NaCl] (not shown)
was observed which may allow the theoretical calculation of
Tm of RPNA-DNA at any salt concentration. The Tm of
CCCCC(b2)-d(TA3G5A3T) drops to 49 °C when the experiment
is performed under simulated physiological salt concentration
(150 mM NaCl). This still represents a remarkably strong
binding interaction when compared to the corresponding DNA-
DNA duplex, d(C5)-d(TA3G5A3T), which has a Tm of 19 °C
under the same conditions.
Figure 5. Nondenaturing PAGE analysis of CCTCC(b2) and d(A3GG-
AGGA3) complex formation. The ratios of CCTCC(b2) to d(A3GGAGGA3)
for lanes 1-8 are 0/1, 1/2, 3/4, 1/1, 5/4, 3/2, 7/4, and 2/1. Lane 9 represents
only CCTCC(b2). (B) The mole ratio of RPNA and DNA (X-axis) is plotted
vs the band intensity of the bound DNA (Y-axis). The titration curve shows
the extent of complex formation. A Kd of ∼ 0.2 µM was derived from the
data.
Collectively, these thermal denaturation studies demonstrated
that RPNAs bind to complementary DNA targets with high
affinity and in a sequence-specific manner, consistent with our
proposed base-pairing model.
Figure 6. Nondenaturing PAGE analysis of TTTTT(b2)-dimer and d(C3T-
(TC2)2A10C(TC2)3) complexes formation. The ratios of TTTTT(b2)-dimer
to d(C3T(TC2)2A10C(TC2)3) for lanes 1-8 are 0/1, 1/2, 1/1, 3/2, 2/1, 5/2,
3/1, and 4/1.
3b. Gel Electrophoresis Studies on rPNA-ssDNA Com-
plexes. Standard nondenaturing polyacrylamide gel electro-
phoresis (PAGE) was performed on RPNA-ssDNA com-
plexes.46 Only b2 RPNAs were studied due to their higher
affinity toward ssDNA and their faster complexation rate
compared with b1 RPNAs. b2 RPNA has six lysine residues,
and the free amino groups of these lysines will be protonated
under the experimental conditions; consequently, the RPNA will
migrate toward the cathode during electrophoresis. The lengths
of the DNA sequences were accordingly chosen to ensure that
the complexes remained negatively charged and migrated toward
the anode. The detection of any new species that migrates more
slowly than the uncomplexed DNA would be taken as direct
evidence for RPNA-ssDNA complex formation.
CCTCC(b2)-d(A3GGAGGA3) was detected (Figure 5A). The
intensity of the unbound DNA d(A3GGAGGA3) band steadily
decreased with increased RPNA and disappeared completely at
a RPNA/DNA ratio of 1.5. Triplex formation is precluded on
the grounds that the resultant ternary complex would have a
net positive charge and therefore be expected to migrate toward
the negative electrode. Quantitative analysis of this gel (see ref
46) allowed us to determine a dissociation constant Kd of ∼
0.2 µM (at 278 K) for this complex.
The analogous gel shift experiment with TTTTT(b2)-dimer
(net charge +12) and d(C3T(TC2)2A10C(TC2)3) (net charge -29)
appeared to produce an additional slower-moving species at the
expense of the initially formed complex (Figure 6). This result
is consistent with the formation of binary 1:1 and ternary 2:1
complexes (recall the two-step melting of this complex in Table
2). It was not possible to distinguish between a tandem duplex
or triplex structure for the 2:1 complex.
For the complex between CCTCC(b2) and d(A3GGAGGA3),
a single, new, and slower-migrating species corresponding to
(43) Von Hippel, P. H.; Schleich, T. Effects of Neutral Salts on the Structure
and Conformational Stability of Macromolecules in Solution. In Structure
and Stability of Biological Macromolecules; Timasheff, S. N., Fasman, G.
D., Eds.; Biological Macromolecules: Vol. 2; M. Dekker Inc.: New York,
1969; p 417.
(46) (a) Taylor, J. D.; Ackroyd, A. J.; Halford, S. E. The Gel Shift Assay for
the Analysis of DNA-Protein Interactions. In DNA-Protein Interactions:
Principles and Protocols; Kneale, G. G., Ed.; Methods in Molecular
Biology, Vol. 30; Humana Press Inc.: Totowa, NJ, 1994; p 263. (b) Lane
D.; Prentki, P.; Chandler, M. Microbiol. ReV. 1992, 56, 509.
(44) Tomac, S.; Sarkar, M.; Ratilainen, T.; Wittung, P.; Nielsen, P. E.; Norde´n,
B.; Gra¨slund, A. J. Am. Chem. Soc. 1996, 118, 5544.
(45) (a) Record, M. T.; Anderson, C. F.; Lohman, T. M. Q. ReV. Biophys. II
1978, 2, 103. (b) Manning, G. S. Ibid., 179.
9
4632 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
Figure 7. Nondenaturing PAGE analysis of CTCCT(b2) and d(A3GA-
GGAA3) (lanes 1-4), CTCCT(b2) and d(A3AGGAGA3) (lanes 5-8)
complexes formation. The ratios of CTCCT(b2) to DNA for lanes 1-4 as
well as 5-8 are 0/1, 1/1, 2/3, and 4/7.
Figure 9. CD spectra of ssDNA (a), RPNA (b), and RPNA-ssDNA
complex (c). Final single-stranded oligomer concentration was 6 µM in
HPLC grade H2O (pH 5.6). Spectra were recorded at 5 °C.
deprotonated with the incremental changes in pH (the pKa of
+
the lysine ꢀ-NH3 in Ala-Lys oligopeptides is approximately
11.5 in 10 mM NaCl at 0 °C),50 the charge-repulsion between
the neighboring lysine residues decreases, as a result of which
the R-helix conformation is favored. Similar behavior has been
reported for poly-lysine or poly-(lysine-alanine) peptides.51
Conformation changes induced in peptide-nucleic acid
complexes can also be detected by CD spectroscopy.52 Although
the CD spectra of peptides and nucleic acids overlap extensively,
they do differ in some regions. For example, above 240 nm,
DNA has strong CD bands, whereas in the 210-230 nm region,
the peptide CD signal is strong and nucleic acid CD signal is
weak. Aromatic and sulfur-containing side chains show a major
contribution around 280 nm. We used CD spectroscopy to study
RPNA-DNA interactions. Trace b in Figure 9 shows the CD
spectrum of RPNA CCTCC(b2) in H2O (pH 5.6) in the absence
of DNA, which suggests an almost random conformation
because of the protonated Lys amines. Upon addition of an
equimolar amount of ssDNA d(A3GGAGGA3), the characteristic
CD signatures of an R-helical peptide were observed (Figure
9, trace c). Because this is a composite spectrum, the R-helix
CD is superimposed upon those of other species present in the
sample. Also, we cannot rule out the possibility of coexisting
310-helix and R-helix structures, since it is not possible to
distinguish between them by CD methods.53 The maximum at
280 nm and minimum at 255 nm are suggestive of an ordered
right- handed DNA helix. Since the control CD spectra of RPNA
and ssDNA indicate different secondary structures than the
RPNA-ssDNA complex, it appears that they are mutually acting
as templates for hybridization. Analogous behavior has been
noted previously in studies on DNA-binding proteins54 and
synthetic peptides that correspond to the DNA-binding region
of the E. coli RecA protein.55 Upon binding, DNA induces the
Rec24 to adopt a highly helical conformation in which the
arrangement of positive charges in Rec24 is analogous to the
Figure 8. CD Spectra of b2 RPNA (CCTCC) obtained at different pH.
Concentration: 6 µM in 17 mM phosphate buffer at 25 °C.
Similar PAGE experiments on CTCCT(b2) + d(A3GAGGAA3)
and CTCCT(b2) + d(A3AGGAGA3) also confirmed that the
parallel (N/5′) orientation resulted in stronger binding (Figure
7). At a particular RPNA/DNA ratio, more parallel complex
was formed compared to its antiparallel counterpart. Again, no
binding was observed between the “abasic” peptide Ac-CysAcm
-
Lys-(Ser-Ala2-Lys)4-Ser-Gly-Lys-NH2) and d(TA3G5A3T), un-
derscoring the role that nucleobases play in RPNA molecular
recognition.
3c. CD Studies on rPNA-ssDNA Complexes. Circular
dichroism (CD) spectroscopy has been used extensively to study
peptide conformation.47 Since RPNA has a chiral peptide
backbone, the RPNA backbone itself is expected to have optical
properties similar to normal peptides. We first studied the tail-
to-tail TTTTT(b1)-dimer. The CD spectra were acquired at
various temperatures (Supporting Information). The double
minima at 220 and 206 nm as well as the maximum at 193 nm
are characteristic of an R-helix.48 With increasing of temperature,
the intensity of the minimum at 200 nm decreased indicating a
transition from R-helix to random structure. An isodichroic point
nearly at 202 nm was suggestive of a temperature-dependent
R-helix to random coil transition.49 The helical content at 20
°C in water was estimated to be 26%. Thus, we concluded that
b1 RPNA dimers are indeed R-helical in solution.
A similar pattern was noticed for b2 RPNA CCTCC(b2). At
neutral pH, CCTCC(b2) is partially helical. An interesting
phenomenon was revealed during the pH titration of this RPNA
(Figure 8). The helicity of the backbone 2 RPNA increases with
the pH. It is highly likely that as more lysines become
(50) Marqusee, S.; Baldwin, R. L. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 8898.
(51) (a) Yasui, S. C.; Keiderling, T. A. J. Am. Chem. Soc. 1986, 108, 5576. (b)
Satake, I.; Yang, J. Biopolymers 1975, 14, 1841. (c) Yaron, A.; Tal, N.;
Berger, A. Biopolymers, 1972,11, 2461.
(52) Talanian, R. V.; McKnight, C. J.; Kim, P. S. Science 1990, 249, 769.
(53) Sudha, T. S.; Vijayakumar, E. K. S.; Balaram, P. Int. J. Pept. Protein Res.
1983, 22, 464.
(47) Fasman, G. D., Ed. Circular Dichroism and the Conformational Analysis
of Biomolecules; Plenum Press: New York, 1996.
(48) Holzwarth, G.; Doty, P. J. Am. Chem. Soc. 1965, 87, 218.
(49) Padmanabhan, S.; Baldwin, R. L. J. Mol. Biol. 1991, 219, 135.
(54) (a) Patel, L.; Abate, C.; Curran, T. Nature 1990, 347, 572. (b) Weiss, M.
A.; Ellenberger, T.; Wobbe, C. R.; Lee, J. P.; Harrison, S. C.; Struhl, K.
Nature 1990, 347, 575.
(55) Zlotnick, A.; Brenner, S. L.; J. Mol. Biol. 1988, 209, 447.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4633

A R T I C L E S
Huang et al.
11 shows the one-dimensional spectrum of RPNA alone (Figure
11A) and RPNA and ssDNA mixture (Figure 11B). Most
resonance lines have different chemical shifts in the RPNA and
in the RPNA-ssDNA mixture indicating complex formation.
The peaks of RPNA and ssDNA in the mixture spectrum are
also broader compared to the spectrum of RPNA alone. This
line broadening may be due to the fact that the complex has
increased molecular weight and is comparatively rigid with
limited segmental mobility for the individual strands.57
Five exchangeable imino proton resonances are observed in
the RPNA-ssDNA complex (Figure 12, the resonance of 12.83
ppm is very weak, but was resolved in the 1D-NMR at 25 °C),
while no imino proton resonances are observed in the RPNA
and DNA alone. These signals are relatively sharp and intense,
implying that they are stabilized via hydrogen bonds through
base pairing. With an increase in temperature, the complex
melted and the imino proton peaks in the RPNA-ssDNA
complex slowly disappeared. The chemical shift of the imino
protons in DNA duplexes not only depends on the intrinsic
contribution from the central residue, but also depends on the
additional contributions induced by 5′- and 3′-nearest-neighbor
(flanking) residues. The imino proton of thymine usually appears
further downfield than the guanine imino proton.58 Referring
to the triplet case59 (13.82 ppm for CTC, 12.80 ppm for AGG,
and 12.66 ppm for GGA), 13.44 ppm is assigned as the thymine
N3-H peak. The signals of the NH (imino) protons of G and T
move upfield as the temperature increases; G-NH (imino) moves
more slowly than does T-NH (imino) due to an increasing
exchange with the water solvent. This is consistent with our
assignment. The two peaks with relatively high intensity (12.72,
12.52 ppm) are assigned as the G5/G7 imino protons. The
remaining peaks (13.15, 12.83 ppm) are assigned to the end
bases G4 and G8 (refer to Figure 13 for the numbering system
of ssDNA and RPNA).
The NOESY spectrum of the complex in phosphate buffer
at 5 °C also exhibits several cross-peaks involving the imino
protons (Figure 13). For the A‚T base pairs, there was a strong
NOE correlation from the thymine imino to the H-2 of the paired
adenine (7.64 ppm) and a weaker NOE correlation to the adenine
N6-H (7.81 ppm).60 For the G‚C base pairs, relatively strong
cross-peaks between the guanine imino protons and the amino
protons on the paired cytosine were seen except for the end
bases G4/G8. For the guanine imino protons (12.52 ppm), cross-
peaks between the guanine imino protons and the amino protons
on the paired cytosine were seen (8.10/6.30 ppm, cross-peaks
between these two amino protons are observed). For the imino
proton at 12.72 ppm, cross-peaks between both N4-H of cytosine
and N2-H of guanine were observed (7.93/6.37, 8.04/5.97; cross-
peaks between 7.93/6.37 and 8.04/5.97 identify the two amino
groups). The strong cross-peaks for the amino protons indicates
either a slow or restricted rotation of the amino groups, which
is expected if they are engaged in the base pairing. A Hoogsteen
base pairing scenario can be excluded on the basis that no cross-
Figure 10. Job plots for CD intensities of the mixtures of DNA and RPNA.
(A) CCCCC(b2)-d(TA3G5A3T) (9) and CCTCC(b2)-d(A3GGAGGA3) (b)
at 258 nm. (B) TTTTT(b2)-dimer-d(A10) (2) at 261 nm. Total concentration
of the RPNA and DNA was 12 µM in HPLC grade H2O (pH 5.6).
3.6 residue periodicity of hydrophobic amino acid residues in
amphipathic R-helices. Taken together, these CD studies
performed on complexes formed between cationic b2 RPNA
and DNA not only provided support for the binding event but
also for the proposed binding model. The induced conforma-
tional change of cationic b2 RPNA was coupled with specific
binding between RPNA and ssDNA.
To probe for any differences between the conformations of
the antiparallel and parallel complexes, CD spectra of the parallel
CTCCT(b2)-d(A3GAGGAA3) and antiparallel CTCCT(b2)-d(A3-
AGGAGA3) complexes were also obtained. These spectra are
distinctly different, especially in the 240-280 nm regions, where
the base pairs make a larger contribution in the parallel complex
(Supporting Information), confirming the preference of a parallel
complex for symmetrical RPNAs.
Finally, CD spectroscopy was also used to obtain accurate
values for the stoichiometry of the RPNA and DNA in a
complex.56 CD spectra were recorded for mixtures of CCCCC-
(b2) + d(TA3G5A3T), CCTCC(b2) + d(A3GGAGGA3) and
TTTTT(b2)-dimer + d(A10) at different molar ratios. The CD
data for monomeric modules when plotted with respect to the
mole percentage of RPNA produced curves having a break point
at a 1:1 molar ratio of the RPNA and DNA strands (Figure 10A).
However, the TTTTT(b2)-dimer and d(A10) showed evidence
of both 1:1 and 2:1 binding stoichiometries (Figure 10B). This
is in line with the UV melting data in which a two-step melting
was observed (Table 2, entries 3 and 4). No minimum or
maximum was observed (data not shown) for mixtures of totally
mismatched RPNA CCCCC(b2) and DNA d(A10), providing
additional evidence that RPNA interacts with DNA in a
sequence specific manner.
3d. NMR Studies on rPNA-ssDNA Complexes. To obtain
direct evidence of specific hydrogen bonding between the
nucleobases in an RPNA-ssDNA hybrid, one- and two-
dimensional NMR spectra of the imino and amino protons of
the nucleobases involved in base pairing were examined. Figure
(57) Wu¨thrich, K. NMR of Proteins and Nucleic Acids; Wiley & Sons: New
York, 1986.
(56) (a) Kim, S. K.; Nielsen, P. E.; Egholm, M.; Buchardt, O.; Berg, R. H.;
Norden, B. J. Am. Chem. Soc., 1993, 115, 6478. (b) Wittung, P.; Kim, S.
K.; Buchardt, O.; Nielsen, P.; Norden, B. Nucleic Acid Res. 1994, 22, 5371.
(c) Wittung, P.; Nielsen, P.; Norden, B. J. Am. Chem. Soc., 1996, 118,
7049. (d) Gray, D. M.; Hung, S.; Johnson, K. H. Methods Enzymol. 1995,
246, 19.
(58) Leijon, M.; Gra¨slund, A.; Nielsen, P. E.; Buchardt, O.; Norde´n, B.;
Kristensen, S. M.; Eriksson, M. Biochemistry 1994, 33, 9820.
(59) Altona, C.; Faber, D. H.; Hoekzema, A. W. Magnetic Reson. Chem. 2000,
38, 95.
(60) Brown, S. C.; Thomson, S. A.; Veal, J. M.; Davis, D. G. Science 1994,
265, 777.
9
4634 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
Figure 11. 1H NMR spectra of RPNA CCTCC(b2) (A) and the CCTCC(b2)-d(A3GGAGGA3) (1 to 1 ratio) complex (B) recorded at 600 MHz and 5 °C.
Figure 12. 1H NMR spectra of the imino proton region of the d(A3GG-
AGGA3)-CCTCC(b2) complex (1:1 ratio) at different temperatures.
peaks are observed between the thymine imino protons and the
H-8 proton on the partner adenine.
In all, the data presented here clearly demonstrates that the
RPNA-ssDNA complex is associated through Watson-Crick
base pairs. Overlap of the ssDNA and RPNA resonances makes
a detailed conformational characterization of the RPNA-ssDNA
complex very challenging. For complete elucidation of the three-
dimensional structure of this RPNA-ssDNA complex, informa-
tion rich data using different NMR experiments should be
acquired and isotopically labeled RPNA or ssDNA may be used.
RPNAs having unsymmetrical amino acid sequences may allow
better dispersion of the data.
Conclusions
Figure 13. (a) NOESY spectrum of the imino proton region of the CCTCC-
A novel platform for nucleic acid recognition that merges
(b2)-d(A3GGAGGA3) complex (1:1 ratio) at 5 °C. (b) Diagrams showing
the imino protons after hydrogen bonding and the numbering system for
ssDNA and RPNA.
the R-helix secondary structure of peptides with the codified
base-pairing capability of nucleic acids has been designed.
Effective syntheses of the constituent Fmoc-protected thymine,
cytosine, and uracil nucleoamino acids 6, 10, and 13 were de-
veloped along with protocols for the solid-phase synthesis of
21mer RPNAs containing five nucleobases as well as their con-
version into symmetrical disulfide-bridged RPNA dimers (con-
taining 10 nucleobases). The binding of RPNA to ssDNA was
examined by thermal denaturation, gel electrophoresis, circular
dichroism, and NMR spectroscopy.
The following conclusions can be drawn from these data.
RPNAs bind to ssDNAs in a cooperative manner with high
affinity and sequence specificity. In general, b2 RPNAs bind
faster and more strongly with ssDNA than do the corresponding
b1 RPNAs. Parallel RPNA-DNA complexes are more stable
than their antiparallel counterparts. CD studies also revealed
that the hybridization event involves the folding of both species
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4635

A R T I C L E S
Huang et al.
mL). The organic layer was dried (Na2SO4) and concentrated by rotary
evaporation, and the residual EtOAc was removed under vacuum to
give a yellow oil. The oil was cooled to -78 °C (dry ice/acetone) and
triturated with hexanes. The yellow solid thus obtained was transferred
to a Bu¨chner funnel (filter paper) and washed with hexanes under
suction until a white solid was obtained (0.563 g, 92% yield). mp 96-
97 °C; Rf ) 0.20, 3:7 EtOAc-hexanes; [R]26D ) +0.63° (c 4.45, DCM);
1H NMR (300 MHz, CDCl3) δ 7.77 (d, J ) 7.6 Hz, 2H), 7.60 (d, J )
7.5 Hz, 2H), 7.26-7.43 (m, 9H), 5.75 (dd, J1 ) 7.0 Hz, J2 ) 1.1 Hz,
NH), 5.23 (s, 2H, CH2Ph), 4.20-4.50 (m, 3H, H-2 + 2 × H-10′), 4.22
(t, J ) 6.7 Hz, 1H, H-9′), 4.07 (dd, J1 ) 16.5 Hz, J2 ) 1.8 Hz, 2H,
CH2OH), 2.15 (broad s, 1H, OH). 13C NMR (75.4 MHz, CDCl3) δ
170.4, 156.2, 143.7, 143.6, 141.2, 135.0, 128.6, 128.4, 128.1, 127.0,
125.0, 119.9, 67.4 (CH2Ph), 67.2 (C-10′), 63.1 (C-3), 56.1 (C-2), 47.0
(C-9′). HRMS (EI) m/z, calcd for C25H23NO5 [M+] 417.1576, obsd
417.1574.
into their helical conformations. Finally, NMR experiments
provided conclusive evidence of Watson-Crick base pairing
in RPNA-ssDNA hybrids.
These studies constitute the proof of principle for RPNA
hypothesis and set the stage for further exploration of this novel
platform for nucleic acid recognition.
Experimental Section
1. Nucleoamino Acid Synthesis. Instrumentation. All reported
melting points were determined using a Mel-Temp capillary melting
point apparatus and are not corrected. Optical rotations were measured
using a Perkin-Elmer 241 polarimeter at either λ ) 589 nm (sodium D
line) or λ ) 578 nm (mercury J line) at RT and were reported as
follows: [R]TD or J, concentration (c ) g/100 mL), and solvent. NMR
spectra were recorded on Varian Inova 500 or Gemini-300 NMR
spectrometers and are reported in parts per million (ppm) on the δ
scale relative to residual CHCl3 (δ 7.25 or δ 77.0), C6H6 (δ 7.15), H2O
(δ 4.75), DMSO-d6 (δ 2.49, δ 39.5), tetramethylsilane (δ 0.00), or DSS
(2,2-dimethyl-2-silapentane-5-sulfonic acid, δ 0.00) for 1H or 13C.
Proton and carbon assignments are based on COSY and HETCOR
experiments. High-resolution mass spectral (HRMS) data were obtained
from a KRATOS Analytical MS25RFA spectrometer and are reported
in units of m/z for M+ or the highest mass fragment derived from M+.
Solvents and Reagents. All moisture-sensitive reactions were
performed in an inert, dry atmosphere of Ar. Reagent grade solvents
were used for chromatography and extraction. The following solvents
and reagents were purified beyond reagent grade: Tetrahydrofuran
(THF) was distilled under Ar or N2 from a purple solution of sodium-
benzophenone ketyl. Dichloromethane (DCM), 1,2-dichloroethane, and
CH3CN were distilled over CaH2 (0-1 mm grain size) under an Ar
atmosphere. Dimethyl sulfoxide (DMSO) was distilled from CaH2 under
reduced pressure and stored over 4 Å MS. 1,1,1,3,3,3-Hexamethyl-
disilazane (HMDS), chlorotrimethylsilane (TMSCl), thymine, cytosine,
uracil, benzyl bromide, benzoyl peroxide, tetrabutylammonium iodide,
4-(dimethylamino)pyridine (DMAP), di-tert-butyl dicarbonate ((Boc)2O),
dimethyl sulfide, N,O-bis-trimethylsilylacetamide (BSA), formic acid
(96% in water), trifluoroacetic acid (TFA), 1,4-cyclohexadiene, and
Pd-C (10% w/w) were used as received from Aldrich.
Chromatography. Thin-layer chromatography (TLC) was performed
using either Merck silica gel 60 F-254 plates or JT-Baker Si 250F plates
(0.25 mm thickness). The plates were visualized first with UV
illumination followed by charring with 0.3% (w/v) ninhydrin solution
in (97:3) EtOH-AcOH. Flash chromatography was performed using
silica gel (230-400 mesh).
HMBC/HMQC NMR Experiments. These experiments were
performed at 500 MHz (Varian, Inova 500) at ambient temperature
using 70-100 mg of sample in 750 µL of D2O as solvent. HMQC
parameters: sw (spectral width) ) 6000 Hz, np (number of points) )
600 and nt (number of transients) ) 8. HMBC parameters: sw ) 6000
Hz, τmb (reciprocal of multiple bond H-C coupling constant) ) 0.05
s and nt ) 16.
Fmoc-Ser(MTM)-OBn (3). A three-neck 250 mL round-bottom
flask equipped with a magnetic stir bar and solid addition assembly
containing benzoyl peroxide (19.6 g, 0.081mol), was charged with 2
(9.66 g, 0.0232 mol). After flushing the system with Ar, 130 mL of
dry CH3CN was added with stirring and the mixture was cooled to 0
°C (ice-water). To the cold solution was added Me2S (11.9 mL, 0.162
mol)), followed by solid benzoyl peroxide over a period of 30 min
from the solid addition assembly. The reaction was complete after
stirring for additional 1 h at 0 °C (a second development, after complete
drying, of a 5.5 cm TLC plate with 3:7 EtOAc-hexanes clearly
demonstrated the absence of starting material). The reaction was then
quenched with 100 mL of water and extracted with ether (3 × 250
mL). The combined organic layer was washed with saturated aq
NaHCO3 (3 × 75 mL) followed by 1 N aq HCl (3 × 75 mL) and brine
(1 × 75 mL), dried (Na2SO4), and concentrated by rotary evaporation
to give a white solid. The solid was purified by flash chromatography
on silica gel (6.5 × 17.5 cm bed, column packed with 1:9 EtOAc-
hexanes, sample loaded onto the column with CHCl3; gradient elution,
1000 mL of 1:9 EtOAc-hexanes then 1000 mL of 1:4 EtOAc-hexanes
followed by 500 mL of 3:7 EtOAc-hexanes) to afford 7.21 g of pure
product and some fractions containing slightly impure product which
were combined and rechromatographed (4 × 20 cm bed, column packed
with 1:9 EtOAc-hexanes; sample loaded with CHCl3; eluted with 600
mL of 3:7 EtOAc-hexanes) to afford another 2.58 g (89% yield) of
pure 3. mp 79-80 °C; Rf ) 0.71, 3:2 hexanes-EtOAc; [R]25D ) -8.08°
1
(c 1.36, DCM); H NMR (300 MHz, CDCl3) δ 7.75 (d, J ) 7.5 Hz,
2H), 7.60 (d, J ) 7.5 Hz, 2H), 7.24-7.42 (m, 9H), 5.70 (d, J ) 8.7
Hz, NH), 5.34 (dd, J1 ) 19.5 Hz, J2 ) 12.3 Hz, 2H, CH2Ph), 4.33-
4.64 (m, 5H, H-2 + 2 × H-10′ + 2 × H-3), 4.22 (t, J ) 7.0 Hz, 1H,
H-9′), 4.04 (dd, J1 ) 9.6 Hz, J2 ) 3.0 Hz, H-3a), 3.76 (dd, J1 ) 9.3
Hz, J2 ) 3.0 Hz, H-3b), 2.00 (s, 3H, SCH3); 13C NMR (75.4 MHz,
CDCl3) δ 170.0, 155.9, 143.9, 143.7, 141.2, 141.2, 135.2, 128.6, 128.2,
127.7, 127.0, 125.1, 125.1, 119.9, 75.6 (CH2SMe), 67.8 (C-3), 67.4
(CH2Ph), 67.2 (C-10′), 54.2 (C-2), 47.1 (C-9), 13.7 (SCH3). HRMS
(EI) m/z, calcd for C27H27NO5S [M+] 477.1609, obsd 477.1619.
Fmoc-SerT-OBn (5) A flask containing 1.48 g of crushed 3 Å MS
was flame dried under vacuum, flushed with argon, and then cooled to
room temperature. A solution of 3 (1.48 g, 3.11 mmol in 6.0 mL dry
THF) was transferred to the flask, followed by the addition of a solution
of 4 (4.2 g, 15.5 mmol, in 6.4 mL of dry THF) and a solution of I2
(1.18 g, 4.67 mmol, in 6.0 mL of dry THF). The reaction mixture was
stirred at room temperature under an Ar atmosphere for 48 h whereupon
TLC analysis showed only the presence of product and some
unconverted 3. The reaction mixture was poured into a 5% (w/v)
aqueous solution of Na2SO3 (100 mL) and stirred vigorously whereupon
a white precipitate formed. The mixture was suction filtered, and the
solid residue in the Bu¨chner funnel was washed with THF until the
washing was free of product (TLC analysis). Excess THF was
evaporated from the filtrate by rotary evaporation, and the aqueous
emulsion so formed was extracted with CHCl3 (3 × 100 mL). The
Fmoc-Ser-OBn (2). Dry DMSO (3 mL) was added to a 25 mL Ar-
flushed flask containing 1 (0.481 g, 1.47 mmol), KHCO3 (0.221 g,
2.20 mmol), and tetrabutylammonium iodide (0.0543 g, 0.147 mmol).
The resulting white suspension was stirred magnetically at room
temperature for 10 min to obtain a homogeneous solution. To this
solution was added benzyl bromide (0.524 mL, 4.40 mmol), and the
colorless reaction mixture was stirred for 8 h at room temperature. TLC
analysis of the resulting brownish-yellow solution showed the formation
of a single product and presence of trace amount of starting material.
The reaction was quenched by addition of 25 mL of water, and the
DMSO-water layer was extracted with EtOAc (3 × 75 mL). The
combined organic layers were washed, first with saturated aq NaHCO3
(3 × 50 mL) to remove 1 completely (analyzed by TLC), then with
saturated aq Na2S2O3 (2 × 50 mL), and finally with brine (1 × 50
9
4636 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
combined CHCl3 layers were washed with brine (2 × 100 mL), dried
over Na2SO4, and concentrated by rotary evaporation to give a white
solid. The solid was purified by flash chromatography using silica gel
(4 × 17.5 cm bed, column packed with 1:9 EtOAc-hexanes) The
sample was loaded onto the column with CHCl3, and 20 mL of
additional CHCl3 was used to embed the sample onto the silica. This
step is necessary to avoid precipitation of solid on the top of the column
upon addition of eluting solvent. The column was eluted initially with
1:1 EtOAc-hexanes (200 mL) to recover 0.539 g of the starting
material and then with 7:3 EtOAc-hexanes to give 1.036 g of pure 5
(60% yield; 95% based on recovered 3): mp 69 °C; Rf ) 0.46, 1:4
hexanes-EtOAc; [R]26D ) +2.20° (c 1.45, DCM); 1H NMR (300 MHz,
CDCl3) δ 8.66 (s, 1H, NH), 7.77 (d, J ) 7.2 Hz, 2H), 7.60 (d, J ) 7.5
Hz, 2H), 7.27-7.43 (m, 9H), 6.97 (s, 1H, H-6′), 5.72 (d, J ) 8.7 Hz,
NH), 5.20 (AB quartet, JAB ) 12.3 Hz, 2H, 2 × H-1′′′), 5.03 (s, 2H,
CH2Ph), 4.60 (m, 1H, H-2), 4.33-4.48 (m, 2H, 2 × H-10′′), 4.22 (t, J
) 6.6 Hz, 1H, H-9′′), 4.04 (dd, J1 ) 10.2 Hz, J2 ) 3.3 Hz, H-3a), 3.88
(dd, J1 ) 9.9 Hz, J2 ) 3.3 Hz, H-3b), 1.88 (s, 3H, 5′-Me); 13C NMR
(75.4 MHz, CDCl3) δ 169.7, 163.7, 155.9, 151.0, 143.7, 141.3, 138.7,
128.7, 128.7, 128.3, 127.80, 127.1, 125.2, 125.1, 76.8 (CH2Ph),
69.6 (C-3), 67.6 (C-1′′′), 67.3 (C-10′′), 54.3 (C-2), 47.1 (C-9′′), 12.3
(5′-CH3). HRMS (FAB, CsI/NaI/glycerol matrix) m/z, calcd for
C31H29N3O7 [MCs+] 688.1060, obsd 688.1046.
The yellow solid was washed first with 1 L of water and then with
500 mL each of EtOAc and hexanes successively to afford 14.97 g
(54% yield) of pure product as a white solid, which was dried in a
P2O5 desiccator. mp 270 °C (decomposed); Rf ) 0.81, 4:1:1 t-BuOH-
1
AcOH-H2O; H NMR (300 MHz, DMSO-d6) δ 7.73 (d, J ) 7.1 Hz,
1H, H-6), 6.87 (d, J ) 7.1 Hz, 1H, H-5), 1.44 (s, 9H, (CH3)3CO). 13
C
NMR (75.4 MHz, DMSO-d6) δ 173.7 (C-6), 162.8 (C-2), 154.5, 146.3,
93.5 (C-5), 80.7 ((CH3)3CO), 27.8 ((CH3)3CO); HRMS (EI) m/z, calcd
for C9H13N3O3 M+ 211.0957, obsd 211.0950.
N4-Boc-Cytosine‚TMS (8). After flushing a dry 250 mL, one-
necked, round-bottom flask, equipped with a magnetic stirring bar and
a rubber septum containing N4-Boc cytosine (6.06 g, 28.7 mmol) with
Ar, CH3CN (121 mL) and BSA (7.10 mL, 28.7 mmol) were added.
The resulting suspension was stirred at room temperature for 2 h, and
more BSA (3.55 mL, 14.4 mmol) was added. The reaction mixture
became a clear light-yellow solution after 15 min. Stirring was continued
at room temperature for another 2 h under Ar atmosphere, and the
contents was transferred to a flame-dried short path distillation assembly
(flushed with Ar). The CH3CN was first distilled under reduced pressure
at room temperature, and then the residual liquid was heated at 40-50
°C under reduced pressure (1-2 mmHg) for 12 h to afford a white
solid. The solid (8.13 g, 100%) in the flask was used directly in the
1
next reaction without further purification. H NMR (300 MHz, C6D6,
Fmoc-SerT-OH (6). Pd-C catalyst (10%, 500 mg) was added as a
slurry in THF to a solution of 5 (1.52 g, 2.73 mmol) in 100 mL of
(1:1) THF-MeOH and stirred while H2 was bubbled through the
reaction mixture at a moderate rate. A solution of 1:1 THF-MeOH
was periodically added to the reaction mixture to replenish the
evaporating solvent. After 45 min, TLC examination showed clean
conversion of starting material to product. The H2 flow was then
replaced by Ar for another 10 min, at which point the reaction mixture
was filtered and washed (THF) through a pad of Celite until the washing
was free of product (TLC analysis). Upon concentration of the filtrate
a sticky foam was obtained, from which the residual MeOH was
removed by coevaporation with CHCl3 (3 mL × 3) to afford 1.280 g
(quantitative yield) of a white solid. mp 80 °C (decomposed); Rf )
0.20, 90:8:2 CHCl3-CH3OH-AcOH; [R]23J ) +34.1° (c 1.11, CHCl3);
1H NMR (300 MHz, CDCl3) δ 10.22 (s, 1H, NH), 9.2 (broad singlet,
1H, COOH), 7.73 (d, J ) 7.1 Hz, 2H), 7.60 (t, J ) 7.2 Hz, 2H), 7.37
(t, J ) 7.8 Hz, 2H), 7.28 (t, J ) 7.2 Hz, 2H), 7.05 (s, 1H, H-6′), 6.11
(d, 7.8 Hz, NH), 5.03 (AB quartet, JAB ) 9.9 Hz, 2H, 2 × H-1′′′), 4.55
(m, 1H, H-2), 4.33-4.47 (m, 2H, 2 × H-10′′), 4.19 (t, 1H, J ) 7.2 Hz,
H-9′′), 4.04 (broad doublet, J ) 7.8 Hz, 1H, H-3a), 3.92 (broad doublet,
J ) 8.1 Hz, 1H, H-3b), 1.82 (s, 3H, 5′-Me); 13C NMR (75.4 MHz,
CDCl3) δ 165.0, 156.4, 151.8, 143.9, 143.8, 141.2, 140.0, 127.8, 127.1,
125.3, 125.2, 120.0, 111.5, 77.2, 69.5 (C-3), 67.2 (C-10′′), 54.6 (C-2),
47.1 (C-9′′), 12.1 (5′-Me); HRMS (FAB, CsI/NaI/glycerol matrix) m/z,
calcd for C24H23N3O7 [MCs+] 598.0590, obsd 598.0620, calcd [MNa+]
488.1433, obsd 488.1436.
N4-Boc-Cytosine (7). Cytosine (14.7 g, 0.132 mol) and DMAP
(0.808 g, 6.61 mmol) were added to a dry 500 mL one-necked round-
bottom flask, equipped with a magnetic stir bar and a rubber septum,
which was vented through an oil bubbler. The apparatus was flushed
with Ar, and 240 mL of dry DMSO was added. The resulting white
suspension was stirred at room temperature for 10 min, and then
(Boc)2O (38.0 mL, 0.165 mmol) was added, when the mixture became
homogeneous. The light yellow solution was stirred for an additional
24 h at room temperature, during which time some white solid
precipitated. Completion of the reaction was confirmed by analysis of
1H NMR (1 mL of the reaction mixture was transferred by a syringe to
a flask containing 10 mL water. The white precipitate formed was
filtered and washed with 10 mL each of water, MeOH, EtOAc, and
hexanes consecutively. After drying under vacuum, the 1H NMR
spectrum of this white solid was obtained). The reaction mixture was
transferred to a 2 L flask containing 1 L of water. The resulting
suspension was stirred vigorously for 30 min and then suction filtered.
distilled from Na) δ 8.54 (s, NH), 8.3 (d, J ) 11.3 Hz, 1H, H-6) 7.67
(d, J ) 11.40 Hz, 1H, H-5), 1.30 (s, 9H, (CH3)3CO), 0.34 (s, 9H, (CH3)3-
SiO).
Fmoc-SerC(Boc)-OBn (9). A flask containing 0.63 g of crushed 3 Å
MS was flame dried under vacuum, flushed with Ar, and then cooled
to room temperature. A solution of 3 (0.628 g, 1.32 mmol in 3 mL dry
THF, 0.44 M) was transferred to the flask, followed by the addition of
a solution of 8 (0.701 g, 1.97 mmol in 1.4 mL of dry THF, 1.0 M) and
a solution of iodine (0.334 g, 1.32 mmol in 4 mL of dry THF). The
reaction mixture was stirred at room temperature under Ar atmosphere
for 48 h whereupon TLC analysis showed only the presence of the
product and unconverted 3. The reaction mixture was worked up
similarly as in the case of compound 5 to give a white solid. The solid
was purified by flash chromatography on silica gel (5.5 × 12 cm bed,
column packed with 1:9 EtOAc-hexanes). The sample was loaded onto
the column with a minimum amount of CHCl3, and 5 mL of additional
CHCl3 was used to embed the sample onto the silica gel. The column
was eluted first with 1:1 EtOAc-hexanes to afford 0.112 g of starting
material 3, and then with 5:2:3 EtOAc-THF-hexanes to afford 0.402
g of pure 9 (46% yield, 96% based on recovered starting material). mp
177-178 °C; Rf ) 0.40, 5:2:3 EtOAc-THF-hexanes; [R]25J ) +1.17°
1
(c 1.96, DCM); H NMR (300 MHz, CDCl3) δ 7.76 (d, J ) 7.4 Hz,
2H), 7.60 (d, J ) 7.1 Hz, 2H), 7.26-7.43 (m, 10H), 7.15 (d, J ) 7.4
Hz, 1H, H-5′), 5.76 (d, J ) 8.5 Hz, NH), 5.60-5.62 (m, 4H, CH2Ph,
2 × H-1′′′), 4.58 (m, 1H, H-2), 4.32-4.48 (m, 2H, 2 × H-10′′), 4.22
(t, J ) 7.0 Hz, 1H, H-9′′), 4.07 (dd, J1 ) 2.9 Hz, J2 ) 9.5 Hz, H-3a),
3.90 (dd, J1 ) 2.6 Hz, J2 ) 9.8 Hz, H-3b), 1.56 (s, 9H, (CH3)3CO);
13C NMR (75.4 MHz, CDCl3) δ 169.7, 163.0, 155.9, 155.7, 150.9, 146.6
(C-6′), 143.8, 143.7, 141.3, 135.1, 128.6, 128.5, 128.3, 127.7, 127.1,
125.1, 125.0, 119.9, 95.8 (C-5′), 83.0 ((CH3)3CO), 78.3, 69.9 (C-3)
67.5, 67.2 (C-10′′) 54.3 (C-2) 47.1 (C-9′′) 28.0 ((CH3)3CO); HRMS
(FAB, CsI/NaI/glycerol matrix) m/z, calcd for C35H36N4O8 [MH]+
641.2611, obsd 641.2614.
Fmoc-SerC(B°c)-OH (10). Compound 9 was dissolved in minimum
volume of THF (approximately 25 mL/0.5 g) by stirring, and equal
volume of MeOH was added. To this solution was added a slurry of 2
equiv (by weight) of Pd-C (10%) in THF. After bubbling Ar through
the reaction mixture for 5 min, 10 equiv of 1,4-cyclohexadiene was
added, and the resulting mixture was stirred until completion of the
reaction (30-45 min). In case of incomplete reaction, an additional 5
equiv of 1,4-cyclohexadiene drives the reaction to completion within
10 min. Filtration and washing (THF) through a Celite pad followed
by rotary evaporation of the solvent gave the product in the form of
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4637

A R T I C L E S
Huang et al.
white foam in quantitative yield. mp 130 °C (formation of foam); Rf )
was added. The water layer was washed with CHCl3 (4 × 20 mL) and
evaporated to an oily residue. The oil was treated with TFA solution
(95% in H2O, 4 mL) and stirred for 45 min. After TFA evaporation
and diethyl ether addition, the product was precipitated as a white solid,
which was further washed with ether (4 × 5 mL). H-SerT-OH (14)
(obtained as a 1.5:1 mixture of 14 and DBU-TFA salt). 1H NMR (500
1
0.40, 4:1 CHCl3-MeOH; [R]25 ) +29.6° (c 0.45, DCM); H NMR
J
(300 MHz, DMSO-d6) δ 10.50 (br s, 1H), 8.04 (d, J ) 7.3 Hz, 1H,
H-6′), 7.87 (d, J ) 7.3 Hz, 2H), 7.73 (d, J ) 7.5 H, 2H), 7.68 (d, J )
8.4 Hz, 1H), 7.40 (dd, J1 ) 6.9 Hz, J2 ) 7.2 Hz, 2H), 7.32 (dd, J1 )
7.3 Hz, J2 ) 6.8 Hz, 2H), 7.08 (d, J ) 6.9 Hz, 1H, H-5′), 6.99 (d, J )
7.4 Hz, NH), 6.86 (br s, 1H), 6.64 (br s, 1H), 5.21 (s, 2H, 2 × H-1′′′),
4.18-4.30 (m, 4H, H-2, 2 × H-10′′, H-9′′), 3.80 (m, 1H, H-3a), 3.58
(m, 1 × H-3b), 1.43 (s, 9H, (CH3)3CO); 13C NMR (75.4 MHz, DMSO-
d6) δ 163.6, 155.7, 155.2, 152.2, 148.7 (C-6′), 143.9, 140.7, 139.3,
127.6, 127.0, 125.2, 124.9, 120.1, 94.6 (C-5′), 81.0 ((CH3)3CO), 78.1,
69.8, 67.0, 65.6, 55.5, 46.7 (C-9′′), 27.8 ((CH3)3CO). HRMS (FAB,
CsI/NaI/glycerol matrix) m/z, calcd for C28H30N4O8 [MCs]+ 683.11180,
obsd 683.11146. calcd for [(M-H)Cs2]+ 815.0094, obsd 815.0093.
MHz, D2O) δ 7.50 (q, J ) 2 Hz, 1H, H-6′), 5.19 (AB quartet, JAB
)
18.0 Hz, 2H, 2 × H-1′′′), 4.23 (t, J ) 5 Hz, 1H, H-2), 4.07 (dd, J1 )
17.5 Hz, J2 ) 6.0 Hz, H-3a), 4.00 (dd, J1 ) 18.5 Hz, J2 ) 5.5 Hz,
H-3b), 1.85 (d, J ) 2 Hz, 3H, CH3); 13C NMR (125.7 MHz, D2O) δ
169.5 (C-1), 166.6 (C-4′), 152.3 (C-2′), 141.3 (C-6′), 76.8 (C-1′′′), 65.9
(C-3), 52.8 (C-2), 10.80 (CH3). H-SerC-OH (15) (obtained as a 1:1
mixture of 15 and DBU-TFA salt). 1H NMR (500 MHz, D2O) δ 7.85
(d, J ) 8.0 Hz, 1H, H-6′), 6.15 (d, J ) 7.5 Hz, 1H, H-5′), 5.25 (m, 2H,
2 × H-1′′′), 4.28 (t, J ) 3.5 Hz, 1H, H-2), 4.10 (dd, J1 ) 10 Hz, J2 )
4.0 Hz, H-3a), 4.04 (dd, J1 ) 11 Hz, J2 ) 3.0 Hz, H-3b); 13C NMR
(125.7 MHz, D2O) δ 169.1 (C-1), 159.6 (C-4′), 149.0 (C-2′), 148.2
(C-6′), 94.9 (C-5′), 78.0 (C-1′), 66.1 (C-3), 52.5 (C-2). H-SerU-OH
Fmoc-SerU-OBn (12) A flask containing 3.0 g of crushed 3 Å MS
was flame dried under vacuum, flushed with Ar, and then cooled to
room temperature. A solution of Fmoc-SerMTM-OBn (3, 2.97 g, 6.23
mmol in 12 mL dry THF) was transferred to the flask, followed by
addition of a solution of U‚2TMS 11 (3.20 g, 12.5 mmol, in 6.8 mL of
dry THF) and a solution of I2 (2.37 g, 9.35 mmol, in 18.3 mL of dry
THF). The reaction mixture was stirred at room temperature under an
Ar atmosphere for 12 h whereupon TLC analysis showed mainly the
presence of one product and some unconverted 3. The reaction mixture
was worked up similarly as in the case of compound 5 to give a white
solid. The solid was purified by flash chromatography using silica gel
(4 × 20 cm bed, column packed with 1:9 EtOAC-hexanes). The sample
was loaded onto the column with CHCl3, and 20 mL of additional
CHCl3 was used to embed the sample onto the silica gel. The column
was eluted initially with 4:5:1 EtOAc-hexanes-THF (1 L) to recover
2.0 g of the starting material and then with 5:3:2 EtOAc-hexanes-
THF to give 0.8161 g of pure product (24% yield; 74% based on
recovered starting material 3). mp 54 °C (thawed) and 68 °C (formed
glass); Rf ) 0.36, 5:3:2 EtOAc-hexanes-THF; [R]25D ) +1.3° (c 2.8,
DCM); 1H NMR (300 MHz, CDCl3) δ 9.34 (s, 1H, NH), 7.75 (d, J )
7.3 Hz, 2H), 7.59 (d, J ) 6.9 Hz, 2H), 7.36-7.41 (t, J ) 7.3 Hz, 2H),
7.27-7.33 (m, 7H), 7.00 (d, J ) 7.9 Hz, 1H, H-6′), 5.80 (d, J ) 8.3
Hz, 1H), 5.60 (dd, J1 ) 8.4 Hz, J2 ) 1.5 Hz, NH), 5.11-5.25 (m, 2H,
2 × H-1′′′), 5.01 (s, 2H, CH2Ph), 4.58 (m, 1H, H-2), 4.33-4.46 (m,
2H, 2 × H-10′′), 4.20 (t, J ) 6.9 Hz, 1H, H-9′′), 4.00 (dd, J1 ) 9.9
1
(16) (obtained as a 3:1 mixture of 16 and DBU-TFA salt). H NMR
(500 MHz, D2O) δ 7.66 (d, J ) 8.5 Hz, 1H, H-6′), 5.83 (d, J ) 8.5
Hz, 1H, H-5′), 5.21 (AB quartet, JAB ) 10.5 Hz, 2H, 2 × H-1′′′), 4.28
(t, J ) 3 Hz, 1H, H-2), 4.11 (dd, J1 ) 10.5 Hz, J2 ) 4.2 Hz, H-3a),
4.04 (dd, J1 ) 10.5 Hz, J2 ) 3.3 Hz, H-3b); 13C NMR (125.7 MHz,
D2O) δ 169.7 (C-1), 166.6 (C-4′), 152.4 (C-2′), 145.8 (C-6′), 102.3
(C-5′), 77.1 (C-1′′′), 65.8 (C-3), 52.5 (C-2) (In all cases, signals from
DBU are not reported).
2. rPNA Synthesis, Purification and Analysis. Reagents and
Materials. Dimethyl sulfoxide (DMSO) was kept over oven-dried 4
Å MS for at least 12 h before use. 1-Methyl-2-pyrrolidinone (NMP)
and N,N-dimethylformamide (DMF) were distilled under reduced
pressure and kept over 4 Å MS for at least 12 h before use.
N,N-diisopropylethylamine (DIPEA) was distilled from CaH2 (0-1 mm
grain size) and kept over 4 Å MS for 12 h before use. Piperidine, DCM,
TFA, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), diisopropylcarbodi-
imide (DIC), N-hydroxybenzotriazole (HOBt) hydrate, and acetic
anhydride were used as received from Aldrich. All amino acids, their
derivatives, and Rink Amide MBHA resin (4-(2′,4′-dimethoxyphenyl-
Fmoc-aminomethyl)-phenoxy-acetamido-norleucyl-MBHA resin, 0.3-
0.8 mmol/g loading capacity, 100-200 mesh) were obtained from
Novabiochem. O-(7-azabenzotriazol-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate (HATU) was obtained from PerSeptive Biosystems
(Applied Biosystems).
Hz, J2 ) 3.0 Hz, H-3a), 3.88 (dd, J1 ) 9.9 Hz, J2 ) 3.0 Hz, H-3b); 13
C
NMR (75.4 MHz, CDCl3) δ 169.2, 162.8, 155.4, 150.5, 143.3, 142.3,
140.8, 134.6, 128.4, 128.2, 128.0, 127.3, 126.6, 125.0, 124.6, 119.5,
102.9, 76.3 (CH2Ph), 69.1 (C-3), 67.1 (C-1′′′), 66.7 (C-10′′), 53.7 (C-
2), 46.6 (C-9′′). HRMS (FAB, CsI/NaI/glycerol matrix) m/z, calcd for
C30H28N3O7 [MH]+ 542.1927, obsd 542.1928.
Automated Solid-Phase Synthesis Protocol. The synthesis of
RPNA was performed on an ACT90 peptide synthesizer (Advance
Chemtech) on a 54 µmol scale. During the couplings, 1.2 equiv of
amino or nucleoamino acids were used with respect to HATU to prevent
tetramethylguanidinium capping.61 After Fmoc cleavage, the NMP wash
is necessary to remove all DBU from the resin. The mixture of acetic
anhydride, DIPEA, and DMF was freshly prepared for the capping
reaction.
Fmoc-SerU-OH (13). The benzyl group was removed from 12 in a
similar way as in the case of compound 9 to give product 13 as a white
foam in quantitative yield. mp 110 °C (thawed) and 124 °C (formed
glass); Rf ) 0.27, 4:1 CHCl3-MeOH; [R]25D ) +16.0° (c 0.45, MeOH);
1H NMR (300 MHz, DMSO-d6) δ 11.34 (br s, 1H), 7.87 (d, J ) 7.6
Hz, 2H), 7.72 (d, J ) 7.3 Hz, 2H), 7.67 (d, J ) 7.9 H, 2H), 7.41 (t, J
) 7.0 Hz, 2H), 7.32 (dd, J1 ) 7.3 Hz, J2 ) 7.6 Hz, 2H), 6.63 (br s,
1H), 5.57-5.60 (dd, J1 ) 7.90 Hz, J2 ) 1.51 Hz, 1H, NH), 5.10 (s,
2H, 2 × H-1′′′), 4.17-4.28 (m, 4H, H-2, 2 × H-10′′, H-9′′), 3.76 (d,
J ) 5.5 Hz, 2H, 2 × H3); 13C NMR (75.4 MHz, DMSO-d6) δ 167.3,
159.6, 152.0, 147.1 (C-6′), 140.8, 140.7, 139.8, 136.7, 135.1, 124.0,
123.6, 123.0, 121.3, 121.2, 121.0, 120.8, 120.7, 116.2, 116.1, 116.0,
97.7 (C-5′), 72.4 (C-1′′′), 64.1, 61.8, 50.1, 42.6. HRMS (FAB,
CsI/NaI/glycerol matrix) m/z, calcd for C23H21N3O7 [MH]+ 452.1458,
obsd 452.1478.
Fmoc Deprotection. a. Resin treated with 2% DBU in DMF (v/v)
and mixed for 8 min.
b. Resin washed with DMF (1 × 25 mL DMF-flush; 1 × 10 mL
DMF, 1 min shake-flush) followed by NMP (3 × 10 mL, 1 min
shake-flush).
Amino Acid Coupling. a. Fmoc-protected amino/nucleoamino acids
(3.6 equiv relative to the loading capacity of resin) were preactivated
by mixing with HATU (3.0 equiv) and DIPEA (9.0 equiv) in 4/1 NMP-
DMSO for 1 min. Final amino/nucleoamino acid concentration was
0.2 M. Fmoc-CysAcm-OH (6.0 equiv) was preactivated by mixing HOBt
hydrate (4.0 equiv) and DIC (4 equiv) in 1:1 DMF-DCM for 5 min.
b. Preactivated Fmoc-protected amino/nucleoamino acid was added
to the resin and shaken with N2 bubbling for 45 min.
General Method for Preparation of Fully Deprotected Nucleo-
amino Acids (14-16). Approximately 100 mg of the protected
nucleoamino acid (6, 10, or 13) was dissolved in a solution of 2%
DBU in DCM (10 mL). The reaction mixture was stirred for 5 min
when TLC analysis showed the disappearance of the starting material.
The reaction mixture was concentrated to 4 mL, and 10 mL of H2O
(61) Alewood, P.; Alewood, D.; Miranda, L.; Love, S.; Meutermans, W.; Wilson,
D. Methods Enzymol. 1997, 289, 14.
9
4638 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004

R-Helical Peptide Nucleic Acid Concept
A R T I C L E S
c. The resin was washed with DMF (1 × 25 mL, flush; 3 × 10 mL,
shake-flush).
Desalting conditions: XTerra RP18 (7 µm, 19 × 150 mm) column,
flow rate ) 15 mL/min, 0-15 min 100% A, 15-30 min 0-100% B.
Mass Spectral Analysis. MALDI-TOF mass spectra were recorded
on a Kratos Kompact MALDI-TOF mass spectrometer. A 1:1 mixture
of the sample and the sinapinic acid matrix were run in a linear mode,
and the spectrometer was calibrated with insulin as internal standard.
Electrospray mass data were recorded on a Micromass Quattro II triple-
quadrupole ESI mass spectrometer. Observed masses were derived from
the mass of multicharged species using MaxEnt and Transform software
packages. The standard deviation value in the observed mass originated
during such calculations is not a reflection of measurement-error and
only represents upper and lower limits of the calculated values. Mass
data for the RPNAs described in this paper are collected in Tables S1
and S2.
d. Coupling efficiency monitoring: The conventional Kaiser (nin-
hydrin) test62 is not ideal for monitoring the solid-phase synthesis of
RPNA. It was observed that once the nucleoamino acids are loaded
onto the resin the test is always slightly positive (light blue), even after
double or triple couplings. Fmoc-counting is the method of choice for
monitoring coupling efficiency and can be done by measuring the UV
absorbance of the dibenzofulvene chromophore (ꢀ ) 7800 M-1 cm-1
at 301 nm) released upon the Fmoc cleavage by a 20% (v/v) solution
of piperidine in DMF. Generally, a single HATU mediated coupling
was sufficient for all natural and nucleoamino acids.
N-Capping (acetylation). a. A mixture of Ac2O (0.5 mL) + DIPEA
(0.85 mL) + DMF (4.0 mL) was added to the resin and shaken for 15
min.
3. Binding Studies. Materials. ssDNA was obtained from various
companies: Cybersyn Inc., Operon Technology, and Integrated DNA
Technology. The commercial oligonucleotides were purified if their
purity was less than 95% by analytical HPLC. For a typical reversed-
phase HPLC analysis of DNA, an XTerra RP18 column (3.5 µm, 4.6 ×
50 mm) was used. The mobile phase was composed of buffer A (0.1
M triethylammonium acetate (TEAA), pH 7.2) and buffer B (0.1 M
TEAA, 20% acetonitrile, pH 7.2), with a linear gradient of B in A (5
to 95% B in 25 min). Flow rate: 1 mL/min. Detection wavelength:
254 nm. For a semipreparative purification, an XTerra RP18 column
(Waters, 7 µm, 19 × 150 mm) was used with the same gradient as the
analytical system and a flow rate of 15 mL/min. The concentrations of
RPNA solutions were determined by UV absorbance using the nearest
neighbor approximation63 assuming that the extinction coefficients of
the nucleobases in RPNAs are the same as in DNA. Concentration of
abasic (control) peptide was determined by quantitative ninhydrin
analysis.64
b. Resin washed with DMF (3 × 10 mL, 1 min shake-flush) and
then DCM (2 × 15 mL, 1 min shake-flush).
CleaVage from Resin. After solid-phase synthesis, the resin was
treated with 95% TFA in water (v/v) [3 mL for 100 mg of resin] at
room temperature for 1 h, with occasional shaking and sonication. The
orange red solution obtained was filtered through glass-wool, and the
resin was washed with 95% TFA in water (5 × 1 mL for 100 mg of
resin). TFA solution was evaporated on a rotary evaporator at RT to
give an orange oil (1-2 mL). Cold ether (ice/water) was added to this
orange oil with mixing until an off-white precipitate formed which was
washed with ether (5 × 3 mL) and purified as described below.
Analysis and Purification. All crude RPNAs and peptides obtained
after resin cleavage were filtered through a cartridge (0.5 × 3 cm) of
C18 resin (eluted with 0.1% TFA in 9:1 CH3CN-H2O) or Sephadex
G-10 gel (eluted with 0.1% TFA H2O) to remove the nonpeptide
impurities. Buffers: A ) 0.1% aqueous TFA, B was 0.1% TFA in
CH3CN, C ) 100 mM TEAA (triethylammonium acetate), pH 7.2; D
) 100 mM TEAA, 20% CH3CN, pH 7.2), E ) 8 mM NaH2PO4 in 9:1
water-CH3CN, pH adjusted to 4 by addition of dilute HCl, F ) E +
0.4 M NaCl. RPNA detection wavelength was 254 nm.
Chromatographic Analysis. Analytical HPLCs were performed on
a XTerra RP18 column (Waters, 3.5 µm 4.6 × 50 mm). A linear gradient
of buffer A and B with of 2% increment in buffer B/min and flow rate
of 1 mL/min was used. Buffers C and D were used for RPNAs that
showed inferior resolution and/or peak broadening in buffers A and B.
In general, sufficiently pure 21 mer RPNAs (>60% by HPLC) were
obtained after cleavage from the resin.
UV Melting Experiments. For b1 RPNA, thermal UV denaturation
curves were obtained with sample solutions made by combining the
RPNA and ssDNA components (final concentration was 5 µM in each
oligomer) in HPLC grade water followed by heating at 75 °C for 5
min and then cooling and storage at 4 °C for 4 days. For b2 RPNA,
thermal UV denaturation curves were obtained with sample solutions
made by combining the RPNA and ssDNA components (final concen-
tration was 5 µM in each oligomer) in TE buffer (10 mM Tris-HCl, 1
mM EDTA, pH 7) followed by heating at 75 °C for 5 min and then
cooling and storage at 4 °C for overnight. All measurements were
conducted in a 1-cm path length quartz cell equipped with a temperature
probe. The absorbance at 260 nm was recorded as a function of
temperature using a Perkin-Elmer Lambda 20 UV/VIS spectrophotom-
eter equipped with a PTP-1 Peltier system thermocontroller with
heating/cooling rates of 1 °C/min over the range of 5 to 75 °C. Below
20 °C, dry N2 gas was passed through the spectrophotometer sample
chamber to prevent moisture condensation. Data were collected and
evaluated using the TempLab software package. The melting temper-
ature, Tm (defined as the temperature at which 50% of a complex is
dissociated into its constituent components) was determined from the
inflection point maximum of the first derivative of the melting curves
and is the average of two separate runs.
SemipreparatiVe and PreparatiVe Scale Purification. For a typical
RPNA purification, a Water’s XTerra RP18 (7 µm, 19 × 150 mm)
column was used. Similar A-B or C-D buffer-gradients at a flow
rate of 15 mL/min were used. Typically 5-6 mg of the crude sample
was purified at a time. An absorbance unit (AU) setting (Varian’s dual
flow cell UV-detector) of 5 was used during the separation for optimal
performance. Preparative scale HPLC purifications were performed with
a C18-Nucleosil (7 µm, 50 × 250 mm) column. Crude RPNA obtained
from a 54 µmol scale synthesis was purified in one run. The same
gradient as the semipreparative HPLC was used with a flow rate of 30
mL/min. An AU of 20 is generally the correct setting to obtain a full-
scale chromatogram.
Circular Dichroism Studies. CD spectra were either recorded on a
Jasco J-600 or J-810 spectropolarimeter. CD spectra were the average
of eight scans obtained by collecting data at 0.10 nm intervals from
320 to 190 nm, with a response time of 2 s, bandwidth of 1 nm and
sensitivity of 20 mdeg. Stoppered optical cells of path length 1 cm
were used. Unless otherwise specified, all the samples were constituted
with HPLC grade water. For the experiments involving only RPNA,
the measured original CD data (expressed as ellipticity [θ]obs: in mdeg)
was converted to the optical constant (molar ellipticity). Molar ellipticity
Dimerization of rPNAs. Dimerization of backbone 2 RPNAs was
achieved by mixing vigorously a solution of RPNA (2.5 mM) in 0.5
M HCl with I2 in methanol (2 equiv of a 73 mM solution) for 30 min.
The reaction mixture was filtered through a Sephadex-G10 gel to
remove I2, followed by a strong cation exchange HPLC purification
and desalting by C18 reverse phase HPLC. The same conditions apply
for the backbone 1 RPNAs; however, the presence of HCl was not
necessary. HPLC conditions: Vydac 400VHP575 SCE column. Linear
E-F gradient at 3.33% increment in F/min, flow rate ) 1 mL/min).
is reported as the mean residue molar ellipticity ([θ], deg‚cm2‚dmol-1
)
and is calculated from the following equation: [θ] ) [θ]obs(mrw)/10lc
(62) Sarin, K. V.; Kent, B. H. S.; Tam, P. J. Anal. Biochem. 1981 117, 6147.
(63) Puglisi, J. D.; Tinoco, I., Jr. Methods Enzymol. 1989, 180, 304.
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4639

A R T I C L E S
Huang et al.
Where mrw is the mean residue molecular weight (molecular weight
of the peptide divided by the number of amino acid residues), c is the
peptide concentration in grams per milliliter, and l is the optical path
length of the cell in centimeter.
Sample Preparation for the NMR Experiments. DNA d(A3-
GGAGGA3) and the RPNA CCTCC (b2) were used for the NMR
experiments. The recommended concentration of the individual oligo-
mer strands (RPNA and ssDNA) for one- and two-dimensional NMR
experiments is 0.5-1 mM for a superior signal-to-noise ratio. The
acidity of the RPNA solution obtained directly from the purification
by HPLC in TFA buffer was adjusted by addition of NH4HCO3 solution
(1 M, 4 equiv to RPNA concentration) and lyophilization of the resulting
solution. The resulting white powder was dissolved in water and re-
lyophilized for four times before the final solution was made.
Commercially available ssDNA was purified by HPLC using a C18
column and TEAA buffer system first, and then the excess TEAA was
exchanged with ammonium bicarbonate. The buffer exchange was
accomplished by a C18 chromatography using an elution gradient of
10 mM NH4HCO3 solution versus CH3CN. The column was eluted
first with 100% 10 mM NH4HCO3, pH 8.0 for 10 min, then ramped to
100% Buffer B (10 mM NH4HCO3, 20% CH3CN, pH 8.2.) within 30
min. The resulting ssDNA solution was lyophilized in manner similar
to RPNA, as described above, before making the final solution. Samples
of ssDNA and RPNA obtained from the above sources were individually
reconstituted with a 20 mM sodium phosphate buffer solution (pH 7.0,
95:5 water-D2O mixture), and mixed together.
For the TTTTT(b1)-dimer, a thermal denaturation profile between
30 and 80 °C was obtained. The helical content at 20 °C was estimated
from a separate CD experiments by taking the ratio of ([θ]219 - [θ]0)/
[θ]100, where [θ]219 is the mean residual helicity at 219 nm () -14 850
deg‚cm2‚dmol-1), [θ]0 is the “background” mean residue ellipticity at
0% helicity () -5368 deg‚cm2‚dmol-1) as determined by a melting
experiment. To account for the length dependence of CD for an R-helix,
the following formula was used to calculate [θ]100;
65 [θ]100 ) [θ]H∞ (1
∞
- k/n). Where [θ]H is the molar ellipticity for an infinite helix
(-38 400 deg‚cm2‚dmol-1), k is a wavelength-dependent factor ()
2.69), and n is the number of residues in the helix () 42). The R-helix
percentage in the structure was determined to be 26% by taking the
ratio of [θ]219(obs)/[θ]219(100% helix).
Gel Mobility Shift Experiments. Solutions were made by com-
bining DNA (300 pmole) with varying amounts of RPNA in 7.5 µL of
TE-buffer followed by heating at 75 °C for 5 min and then cooling
and storage at 4 °C for overnight. Before electrophoresis, 2.5 µL of
loading buffer (0.01% xylene cyanol FF, 0.01% bromophenol blue
solution, 60% (w/v) glycerol in 44 mM Tris-borate, pH 7.2) was added
to the sample and mixed. An amount of 3.5 µL of the sample was
loaded onto the gel. Free and RPNA bound DNA were resolved by
nondenaturing 14% polyacrylamide gel electrophoresis in 44 mM Tris-
borate, pH 7.2 for 1 h at 14 V/cm at 4 °C. The gel was visualized by
silver stain using the PlusOne DNA silver staining kit (Pharmacia
Biotech).
NMR Spectroscopy. All NMR spectra were recorded on a Varian
Inova 600 spectrometer equipped with a triple-resonance gradient probe.
Unless stated otherwise, the spectra were recorded at a probe temper-
ature of 5 °C. All experiments were performed with WATERGATE
water suppression.66 The NMR data were processed and analyzed using
the program FELIX 98 (Molecular Simulations, Inc.).
Acknowledgment. This work was supported by grants from
the National Institutes of Health (GM-54796) and by the Center
for AIDS Research at Case Western Reserve/University Hos-
pitals of Cleveland (AI-36219). Thanks to the Orlando Research
Group (Complex Carbohydrate Research Center, University of
Georgia) for performing the MALDI-TOF mass spectrometry
and Dale Ray (Case) for assisting us with HMBC/HMQC NMR
experiments. Special thanks Prof. Tony Berdis, Prof. Irene Lee,
Prof. Peter DeHaseth, and Gianina Panaghie for technical
assistance and helpful comments.
Determination of Dissociation Constant (Kd) by Gel Shift Assay.
For determination of Kd, the following equation was assumed for DNA
(D) and RPNA (P) interaction to form the complex RPNA-DNA
(DP): D + P T DP; The concentrations of bound and free DNA, [Db]
and [Df] respectively, can then be related to equilibrium constant: Kd
) [Df]([Po] - [Db])/[Db]; [Po] and [Do] are the total concentration of
RPNA and DNA, respectively. In the gel shift experiment, the total
concentration of DNA was fixed (30 µM) and the concentration of
RPNA was varied. F is defined as the molar ratio of the RPNA and
DNA ([Po] ) F[Do]). [Db] was determined from the intensity of the
silver stained gel band corresponding to the retarded RPNA-DNA
complex using Quantity One software (Bio-Rad). It is assumed that
no dissociation of the RPNA-DNA occurs during electrophoresis. Kd
was obtained by plotting either F against [Db] or 1/([Do] - [Db]) against
F/Db (linear plot) using the following equation: F/[Db] ) KD/{[Do]-
([Do] - [Db])} + 1/Do.
Supporting Information Available: MS data for all new
compounds and experimental protocols for the RPNA-ssDNA
binding studies. This material is available free of charge via
the Internet at http://pubs.acs.org/.
JA038434S
(64) Sarin, V. K.; Kent, S. B.; Tam, J. P.; Merrifield, R. B. Anal. Biochem.
1981, 117, 147.
(65) Chen, Y.; Yang, J. T.; Chau, K. H. Biochemistry 1974, 13, 3350.
(66) Piotto, M.; Saudek, V.; Sklenar, V. J. Biomolecular NMR 1992, 2, 661.
9
4640 J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004