Synthesis and RNA binding selectivity of oligonucleotides modified with five-atom thioacetamido nucleic acid backbone structures

Article abstract of DOI:10.1021/ol070990u

Equation Presented Convenient chemical synthesis and incorporation of dithymidine and thymidine-cytidine dimer blocks connected with a five-atom amide linker N3′-CO-CH₂-S-CH₂ into oligonucleotides (ONs) are reported. The UV-T_m

Full text of DOI:10.1021/ol070990u

ORGANIC
LETTERS
2007
Vol. 9, No. 14
2697-2700
Synthesis and RNA Binding Selectivity
of Oligonucleotides Modified with
Five-Atom Thioacetamido Nucleic Acid
Backbone Structures
Khirud Gogoi, Anita D. Gunjal, Usha D. Phalgune, and Vaijayanti A. Kumar*
DiVision of Organic Chemistry, National Chemical Laboratory,
Dr. Homi Bhabha Road, Pune 411008, India
Va.kumar@ncl.res.in
Received April 27, 2007
ABSTRACT
Convenient chemical synthesis and incorporation of dithymidine and thymidine−cytidine dimer blocks connected with a five-atom amide
linker N3′−CO CH2 CH2 into oligonucleotides (ONs) are reported. The UV Tm experiments for binding affinities of these mixed backbone
−
−S−
−
ONs with complementary DNA and RNA sequences revealed important results such as significantly higher RNA-binding selectivity as compared
with complementary DNA. NMR studies of the dimer blocks suggested a marginal increase in the N-type sugar conformations over that of the
native DNA.
The last two decades have witnessed an upsurge in the
synthesis of several modified nucleic acid derivatives. The
intentions have been to synthesize therapeutically suitable
and commercially viable nucleic acid analogues.1 The more
recent developments such as splice correcting and exon
skipping strategies require highly robust nucleic acid ana-
logues2 that are stable under physiological conditions as
single strands as well as in the form of duplexes with
complementary RNA sequences. The enzymatic stability3 of
peptide nucleic acids (PNAs) gains importance for such
applications.2 The uncharged PNAs (Figure 1) are poorly
soluble in water and bind to DNA in either parallel or
antiparallel orientation.4 Also, PNAs require assistance in
the form of covalent conjugation with cell-penetrating
peptides (cpp)5 or additional positive charges on PNA6,7 to
cross the cell membranes for their activity. The replacement
of a few or several internucleoside phosphate linkages by
the robust amide bond as in PNA could be an alternate
solution so that the advantages of chirality and 3′-5′
(1) (a) Micklefield, J. Curr. Med. Chem. 2001, 8, 1157. (b) Kurreck, J.
Eur. J. Biochem. 2003, 270, 1628. (c) Turner, J. J.; Fabani, M.; Arzumanov,
A. A.; Ivanova, G.; Gait, M. J. Biochim. Biophys. Acta 2006, 1758, 290.
(d) Kumar, V. A.; Ganesh, K. N. Curr. Top. Med. Chem. 2007, 7, 715.
(2) (a) Dominski, Z.; Kole, R. Proc. Natl. Acad. Sci. U.S.A. 1993, 90,
8673. (b) Sazani, P.; Kole, R. J. Clin. InVest. 2003, 112, 481.
(3) (a) Egholm, M.; Buchardt, O.; Nielsen, P. E. J. Am. Chem. Soc. 1992,
114, 1895-1897. (b) Nielsen, P. E.; Egholm, M.; Buchardt, O. Bioconjugate
Chem. 1994, 5, 3.
(4) (b) Uhlmann, E.; Peyman, A.; Breipohl, G.; Will, D. W. Angew.
Chem., Int. Ed. 1998, 37, 2796.
(5) Bendifallah, N.; Rasmussen, F. W.; Zachar, V.; Ebbesen, P.; Nielsen,
P. E.; Koppelhus, U. Bioconjugate Chem. 2006, 17, 750.
(6) (a) Zhou, P.; Wang, M.; Du, L.; Fisher, G. W.; Waggoner, A.; Ly,
D. H. J. Am. Chem. Soc. 2003, 125, 6878. (b) Englund, E. A.; Appella, D.
H. Angew. Chem., Int. Ed. 2007, 46, 1414.
(7) Kumar, V. A.; Ganesh, K. N. Acc. Chem. Res. 2005, 38, 404.
10.1021/ol070990u CCC: $37.00
© 2007 American Chemical Society
Published on Web 06/07/2007

complementary RNA sequences significantly better than their
DNA counterparts, and the binding efficiency was found to
be as good as PNA itself. However, introduction of these
amino acid nucleoside derivatives in PNA sequences was
found to be detrimental for RNA/DNA binding. In this
communication, we present the synthesis and incorporation
of thymidine and thymidine-cytidine dimer blocks (tst and
cst) connected with a five-atom amide linker N3′-CO-
CH2-S-CH2 (TANA) (Figure 2) into oligonucleotide oli-
Figure 1. Structures of DNA, PNA, and TANA.
directionality of the sugar are maintained in the analogue
along with the enzymatic stability of the amide bond. The
retention of some negatively charged phosphate groups would
be advantageous for water solubility. Cell-penetration experi-
ments then could be performed using cationic lipids,8
complexation, or conjugation with cpp.8 Several four- and
five-atom amide-linked oligonucleotide analogues and those
containing thioether linkages are known in the literature.1,9
The five-atom amide linkers were introduced to compensate
the shorter amide bond in the dinucleoside as compared with
the four-atom phosphate linkers to maintain the internucleo-
side distance complementarity.9 The strategy was found to
be successful as some of the five-atom-linked dimers when
substituted in oligonucleotide (ON) sequences lead to
significantly higher RNA affinities compared with that of
the native DNA. It has been shown by CD spectroscopy and
crystal structure that the longer amide backbones do not
disrupt the duplex geometries.10 The application of these ONs
with superior binding properties could be stymied by the
multistep synthetic procedures to arrive at the desired
monomers. Our continuing efforts toward finding probable
optimum oligonucleotide mimics resulted in a new five-atom
thioacetamido nucleic acid (TANA, Figure 1) backbone.11
The convenient synthetic methodology to convert pyrimidine
nucleosides to a sugar-amino acid monomer unit by using
thiol functionality in ethyl mercaptoacetate as a requisite
nucleophile was reported in our previous communication.
The homooligomeric pyrimidine ONs were found to bind to
Figure 2. tst and cst building blocks.
gomers. The assessment of the compatibility of the TANA
dimer blocks in a sugar-phosphate backbone was studied
by UV-Tm measurements of the resulting mixed-backbone
ON complexes with DNA and RNA.
Thymidine was selectively tosylated at the 5′ position, and
the 3′-OH was then protected as a 3′-O-TBDMS group.
Treatment of 3′-O-protected 5′-O-tosyl-thymidine with ethyl
mercaptoacetate followed by ester hydrolysis gave the acid
synthon 4 (Scheme 1). The amine synthon of thymidine was
Scheme 1. Synthesis of a Thymidine Monomer Unit
(8) (a) Venkatesan, N.; Hyean Kim, B. Chem. ReV. 2006, 106, 3712. (b)
Turner, J. J.; Jones, S.; Fabani, M. M.; Ivanova, G.; Arzumanov, A. A.;
Gait, M. J. Blood Cells, Mol. Dis. 2007, 38, 1.
(9) (a) De Mesmaeker, A.; Lesueur, C.; Btvikrre, M.-O.; Waldner, A.;
Fritsch, V.; Wolf, R. M. Angew. Chem., Int. Ed. 1996, 35, 2790. (b) Nina,
M.; Fonne-Pfister, R.; Beaudegnies, R.; Chekatt, H.; Jung, P. M. J.; Murphy-
Kessabi, F.; Mesmaeker, A.; Wendeborn, S. J. Am. Chem. Soc. 2005, 127,
6027. (c) Govindaraju, T.; Kumar, V. A. Chem. Commun. 2005, 495. (d)
Govindaraju, T.; Kumar, V. A. Tetrahedron 2006, 60, 2321. (e) Meng, B.;
Kawai, S. H.; Wang, D.; Just, G.; Giannaris, P. A.; Damha, M. Angew.
Chem., Int. Ed. Engl. 1993, 32, 729. (f) Damha, M. A.; Meng, B.; Kawai,
S. H.; Wang, D.; Yannopoulos, C. G.; Just, G. Nucleic Acids Res. 1995,
23, 3967.
(10) (a) Pallan, P. S.; von Matt, P.; Wilds, C. J.; Altmann, K.-H.; Egli,
M. Biochemistry 2006, 45, 8048. (b) Wilds, C. J.; Minasov, G.; von Matt,
P.; Altmann, K.-H.; Egli, M. Nucleosides, Nucleosides Nucleic Acids 2001,
20, 991.
synthesized from 3′-azidothymidine 5 by protection of the
free hydroxy group as 5′-O-DMTr in 6 and reduction of azide
to amine to get 7. The 5′-O-DMTr, 3′-azidothymidine 6 was
converted12 to a cytidine derivative by known procedures
via C4-triazolide (8) followed by amination (9) and ben-
zoylation (10), and reduction of the azide gave the desired
(11) Gogoi, K.; Gunjal, A. D.; Kumar, V. A. Chem. Commun. 2006,
2373.
2698
Org. Lett., Vol. 9, No. 14, 2007

3
3
13
constants J1′2′ and J1′2′′
.
The population of the 2′-endo
sugar conformations (% S) was calculated from the equation
used in this analysis.20
Scheme 2. Synthesis of a Cytidine Monomer Unit
A series of chimeric ONs (Tables 1 and 2) containing one
Table 1. Modified ON-Pyrimidine Sequences and UV-Tm of
Complexes with Complementary DNA and RNAa
ON sequences mass
(calcd/obsd)
ON:DNA
Tm (°C)
ON:RNA
Tm (°C)
1
2
5′ CGTTTTTTTTGC 33
33:34
40
20:34
23.7
21:34
nd
33:35
32.0 (-8.0)b
20:35
32.3 (+8.6)
21:35
50.0
5′ CG TTtstTTT TGC 20
(3606.5/3606.6)
5′ CGTT tst TT tst GC 21
(3596.6/5597.0)
3
4
5′ CG tst tst tst tst GC 22
(3576.8/3576.6)
22:34
nd
22:35
47.81
5
5′ TCT CTT TCT T 39
39:40
23.6
23:40
nd
39:41
25.4 (+0.8)
23:41
29.6
6
5′ TCT C tst TCT T 23
(2933.1/2933.1)
7
5′ TCT C tst TC tst 24
(2923.2/2924.0)
24:40
nd
24:41
33.8
8
5′ Tcst CTT TCTT 29
(2948.2/2948.1)
5′ T cst CTT T cst T 30
(2938.3/2938.7)
5′ Tcst cst T T cst T 31
(2928.4/2928.7)
29:40
19.5
30:40
nd
31:40
nd
29:41
26.2
30:41
33.8
31:41
39.7
cytidine derivative 11 (Scheme 2). The synthesis of the dimer
building blocks tst 14 and cst 17 was then achieved by using
peptide-coupling chemistry from the monomer units, depro-
tection, and phosphitylation (Scheme 3). All the new
9
10
a
DNA (34, 40) and RNA (35, 41) sequences 34 5′ GCAAAAAAAACG
3′, 40 5′AAG AAA GAG A3′, 35 r (5′ GCAAAAAAAACG 3′), and 41 r
(5′ AAG AAA GAG A 3′). b Figures in parentheses indicate the difference
in Tm between complexes with RNA and DNA.
Scheme 3. Synthesis of tst and cst Dimer Blocks
to four tst and cst blocks were synthesized by automated
solid-phase synthesis using the phosphoramidite approach
and an Applied Biosystems 3900 DNA Synthesizer. After
cleavage from the support, the oligomers were purified by
Table 2. Modified ON-Purine-Pyrimidine Sequences and
UV-Tm of Complexes with Complementary DNA and RNAa
ON sequence
ON:DNA
ON:RNA
(mass cald/obsd)
Tm (°C)
Tm (°C)
1
2
3
4
5
5′ CCT CTT ACC TCA GTT ACA
36
5′ CCT C tst ACC TCA G TT ACA
26 (5366.7/5366.8)
5′ CCT C tst ACC TCA G tst ACA
27 (5356.9/5356.9)
5′ TCA CTA GAT G
42
36:37
54.6
26:37
39.6
27:37
43.5
42:43
24.3
36:38
54.7 (+0.1)
26:38
47.5 (+7.9)
27:38
52.8 (+9.3)
42:44
24.7 (0.4)
32:44
29.6 (+13)
compounds were characterized by 1H NMR, 13C NMR, and
mass spectral analysis. The structures of the dimer blocks
18 and 19 (Figure 2) were extensively studied by using 2D
NOESY, COSY, and TOCSY 1H spectral analysis. The sugar
ring conformations in the dinucleosides were elucidated using
an empirical method of analysis of the vicinal coupling
5′ TCA cst A GAT G 3′
32 (3036.2/3037.0)
32:43
16.3
a DNA (37, 43) and RNA (38, 44) sequences 37 5′ TGT AAC TGA
GGT AAG AGG 3′, 43 5′CAT CTA GAG A3′, 38 r (5′ UGU AAC UGA
GGU AAG AGG 3′), and 44 r (5′CAU CUA GAG A3′). bFigures in
parentheses indicate the difference in Tm between complexes with RNA
and DNA.
(12) Divakar, K. J.; Reese, C. B. J. Chem. Soc., Perkin Trans. 1 1982,
1171.
Org. Lett., Vol. 9, No. 14, 2007
2699

gel filtration and reverse-phase HPLC. The purity of the
oligomers was checked by reverse-phase HPLC analysis on
a C18 column and was characterized by mass spectrometry.
These ONs were tested for their binding affinity to comple-
mentary DNA and RNA sequences in thermal denaturation-
UV measurement experiments and the data are summarized
in Tables 1 and 2. The unmodified sequence GCT8CG 33
was found to form complexes with both cDNA and RNA
with a higher melting temperature for the ON:DNA complex
over ON:RNA (∆Tm ) -8).14 Introduction of a single
thioacetamido-linked tst dimer unit in 20 reversed this
selectivity, and the ON:RNA duplex, 20:35, was more stable
(∆Tm ) +8.6) than its complex with cDNA (20:34). A
cumulative effect was observed in stabilizing the ON:RNA
complex, 21:35, and at the same time, the complex with
DNA (21:34) was destabilized, when two units of modified
tst dimer were incorporated in the ON. Similarly, the ON
sequence 22 with alternating phosphate-TANA linkers
exhibited binding only with RNA 35 and no observable
melting transition with cDNA 34. The unmodified mixed
pyrimidine ON 39 formed complexes with either cDNA or
RNA and exhibited almost equal binding strength. The
preferential binding to RNA was consistently observed when
one or two phoshate linkers were replaced by TANA by
incorporation of either one or two modified (tst or cst) dimer
units, almost independent of their position in the ON.
Inclusion of two or more modified units in ONs leads to the
significant stabilization of their complexes with complemen-
tary RNA, whereas complexes with the DNA counterpart
did not show a detectable transition (Table 1, entries 3, 4, 7,
9, and 10). To verify the usefulness of these modified units
in the mixed purine-pyrimidine sequence context, we
synthesized two different unmodified sequences 36 and 42
(Table 2). The 18mer sequence 36 was modified by
introduction of one tst unit in 26 and two tst units in 27.
The unmodified 18mer 36 recognized both cDNA 37 and
RNA 38 with equal affinity (∆Tm ) +0.1). One tst unit
caused destabilization of complexes with both DNA
(36:37) and RNA (36:37), but to a much lesser extent with
RNA than with DNA. The discrimination between RNA and
DNA recognition was observed with a ∆Tm of about 8 °C.
Introduction of two tst units in 27 increased this stabilization
of ON:DNA/RNA (27:37/27:38) complexes and RNA vs
DNA discrimination to 9.3 °C. In a 10mer ON 32, a single
cst unit caused stabilization of the ON:RNA complex
(32:44) compared to the control of the unmodified complex
(42:44) and destabilized the ON:cDNA complex (32:43,
Table 2, entries 4 and 5). The binding was found to be
sequence specific as a single mismatch in the target RNA
highly destabilized the modified DNA:RNA complexes.20
Thus, the modified units consistently destabilized the com-
plex formation of the modified ONs with cDNA and
stabilized the ON:RNA complexes.
A single modified unit of LNA with a locked N-type sugar
conformation in an ON is known to effectively stabilize
duplex structures with both DNA and RNA.15 The electro-
negativity of the 3′-substituents in 3′-N-phoramidates was
shown to set the sugars in a preferred N-type conformation
but, unlike LNA, to show preferential binding to RNA over
DNA.14,16 Several other examples in the literature such as
2′-5′ DNA17/RNA18 prefer to bind to RNA over DNA, and
it is not therefore entirely certain which factors differentiate
the DNA vs RNA selectivity.1d Native DNA and RNA prefer
to be in S- or N-type sugar conformations giving rise to either
B- or A-form structures in equilibrium. In this particular case,
1
however, the H NMR studies point out conformational
equilibration in either the 3′-amino or 5′-thioacetamido sugars
to be similar to the native 3′-5′ phosphate linked DNA.20
The structural similarity between unmodified and modified
RNA:DNA complexes was also evident by CD studies.20 The
RNA selectivity of binding seems to be arising from the
extended backbone linker that is probably inherently folded
to be competent to bind to RNA over DNA as was found
with the reported five-atom-linked ON analogues.10,19 The
tst and cst dimer blocks were found to be compatible in the
DNA backbone to selectively stabilize the ON:RNA com-
plexes. Further work to exploit their utility is currently in
progress in our laboratory. The preferential sequence-
independent RNA binding ability of these evolved modified
ONs will find applications in current antisense research.
Acknowledgment. K.G. thanks CSIR, New Delhi, for a
senior research fellowship, and V.A.K. thanks DST, New
Delhi, for a research grant. We acknowledge a generous gift
of phosphitylating reagent from Innovassynth Technologies
(I) Ltd., Khopoli, India.
Supporting Information Available: Experimental and
1
spectral data of the compounds in Schemes 1-3. H NMR
of dinucleosides. Mass and 31P spectra of 14 and 17. Mass
spectra and CD and UV-Tm plots of the synthetic ONs with
complementary DNA/RNA/mismatched RNA. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL070990U
(15) Petersen, M.; Wengel, J. Trends Biotechnol. 2003, 21, 74.
(16) Gryaznow, S.; Chen, J.-K. J. Am. Chem. Soc. 1994, 116, 3143.
(17) Prakash, T. P.; Kraynack, B.; Baker, B. F.; Swayze, E. E.; Bhat, B.
Bioorg. Med. Chem. Lett. 2006, 16, 3238.
(18) (a) Giannaris, P. A.; Damha, M. J. Nucleic Acids Res. 1993, 21,
4742. (b) Wasner, M.; Arion, D.; Borkow, G.; Noronha, A.; Uddin, A. H.;
Parnaik, M. A.; Damha, M. J. Biochemistry 1998, 37, 7478.
(19) Alternatively, in duplexes, the extra distance between sugars might
be compensated by compact N-type sugar conformations that favor RNA
binding (we thank an annonymous referee for this useful suggestion).
(20) See Supporting Information.
(13) Rinkel, L. J.; Altona, C. J. Biomol. Struct. Dyn. 1987, 4, 621.
(14) Nawrot, B.; Boczkowska, M.; Wojcik, M.; Sochaki, M.; Kazmierski,
S.; Stec, W. J. Nucleic Acids Res. 1998, 26, 2650.
2700
Org. Lett., Vol. 9, No. 14, 2007