Autonomously pairing cysteinyl-linked nucleotide analogues with a unique architecture

Article abstract of DOI:10.1021/ja200829s

We report the efficient pairing in water of the first representative of oligonucleotide analogues in which the backbone is replaced by linking elements between the nucleobases. The architecture of the new analogue demonstrates that the structural differen

Full text of DOI:10.1021/ja200829s

COMMUNICATION
pubs.acs.org/JACS
Autonomously Pairing Cysteinyl-Linked Nucleotide Analogues with a
Unique Architecture
Manuel Peifer and Andrea Vasella*
Laboratory for Organic Chemistry, ETH Z€urich, CH-8093 Z€urich, Switzerland
S Supporting Information
b
ABSTRACT: We report the efficient pairing in water of the
first representative of oligonucleotide analogues in which
the backbone is replaced by linking elements between the
nucleobases. The architecture of the new analogue demon-
strates that the structural differentiation of oligonucleotides
into a contiguous backbone and nucleobases, as embodied
by the natural nucleic acids and all nucleotide analogues
analyzed to date, is not a prerequisite for pairing. UV and
circular dichroism analyses of self-complementary and non-
self-complementary octanucleotide analogues strongly sug-
gest the fully reversible, sequence-specific association of our
new analogues to form a left-handed double helix with an
antiparallel strand orientation that is characterized by melt-
ing temperatures and free enthalpies higher than those of
natural RNA and DNA of the same sequence. The linking
Figure 1. a) Definitions of the torsion angles and numbering of a
cysteinyl-linked pyrimidine-purine (UA) dinucleotide analogue. (b)
element incorporates an L-cysteine moiety that allows a
short and efficient synthesis of the monomeric building
Conformation of the UA dinucleotide analogue with torsion angles κ-φ
blocks and, through the choice of either L- or D-cysteine,
required for pairing [termini on N(1) and C(8) have been replaced by
methyl groups for clarity].
gives access to either one of the enantiomeric oligomers and
thus to left- or right-handed helices.
desired oligomers by one of the numerous established methods
of amide coupling and deprotection in solution or on a solid
support. We aimed at designing the new linker to be short
enough to allow favorable twist angles but long enough to
maintain sufficient flexibility, allowing the oligonucleotide ana-
xtensive work by a number of research groups, including
E
those of Eschenmoser,1 Nielsen,2 and Kool,3 has addressed
logues to adapt to the stereochemical requirements of pairing
conformers. We realized this intent by incorporating an L-
cysteinyl moiety (Figure 1). According to a conformational
analysis based on vibrational spectra,11 NMR data,8 and crystal
structures,12 we conclude that the sulfur atom favors a first
synclinal arrangement of CH2(10) and CH(40) and that it further
contributes to the flexibility of our new analogues on account of
the low energy barriers for rotation about C-S bonds.11 The
polar side chain attached to C(40) allows for a second crucial
synclinal arrangement of S(20) and NH(50) and ensures the
water solubility of the oligomers at neutral or basic pH. The
preferred conformation of the aminoacetyl fragment attached to
N(9) (as shown in Figure 1) is well-precedented by crystal and
NMR structures of peptide nucleic acids (PNAs).13 These
conformational aspects suggest that the cysteinyl linker adopts
a conformation similar to the one shown in Figure 1, where the
nucleobases are positioned at a favorable distance for stacking
and in the right orientation for hydrogen bonding in the
Watson-Crick mode. In agreement with the common structural
the synthesis and analysis of oligonucleotide analogues with
altered backbones and nucleobases. In accord with the structure
of naturally occurring nucleic acids and oligonucleotides, all of
these analogues are characterized by the structural differentiation
between a contiguous backbone and the nucleobases attached to
it. Over the last several years, our group has described a series of
oligonucleotide analogues that diverge from this common struc-
tural motif. Instead of a contiguous backbone, they possess
linking elements directly between adjacent nucleobases. Such
oxymethylene-,4 ethynylene-,5 (Z)-ethenylene-,6 thiomethy-
lene-,7,8 sulfonyl- and sulfinylmethylene-,9 and aminomethylene-
linked10 di- and tetranucleosides, called “oligonucleotide analo-
gues with integrated bases and backbone” (ONIBs), pair effi-
ciently in organic solvents, thus deepening our insight into the
limits within which the structure of pairing nucleic acids can be
varied.
Having shown that ONIBs are capable of pairing in organic
solvents, we designed a new type of these analogues that allows
their pairing properties in aqueous solution to be evaluated. As
pairing in water requires longer oligomers, we chose an amide
group as a key element of the new linker to allow assembly of the
Received: February 3, 2011
Published: March 08, 2011
r
2011 American Chemical Society
4264
dx.doi.org/10.1021/ja200829s J. Am. Chem. Soc. 2011, 133, 4264–4267
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. Structures of the Monomeric Building Blocks
3-6 and Summary of the Synthesis of the Octanucleotide
Analogues U2C2G2A2 (1) and U2C2U2C2 (2) of the
Type Illustrated by the Depicted General Structure of a
Dimeric Unit
Figure 2. Model of the octanucleotide analogue U2C2G2A2 (1): (a) side
view; (b) top view.
motif of ONIBs, the linker is attached to N(1) and C(6) of a
pyrimidine base and to N(9) and C(8) of a purine base, thus fusing
the nucleobases and the linker into a single linear structural unit.
Amber* force field calculations14 of cysteinyl-linked dimers
suggest that the surface overlap of the nucleobases in the
cysteinyl-linked analogues depends even more strongly on the
sequence than in nucleic acids14-16 and PNAs.13 The surface
overlap decreases in the order pyr pur ≈ pyr pyr ≈ pur pur >
3
3
3
pur pyr (from the C- to the N-terminus, with pyr representing a
3
pyrimidine base and pur representing a purine base). We thus
synthesized the self-complementary octamer 1 having the se-
quence U2C2G2A2 and the non-self-complementary octamer 2
having the sequence U2C2U2C2 (naming from the C- to the
N-terminus), which are devoid of a pur/pyr switch. Minimization
of the conformational energy of 1 after releasing all of the
constraints between the hydrogen bond donors and acceptors
(Amber* force field calculation14) resulted in a left-handed
double helix (Figure 2) with the left-handedness dictated by
the L-cysteine moiety.
The helix is stabilized by Watson-Crick-type base pairing and
base stacking, with the cysteinyl linker adopting a conformation
having the values of the torsion angles κ-φ that we deduced for
the pairing conformation shown in Figure 1. The helix is
characterized by a rise of 3.3-3.4 Å per base pair (cf. B-DNA,
3.4 Å; A-RNA, 2.8 Å), a pitch of ∼26 Å (cf. 34 Å for B- and
A-RNA), and a twist angle of ∼42° (cf. B-DNA, 36°; A-RNA,
31°).14,15
We synthesized the fully protected monomeric building
blocks 3-5 in short sequences of four to five steps and overall
yields of 19-36% from uracil, cytosine, and adenine, respec-
tively. The fully protected monomeric building block 6 was
obtained in eight steps and 18% overall yield from commercially
available 2,6-diamino-4-chloropyrimidine (Scheme 1). Self-com-
plementary octameric 1 and non-self-complementary octameric
2 were synthesized by convergent coupling of the N-Fmoc- or N-
tert-butoxycarbonyl-deprotected building blocks 3-6 via di- and
tetrameric intermediates and a final deprotection in 24 steps and
6% overall yield for 1 and 15 steps and 24% overall yield for 2 (see
the Supporting Information).
The UV absorbance of the self-complementary octanucleotide
analogue 1 (c = 10 μM in 10 mM sodium phosphate buffer,
100 mM NaCl, 0.1 mM EDTA, pH 7.0) changed nonlinearly
with linearly increasing temperature, evidencing cooperative
stacking and thus pairing (Figure 3, top plot). The UV
absorbance of the non-self-complementary octanucleotide
Figure 3. (top) UV melting curves and (bottom) temperature-depen-
dent CD spectra (solid lines, 10 °C steps from 10 to 90 °C) and UV
spectra (dotted line at 4 °C, dashed line at 80 °C) of self-complementary
U2C2G2A2 (1) in 10 mM sodium phosphate buffer, 100 mM NaCl, and
0.1 mM EDTA at pH 7.0 (c = 10 μM, 1 cm cell).
analogue 2, however, changed only linearly with linearly
increasing temperature, evidencing the absence of coopera-
tive stacking and thus suggesting the absence of pairing.17 The
self-recognition of 1 unambiguously proves sequence-specific
hydrogen bonding between complementary nucleobases and
thus an antiparallel orientation of the associating strands.
4265
dx.doi.org/10.1021/ja200829s |J. Am. Chem. Soc. 2011, 133, 4264–4267

Journal of the American Chemical Society
COMMUNICATION
Similar to the UV melting curves of the octanucleotides
U2(T2)C2G2A2 of RNA and DNA (for the spectra of 2,
RNA, and DNA, see the Supporting Information), the UV
melting curves resulting from heating (denaturation) and
cooling (renaturation) experiments with self-complementary 1
were identical, evidencing fully reversible pairing and ther-
modynamic control of the association. The remarkable
stability of the associate of 1 is reflected by its melting
temperature of 46.8 °C (c = 10 μM, 10 mM sodium phos-
phate buffer, 100 mM NaCl, 0.1 mM EDTA, pH 7.0), which is
higher than those of RNA and DNA with the same sequence,
as obtained under the same conditions (cf. RNA, 44.9 °C;
DNA, 31.4 °C).
The transition of 1 in the UV melting profile is considerably
sharper than those for RNA and DNA and is also reflected by the
abrupt decrease of the strong circular dichroism (CD) signal of 1
precisely at the melting temperature of 47 °C (compare the CD
spectra at 40 and 50 °C in the bottom panel of Figure 3),
evidencing a high cooperativity during the pairing of 1. The CD
spectra of the non-self-complementary cysteinyl-linked analogue
2 at various temperatures, in contradistinction, show very small
ellipticities and no temperature dependence (see the Supporting
Information).
(monomolecular equilibrium), since the melting temperature
of 1 depends on the concentration (cf. c = 5 μM, Tm = 44.7 °C;
c = 10 μM, Tm = 46.8 °C; c = 100 μM, Tm = 55.4 °C; 10 mM
sodium phosphate buffer, 100 mM NaCl, 0.1 mM EDTA, pH
7.0). Triplex formation (trimolecular equilibrium) requires a
homopyrimidine strand and a homopurine strand consisting of
pyr pur pyr and/or pyr pur pur triads. CG Cþ and TA Tþ
3
3
3
3
3
3
triads require protonation, while the stability of pyr pur pur
3
3
triads depends on the presence of a bivalent metal cation.21
However, 1 does not possess a pure homopyrimidine or pure
homopurine sequence; moreover, its association does not re-
quire bivalent cations, and its Tm does not decrease with increas-
ing pH (cf. pH 7.0, Tm = 46.8 °C; pH 8.0, Tm = 47.5 °C; pH 8.8,
Tm = 47.0 °C; c = 10 μM in 10 mM sodium phosphate buffer,
100 mM NaCl, 0.1 mM EDTA). Also, considering the single
transition of 1 in its UV melting profile, the formation of a
triplex by 1 is highly improbable. Association of 1 to form a
G-quadruplex or an i-motif (tetramolecular equilibria) requires
stabilization of the associate by mono- or bivalent cations
[preferentially Kþ, Rbþ, or Naþ (150 mM-1 M) or Mg2þ
,
Ca2þ, Ba2þ, or Sr2þ (>10 mM)] or by protonation.22 The
octamer 1, however, associates even more strongly in the
presence of the only very weak G-quadruplex initializer Liþ (cf.
10 mM sodium phosphate buffer þ 100 mM NaCl, Tm = 46.8 °C;
10 mM lithium phosphate buffer þ 100 mM LiCl, Tm = 50.9 °C
at 10 μM concentration and pH 7.0) as well as at basic pH (see
above). Also, the hysteresis characteristic of tetramolecular
G-quadruplexes22 was not observed. The progressive increase
in the melting temperature of 1 with increasing salt concentration
is in agreement with the findings for DNA and RNA duplexes (cf.
10 mM sodium phosphate buffer, no NaCl, Tm = 24.9°; 10 mM
sodium phosphate buffer, 1 M NaCl, Tm = 63.7° at 10 μM
concentration and pH 7.0).23
The UV spectrum of 1 at 4 °C shows a bathochromic shift
relative to its UV spectrum at 80 °C (λmax from 272 to 268 nm;
see Figure 3, bottom panel, dashed and dotted lines). We
attribute this observation to a pronounced stacking of the inner
C2G2 fragment, as compared with the somewhat weaker stacking
of the C-terminal U2 and N-terminal A2 fragments (λmax for a
C G base pair, ∼280 nm; λmax for a T A base pair, ∼260 nm).18
3
3
As a consequence, 1 shows only a small hyperchromic shift in the
UV melting profile at a wavelength of 260 nm (5%), whereas one
observes remarkable hypochromic shifts at λ = 280 nm (17%)
and 290 nm (66%) (Figure 3, top panel). The melting tempera-
tures calculated from the curves at 260 and 290 nm are very close
(46.8 and 47.5 °C).
The analytically evidenced stacking, the pairing via hydro-
gen bonding (as evidenced by the self-recognition of 1 and the
large values of -ΔH298 and -ΔG298), and the considerations
regarding the stoichiometry provide evidence for the associa-
tion of 1 in a double-helical structure similar to that proposed
in Figure 2.
In conclusion, efficient pairing of the first oligonucleotide
analogue with integrated bases and backbone (ONIB) in water
has been demonstrated. Analytical data and molecular modeling
suggest that L-cysteinyl-linked ONIBs associate to form a left-
handed double helix. The chirality of the cysteine moiety allows
the synthesis of either one of the enantiomeric cysteinyl-linked
helical oligomers.
The stability of the associate of the cysteinyl-linked octamer 1
is expressed not only by the higher melting temperature but also
by the high value of the negative free energy (-ΔG298) (cf. 1,
13.4 kcal/mol; RNA, 9.8 kcal/mol; DNA, 8.0 kcal/mol) and of
the negative energy of association (-ΔH298) (cf. 1, 89.1 kcal/
mol; RNA, 48.1 kcal/mol; DNA, 55.4 kcal/mol). The value
of -TΔS for 1 was calculated to be 75.7 kcal/mol (cf. RNA,
38.3 kcal/mol; DNA, 47.4 kcal/mol).19 In fair agreement with
the values obtained by Kool,20 -ΔG298 per base pair for 1
amounts to 1.7 kcal/mol (cf. RNA, 1.2 kcal/mol; DNA, 1.0 kcal/
mol; eight Watson-Crick base pairs, seven stacking inter-
actions). The large value of -ΔH for 1 suggests a combination
of stabilizing H-bonding and stacking,20 possibly combined with
a weaker strand repulsion of 1 relative to RNA and DNA. The
large value for -TΔS may be due to the conformational equilibria of
the dissociated strands (λ = -60° and λ = 180° and/or ξ = þ90°
and ξ = -90°) (Figure 1), which are “frozen” in the associated state
(λ = -70° and ξ = þ85°). The hydrophobic effect could also have
an influence on -TΔS, in that 1 could release less hydrated water
upon association than RNA and DNA.16
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures, spec-
b
tral data, and additional UV and CD spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
vasella@org.chem.ethz.ch
The self-complementarity of 1, its complex NMR spectra in
water, and the fact that it could not be crystallized prevented us
from unambiguously proving the stoichiometry of the associate
of 1. However, on the basis of a series of UV experiments, we are
confident that our assumption of a bimolecular associate is
correct. Specifically, we can exclude an intramolecular associate
’ ACKNOWLEDGMENT
We thank ETH Z€urich and Syngenta (Basel, Switzerland) for
generous support and Dr. Bruno Bernet for helpful discussions.
4266
dx.doi.org/10.1021/ja200829s |J. Am. Chem. Soc. 2011, 133, 4264–4267

Journal of the American Chemical Society
COMMUNICATION
’ REFERENCES
(22) (a) Sen, D.; Gilbert, W. Nature 1990, 344, 410–414. (b)
Venczel, E. A.; Sen, D. Biochemistry 1993, 32, 6220–6228. (c) Mergny,
J. L.; Phan, A. T.; Lacroix, L. FEBS Lett. 1998, 435, 74–78. (d) Hardin,
C. C.; Perry, A. G.; White, K. Biopolymers 2000, 56, 147–194.
(23) Schildkraut, C.; Lifson, S. Biopolymers 1965, 3, 195–208.
(1) (a) Beier, M.;Reck, F.;Wagner, T.;Krishnamurthy, R.;Eschenmoser,
A. Science 1999, 283, 699–703. (b) Schoning, K. U.; Scholz, P.; Guntha, S.;
Wu, X.; Krishnamurthy, R.; Eschenmoser, A. Science 2000, 290, 1347–1351.
(c) Eschenmoser, A. Tetrahedron 2007, 63, 12821–12843.
(2) (a) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science
1991, 254, 1497–1500. (b) Egholm, M.; Buchardt, O.; Nielsen, P. E.;
Berg, R. H. J. Am. Chem. Soc. 1992, 114, 1895–1897. (c) Egholm, M.;
Nielsen, P. E.; Buchardt, O.; Berg, R. H. J. Am. Chem. Soc. 1992,
114, 9677–9678.
(3) (a) Liu, H. B.; Gao, J. M.; Lynch, S. R.; Saito, Y. D.; Maynard, L.;
Kool, E. T. Science 2003, 302, 868–871. (b) Lynch, S. R.; Liu, H. B.; Gao,
J. M.; Kool, E. T. J. Am. Chem. Soc. 2006, 128, 14704–14711. (c) Wilson,
J. N.; Teo, Y. N.; Kool, E. T. J. Am. Chem. Soc. 2007, 129, 15426.
(4) Matthews, A. J.; Bhardwaj, P. K.; Vasella, A. Helv. Chim. Acta
2004, 87, 2273–2295.
(5) (a) Zhang, X.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2006,
89, 2861–2917. (b) Zhang, X.; Bernet, B.; Vasella, A. Helv. Chim. Acta
2007, 90, 792–819.
(6) Zhang, X.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2007,
90, 864–890.
(7) (a) Ritter, A.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2008,
91, 1675–1694. (b) Bernet, B.; Johar, Z.; Ritter, A.; Jaun, B.; Vasella, A.
Helv. Chim. Acta 2009, 92, 2596–2630.
(8) Ritter, A.; Egli, D.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2008,
91, 673–714.
(9) Bogliotti, N.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2010,
93, 659–667.
(10) (a) Chiesa, K.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2010,
93, 1822–1843. (b) Chiesa, K.; Shvoryna, A.; Bernet, B.; Vasella, A. Helv.
Chim. Acta 2010, 93, 668–691.
(11) (a) Plusquellic, D. F.; Suenram, R. D.; Mate, B.; Jensen, J. O.;
Samuels, A. C. J. Chem. Phys. 2001, 115, 3057–3067. (b) Vansteenkiste,
P.; Pauwels, E.; Van Speybr€ock, V.; Waroquier, M. J. Phys. Chem. A 2005,
109, 9617–9626.
(12) (a) Kashino, S.; Ashida, T.; Kakudo, M. Acta Crystallogr., Sect. B
1974, 30, 2074–2076. (b) Voliotis, S.; Cordopatis, P.; Main, P.; Nastoroulos,
V.; Germain, G. Acta Crystallogr., Sect. C 1986, 42, 1777–1780.
(13) (a) Rasmussen, H.; Sandholm, J. Nat. Struct. Biol. 1997,
4, 98–101. (b) Betts, L.; Josey, J. A.; Veal, J. M.; Jordan, S. R. Science
1995, 270, 1838–1841. (c) Eriksson, M.; Nielsen, P. E. Nat. Struct. Biol.
1996, 3, 410–413. (d) Brown, S. C.; Thomson, S. A.; Veal, J. M.; Davis,
D. G. Science 1994, 265, 777–780.
(14) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Caufield, C.; Lipton, M.; Chang, G.; Hendrickson, T.; Still, W. C.
J. Comput. Chem. 1990, 11, 440.
(15) Evaluation of the structural properties of DNA and RNA was
performed according to the default settings for nucleic acids in Macro-
model 7.0.
(16) Saenger, W. Principles of Nucleic Acid Structure; Springer-Verlag:
New York, 1984.
(17) (a) Marky, L. A.; Breslauer, K. J. Biopolymers 1987,
26, 1601–1620. (b) Mergny, J. L.; Lacroix, L. Oligonucleotides 2003,
13, 515–537. (c) Owczarzy, R. Biophys. Chem. 2005, 117, 207–215.
(18) Puglisi, J. D.; Tinoco, I. Methods Enzymol. 1989, 180, 304–325.
(19) (a) SantaLucia, J., Jr.; Allawi, H. T.; Seneviratne, P. A. Biochem-
istry 1996, 35, 3555–3562. (b) Xia, T. B.; SantaLucia, J., Jr.; Burkard,
M. E.; Kierzek, R.; Schr€oder, S. J.; Jiao, X. Q.; Cox, C.; Turner, D. H.
Biochemistry 1998, 37, 14719–14735. (c) SantaLucia, J., Jr. Proc. Natl.
Acad. Sci. U.S.A. 1998, 95, 1460–1465.
(20) (a) Kool, E. T. Annu. Rev. Biophys. 2001, 30, 1–22. (b) Kool,
E. T. Chem. Rev. 1997, 97, 1473–1487. (c) Kool, E. T.; Morales, J. C.;
Guckian, K. M. Angew. Chem., Int. Ed. 2000, 39, 990–1009. (d)
Yakovchuk, P.; Protozanova, E.; Frank-Kamenetskii, M. D. Nucleic Acids
Res. 2006, 34, 564–574.
(21) Frank-Kamenetskii, M. D.; Mirkin, S. M. Annu. Rev. Biochem.
1995, 64, 65–95.
4267
dx.doi.org/10.1021/ja200829s |J. Am. Chem. Soc. 2011, 133, 4264–4267