Control of the chirality and helicity of oligomers of serinol nucleic acid (SNA) by sequence design

Article abstract of DOI:10.1002/anie.201006498

Mimicking the masters: An artificial nucleic acid, serinol nucleic acid (SNA), formed oligomers in which natural nucleobases were tethered through its 2-amino-1,3-propanediol (serinol) scaffold. These SNA oligomers have two unique properties: their chiral

Full text of DOI:10.1002/anie.201006498

DOI: 10.1002/anie.201006498
Artificial Nucleic Acids
Control of the Chirality and Helicity of Oligomers of Serinol Nucleic
Acid (SNA) by Sequence Design**
Hiromu Kashida, Keiji Murayama, Takasuke Toda, and Hiroyuki Asanuma*
The construction of an artificial double helix that mimics
natural DNA or RNA has been one of the most challenging
endeavors in chemistry.[1–9] Recently, Meggers and co-workers
showed that even a simple acyclic propylene glycol with two
carbon atoms in the main chain (see (S)-GNA in Figure 1) as
a nucleobase tether could form a more stable duplex than that
of native DNA or RNA.[10–13] This pioneering work prompted
us to synthesize a new foldamer with three carbon atoms in
the main chain, acyclic threoninol nucleic acid (aTNA), from
d-threoninol.[14] We found that aTNA has the following
properties: 1) duplex formation involves complementary
pairing in an antiparallel fashion, as for natural DNA or
RNA; 2) owing to the flexibility of the backbone, the single-
stranded state does not adopt a characteristic preorganized
structure; 3) the thermal stability of the duplex is far greater
than that of the natural DNA or RNA duplex and even higher
than that of the GNA duplex. Studies on aTNA as well as
GNA and peptide nucleic acid (PNA)[15,16] have confirmed
that scaffold rigidity is not a prerequisite for stable duplex
formation as previously thought. However, unlike PNA, with
an acyclic scaffold, aTNA can not cross-hybridize with either
natural DNA or natural RNA. Although A15 of (S)-GNA can
hybridize with U15, the incorporation of several GC pairs
severely destabilizes the duplex with RNA.[11] Thus, there are
no artificial nucleic acids comprising a fully acyclic backbone
with a phosphodiester linkage that can cross-hybridize with
DNA or RNA without sequence limitation. We hypothesize
that the threoninol scaffold is still not flexible enough to form
a duplex with natural DNA or RNA.
Figure 1. a) Chemical structures of DNA and SNA. S and R termini
were named according to the chirality of the terminal residue. The
SNA monomer for the DNA synthesizer is also shown (DMT=di-
methoxytrityl). b) The mirror image of SNA with an asymmetrical
Herein, we propose a new artificial nucleic acid, serinol
nucleic acid (SNA, see Figure 1a), with a 2-amino-1,3-
propanediol (serinol) scaffold, which is even more flexible
than threoninol. In comparison with aTNA (Figure 1a), the
sequence ((S)-AT-(R)) is identical to SNA with the reverse sequence
only structural difference is the lack of a methyl group next to
the amino group. However, this small change affords the SNA
oligomer a unique stereochemical property: since this methyl
group provides chirality, its absence makes the scaffold
achiral as well as flexible. Accordingly, the chirality of the
“pure” SNA oligomer synthesized from four SNA monomers
((S)-TA-(R)). c) SNA with a symmetrical sequence ((S)-TT-(R)) is
identical to its mirror image.
(or the helicity of its duplex) depends only on its sequence
(see below). This property is specific to the SNA oligomer;
DNA, RNA, and previously synthesized aTNA all have
chirality (or helicity) that is inherently determined by the
chirality of the scaffold.[17] In the present study, we first
demonstrated this unique stereochemical property and then
cross-hybridized the SNA oligomer, which was found to
recognize both DNA and RNA sequence specifically.
[*] Dr. H. Kashida, K. Murayama, T. Toda, Prof. Dr. H. Asanuma
Graduate School of Engineering, Nagoya University
Furo-cho, Chikusa-ku, Nagoya 464-8603 (Japan)
Fax: (+81)52-789-2488
E-mail: asanuma@mol.nagoya-u.ac.jp
The chemical structure of the SNA oligomer is shown in
Figure 1a. Serinol (2-amino-1,3-propanediol), which, like
DNA, has three carbon atoms in its backbone, is an achiral
diol. However, modification of the two hydroxy groups with
different functional groups to form an SNA monomer (Fig-
[**] This research was supported by a Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports, Science
and Technology (Japan).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201006498.
Angew. Chem. Int. Ed. 2011, 50, 1285 –1288
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1285

Communications
ure 1a) results in a chiral monomer. We synthesized four
chiral SNA phosphoramidite monomers by attaching four
nucleobases (A, T, G, C) through the amide bond of l-
serine[18–20] to avoid racemization, as described in the Sup-
porting Information.[21–23] When these optically pure mono-
mers were conjugated, the obtained SNA oligomers showed
unique stereochemical properties, as depicted in Figure 1b,c.
For example, the mirror image of a symmetrical oligomer,
such as T!T, is identical to the original dimer; thus, an SNA
oligomer with a symmetrical sequence is achiral (i.e., a meso
compound). On the other hand, when an A!T dimer is
synthesized from monomers, its mirror image does not
coincide with the original enantiomer because the unsym-
metrical A!T dimer is chiral. However, interestingly, its
enantiomer is identical to the dimer of reversed sequence,
T!A. More specifically, the chirality of the oligomer can be
exactly inverted by reversing the sequence of the chiral SNA
monomers: two enantiomers can be synthesized from the
same chiral monomers by programming the sequences
correctly, but not from their enantiomeric monomers.[24]
These unique properties of SNA might provide insight into
the selection of d-ribose as a scaffold for nucleic acids.
Sequences of SNA oligomers are shown in Figure 2. R and
S termini are named according to the chirality of the terminal
residues: the terminal residue incorporated in the first step of
oligomer synthesis has the R configuration, whereas the other
terminal residue incorporated in the last step of synthesis has
the S configuration. S1 is a random 8-mer SNA oligomer, and
S2 is its complementary strand in an antiparallel orientation.
S4, which has the reverse sequence of S1, is the enantiomer of
S1. Similarly, S3 is the enantiomer of S2. Accordingly, the S3/
S4 duplex is the enantiomer of the S1/S2 duplex, and the
helicity of the duplex should be inverted. Furthermore, the
two enantiomeric duplexes S1/S3 and S2/S4 are parallel
duplexes. On the other hand, the duplex composed of S5 and
S6, which have symmetrical, complementary sequences, is
achiral.
Figure 3. a) Melting profiles of S1/S2 and single-stranded S1 and S2.
Inset: Job plot of the hybridization between S1 and S2. b) CD spectra
at 208C of S1/S2, single-stranded S1, and single-stranded S2, and the
sum of the spectra of S1 and S2. c) Melting profiles of antiparallel
duplexes (S1/S2, S3/S4) and parallel duplexes (S1/S3, S2/S4). d) CD
spectra of duplexes S1/S2, S3/S4, and S5/S6 at 208C. H is hyper-
chromicity; De is the molar circular dichroism per duplex
(Lmolꢀ1 cmꢀ1). Conditions: [NaCl]=100 mm, pH 7.0 (10 mm phos-
phate buffer), [oligonucleotide]=2.0 mm (for melting profiles) or
4.0 mm (for CD measurements).
sigmoidal curve, whereas single-stranded S1 and S2 did not
show any transition above 208C. A Job plot of absorbance at
260 nm revealed that S1 and S2 formed a 1:1 complex
(Figure 3a, inset). Furthermore, the CD spectrum of S1/S2
was different from the sum of the spectra of single-stranded
S1 and S2 (Figure 3b).[25] These results clearly demonstrate
that S1 and S2 form a duplex. The Tm value of the duplex was
determined to be 51.18C (Table 1), which is
significantly higher than that of the correspond-
ing DNA (D1/D2: 29.08C) and RNA duplexes
(R1/R2: 38.98C). Thus, the SNA duplex was
much more stable than the DNA and RNA
duplexes.[26] Because the S3/S4 duplex is the
enantiomer of S1/S2, the Tm value of S3/S4
(51.28C) was the same as that of S1/S2 (51.18C;
Table 1 and Figure 3c). We also found that the
parallel combination of S1/S3 showed a typical
sigmoidal curve, and we determined its
Tm value to be 15.98C. This Tm value is about
358C lower than that of the antiparallel S1/S2
duplex. Thus, the SNA oligomer strongly rec-
ognized antiparallel complementary strands in
the same way that DNA and RNA do. We
further examined the sequence specificity of
the SNA oligomer. The incorporation of one
base mismatch (T–T mismatch) into the SNA
duplex (S1/S7) lowered its Tm value to 38.78C,
First, we measured the melting temperatures (Tm) of these
duplexes (Figure 3). The S1/S2 duplex showed a typical
Figure 2. a) Schematic illustration of the relationship between SNA duplexes.
b) Sequences of SNA, DNA, and RNA.
1286
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 1285 –1288

Table 1: Melting temperatures of SNA, DNA, and RNA duplexes.
Sequence
Direction
of duplex
T
m [8C][a]
RNA
SNA
DNA
SNA/DNA
SNA/RNA
1/2
3/4
1/3
2/4
5/6
1/7
antiparallel
antiparallel
parallel
51.1 (S1/S2)
29.0 (D1/D2)
38.9 (R1/R2)
23.5 (S1/D2)
21.2 (D1/S2)
21.4 (D3/S4)
20.0 (S3/D4)
<10 (S1/D3)
<10 (D1/S3)
<10 (D2/S4)
<10 (S2/D4)
19.1 (S5/D6)
18.5 (D5/S6)
<10 (D1/S7)
35.0 (S1/R2)
32.2 (R1/S2)
33.3 (R3/S4)
36.2 (S3/R4)
<10 (S1/R3)
<10 (R1/S3)
<10 (R2/S4)
<10 (S2/R4)
30.3 (S5/R6)
26.3 (R5/S6)
14.5 (R1/S7)
51.2 (S3/S4)
15.9 (S1/S3)
15.9 (S2/S4)
47.0 (S5/S6)
38.7 (S1/S7)
23.1 (D3/D4)
<10 (D1/D3)
<10 (D2/D4)
28.1 (D5/D6)
<10 (D1/D7)
37.9 (R3/R4)
<10 (R1/R3)
<10 (R2/R4)
34.3 (R5/R6)
23.1 (R1/R7)
parallel
–
antiparallel
[a] Conditions: [oligonucleotide]=2.0 mm, [NaCl]=100 mm, pH 7.0 (10 mm phosphate buffer).
which is 128C lower than that of fully matched S1/S2. The
decrease in Tm was comparable to that observed for the DNA
and RNA duplexes, which indicates that the bases in the SNA
oligomer formed Watson–Crick base pairs and recognized
complementary nucleobases, as in DNA and RNA.
melting profile and CD spectrum of S1/R2 revealed that the
SNA oligomer could also form a duplex with RNA: the
Tm value of S1/R2 was 35.08C, which is also comparable to
that of R1/R2 (38.98C). The Tm value of S1/R2 was even
higher than that of D1/R2 (27.38C), which indicates the
potential of SNA as an antisense agent. On the other hand, S4,
the enantiomer of S1, hybridized with neither D2 nor R2 (see
Table 1). On the basis of these results, we regarded the S and
R termini of the SNA oligomer as equivalent to the 5’ and 3’
termini of natural oligonucleotides with a ribose scaffold,
respectively. Interestingly, both S3 and S4 formed a stable
duplex with DNA and RNA in an “antiparallel” fashion,
We measured the CD spectra of the duplexes to evaluate
their helicity (Figure 3d). S1/S2 showed symmetrical positive
and negative Cotton effects, which indicated that this duplex
formed a right-handed helix. As expected, S3/S4 (the
enantiomer of S1/S2) with its reversed sequence showed
exactly inverse CD signals: negative and positive Cotton
effects appeared at around 260 nm. On the other hand, S5/S6,
which has symmetrical sequences, showed no induced CD,
because the S5/S6 duplex is achiral. Thus, the helicity of the
SNA duplex can be modulated by the sequence. Since the
symmetrical SNA duplex S5/S6, which does not exhibit
helicity, showed a similar Tm value (47.08C) to that of chiral
SNA (S1/S2: 51.18C), the helical preference (i.e., helical
preorganization) of the corresponding single strands did not
strongly affect the stability of the SNA duplexes. Although an
SNA monomer has a simple and symmetrical structure, SNA
has the prerequisites for a genetic carrier: SNA oligomers can
form a highly stable duplex (considerably more stable than
DNA or RNA duplexes) with an antiparallel orientation and
also specifically recognize the complementary SNA sequence.
However, the conformation of the SNA duplex is significantly
sequence-dependent, which might be a severe disadvantage
for recognition by proteins and other biomolecules. We
believe that one the reasons for the selection of ribose as the
scaffold for DNA and RNA is that the conformation of the
resulting duplexes is not sequence-dependent.
Next, we investigated the cross-hybridization ability of
SNA with DNA or RNA. As mentioned above, both GNA
and aTNA could not hybridize with DNA or RNA.[14]
Surprisingly, however, SNA can form a heteroduplex with
DNA and RNA: the S1/D2 duplex showed a sigmoidal
melting curve (Figure 4a), unlike single-stranded S1 or D2.
Furthermore, the CD spectrum of S1/D2 was completely
different from the sum of the spectra of single-stranded S1
and D2 (Figure 4b). These results demonstrate that the SNA
oligomer can hybridize with DNA. The Tm value of S1/D2 was
determined to be 23.58C, which is comparable to that of the
corresponding native duplex D1/D2 (29.08C). Similarly, the
Figure 4. a) Melting profiles of S1/D2, S1/R2, and single-stranded S1,
D2, and R2. b) Comparison of the CD spectra at 08C of S1/D2 and S1/
R2 with those of single-stranded S1, D2, and R2, and the sum of the
spectra of S1 and D2, and S1 and R2. c) Comparison of the CD spectra
at 08C of SNA–DNA (S1/D2), SNA–RNA (S1/R2), DNA (D1/D2), and
RNA (R1/R2) duplexes. d) Dependence of the Tm values of S1/S2, S1/
D2, S1/R2, D1/D2, and R1/R2 on salt concentration. Conditions:
[NaCl]=100 mm, pH 7.0 (10 mm phosphate buffer), [oligonucleo-
tide]=2.0 mm (for melting profiles) or 1.0 or 4.0 mm (for CD measure-
ments).
Angew. Chem. Int. Ed. 2011, 50, 1285 –1288
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1287

Communications
although the S3/S4 duplex formed a left-handed helix (see 3/4
and may find application in the development of nanomate-
rials.
in Table 1). The achiral SNA oligomers S5 and S6 also formed
duplexes both with DNA and RNA (see 5/6 in Table 1).
Accordingly, the helical preference of the SNA oligomer in
the single-stranded state did not much affect its ability to
recognize DNA or RNA. CD spectra of the S1/D2 and S1/R2
duplexes are shown in Figure 4b,c. The exciton couplet shows
that these heteroduplexes form a right-handed helix. Since
these CD spectra were similar to that of R1/R2, rather than
that of D1/D2, we think that SNA–DNA and SNA–RNA
duplexes both form an A-form-like structure.
Received: October 16, 2010
Published online: January 5, 2011
Keywords: antisense agents · artificial oligonucleotides ·
.
chirality · helical structures · nucleic acids
[1] D. J. Hill, M. J. Mio, R. B. Prince, T. S. Hughes, J. S. Moore,
Chem. Rev. 2001, 101, 3893.
Finally, we determined thermodynamic parameters for S1/
S2 as well as for cross-hybridized S1/D2 and S1/R2 (see
Table S2 in the Supporting Information). S1/S2 was 5.6 and
3.8 kcalmolꢀ1 more stable than D1/D2 and R1/R2, respec-
tively, in terms of the DG837 values of these duplexes. The
DH value of S1/S2 was more negative than those of D1/D2
and R1/R2, and the DS value was also more negative. These
results indicate that the SNA duplex is stabilized by a larger
enthalpy change. Since the dependence of the Tm value of S1/
S2 on the salt concentration was smaller than that of D1/D2
and R1/R2 (Figure 4d: the slope of the plot was greater for
D1/D2 and R1/R2 (at 3.6 and 3.4, respectively) than for S1/S2
(2.3)), the high stability of the SNA duplex is partly
attributable to the decrease in electrostatic repulsion between
phosphate anions.[27,28] The SNA monomer has three atoms
between the backbone and the base, whereas DNA and RNA
have only two: the one-atom-longer interior linker should
expand the helix radius and lessen the electrostatic repulsion
between phosphate groups. The dependence of the Tm values
of S1/D2 and S1/R2 on salt concentration was also weaker
than that observed for the DNA and RNA duplexes. This
observation indicates that decreased electrostatic repulsion as
well as the flexibility of the serinol scaffold also contribute to
the stabilities of the SNA–DNA and SNA–RNA duplexes. In
previous studies, the incorporation of SNA monomers into
DNA destabilized the duplex.[18–20] This destabilization might
reflect the structural difference between SNA and natural
nucleic acids. On the other hand, the oligomer containing only
SNA monomers could hybridize with DNA and RNA, in
contrast to GNA and aTNA. As we had intended with our
design, the flexible structure of the serinol scaffold facilitated
cross-hybridization with natural oligonucleotides by relaxing
the structural difference.
[2] M. Albrecht, Chem. Rev. 2001, 101, 3457.
[3] E. Yashima, K. Maeda, H. Iida, Y. Furusho, K. Nagai, Chem.
Rev. 2009, 109, 6102.
[4] K. C. Schneider, S. A. Benner, J. Am. Chem. Soc. 1990, 112, 453.
[5] S. A. Benner, Acc. Chem. Res. 2004, 37, 784.
[6] H. Kaur, B. R. Babu, S. Maiti, Chem. Rev. 2007, 107, 4672.
[7] D. H. Appella, Curr. Opin. Chem. Biol. 2009, 13, 687.
[8] H. Kashida, X. G. Liang, H. Asanuma, Curr. Org. Chem. 2009,
13, 1065.
[9] S. Zhang, C. Switzer, J. C. Chaput, Chem. Biodiversity 2010, 7,
245.
[10] L. Zhang, A. Peritz, E. Meggers, J. Am. Chem. Soc. 2005, 127,
4174.
[11] M. K. Schlegel, A. E. Peritz, K. Kittigowittana, L. Zhang, E.
Meggers, ChemBioChem 2007, 8, 927.
[12] M. K. Schlegel, X. Xie, L. Zhang, E. Meggers, Angew. Chem.
2009, 121, 978; Angew. Chem. Int. Ed. 2009, 48, 960.
[13] M. K. Schlegel, L.-O. Essen, E. Meggers, Chem. Commun. 2010,
46, 1094.
[14] H. Asanuma, T. Toda, K. Murayama, X. Liang, H. Kashida, J.
Am. Chem. Soc. 2010, 132, 14702.
[15] P. Wittung, P. E. Nielsen, O. Buchardt, M. Egholm, B. Norden,
Nature 1994, 368, 561.
[16] P. E. Nielsen, Acc. Chem. Res. 1999, 32, 624.
[17] Although a chiral SNA monomer tethering T was reported
previously, this unique property was never expressed by the
conjugation of SNA with natural DNA or RNA.[18–20]
.
[18] K. S. Ramasamy, W. Seifert, Bioorg. Med. Chem. Lett. 1996, 6,
1799.
[19] R. Benhida, M. Devys, J.-L. Fourrey, F. Lecubin, J.-S. Sun,
Tetrahedron Lett. 1998, 39, 6167.
[20] A. K. Sharma, P. Kumar, K. C. Gupta, Helv. Chim. Acta 2001, 84,
3643.
[21] L. Christensen, H. F. Hansen, T. Koch, P. E. Nielsen, Nucleic
Acids Res. 1998, 26, 2735.
[22] Z. Timꢀr, L. Kovꢀcs, G. Kovꢀcs, Z. Schmꢁl, J. Chem. Soc. Perkin
Trans. 1 2000, 19.
[23] T. Kofoed, H. F. Hansen, H. Ørum, T. Koch, J. Pept. Sci. 2001, 7,
402.
[24] Chirality control by sequence design is not possible with natural
oligonucleotides or linear peptides: M. Goodman, M. Chorev,
Acc. Chem. Res. 1979, 12, 1.
[25] S1 and S2 showed small induced CD signals, which indicated that
these strands had slight right-handed helicity even in the single-
stranded state.
In conclusion, we synthesized a unique artificial oligonu-
cleotide, the SNA oligomer. The helicity of the SNA duplex
could be controlled by appropriate design of the SNA-
oligomer sequence. The SNA duplex showed remarkably high
stability in comparison with DNA and RNA, partly as a result
of the decrease in electrostatic repulsion. Furthermore, the
SNA oligomer sequence-specifically recognized its comple-
mentary sequence in an antiparallel fashion with sufficient
thermal stability. SNA is the first example of an artificial
nucleic acid composed of a fully acyclic scaffold with a
phosphodiester linkage that can cross-hybridize with DNA
and RNA irrespective of the sequence. SNA oligomers may
be useful as biological tools, such as antigene/antisense agents,
[26] The Tm value of the S1/S2 duplex was about 108C lower than that
of the corresponding aTNA duplex, which indicates that the
methyl group of threoninol contributed to the high stability of
the aTNA duplex (see Table S1 in the Supporting Information).
[27] M. T. Record, C. F. Anderson, T. M. Lohman, Q. Rev. Biophys.
1978, 11, 103.
[28] S. Nakano, M. Fujimoto, H. Hara, N. Sugimoto, Nucleic Acids
Res. 1999, 27, 2957.
1288
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 1285 –1288