2. fraction boiling at 90°C contained the pure product. Yield was
5.8 g (40.7%). The IR spectrum had peaks at 1965, 1740, 1230,
and 1145 cmϪ1
. 1
H NMR (CDCl3) 2.37, S, CH3. This singlet in
isothiocyanate moved upfield compared with CH3COCl (2.66
p). 13
C NMR (CDCl3) 165.3, CO; 146.9, NCS, and 27.7, CH3.
HRCI. m͞z 101.12 calculated for C3H3NOS (M ϩ H)ϩ
102.1361.
N-Acetyl-N-(3-deoxythymidine-3-yl)-5-O-(4-methoxy-
phenyl)-diphenyl methyl thiourea (Compound 4). The solution
of 5Ј-O-MMTr-3Ј-amino-3Ј-deoxythymidine 2 (1.02 g, 2 mmol)
in anhydrous dicholoromethane (15 ml) was cooled in ice bath.
To this solution, acetylisothiocyanate 3 (0.185 ml, 2.2 mmol)
was added slowly. The reaction mixture was allowed to attain
RT and stirred for 3 hr. Reaction was monitored by TLC,
evaporated to dryness, and then purified by column chroma-
tography by using CH2Cl2:MeOH solvent system. Yield was
0.760 g (62%). TLC (CH2Cl2:MeOH, 95:5) Rf ϭ 0.35; IR (KBr)
spectrum had peaks at 3330, 3200, 2106, 1678, 1473, 1277, and
1085 cmϪ1
. 1
H NMR (400 MHz DMSO-d6): ␦ (ppm) 1.5 (s, 3H,
CH3 Thiourea), 2.0 (s, 3H, CH3 Thym), 2.25–2.35 (m, 2H, 2Ј
and 2ЈЈ H2), 3.2–3.4 (m, 2H, 5Ј and 5ЈЈ H2), 3.8 (s, 3
H, OOCH3
MMTr), 4.1 (m, 1H, 4ЈH), 5.1 (m, 1H, 3ЈH), 6.2 (t, 1H, 1ЈH),
6.9 (d, 2H, MMTr), 7.2–7.4 (m, 12H, MMTr), 7.6 (s, 1H, 6-H),
10.8 (d, 1H, exch NH Thiourea), 11.28 (s, 1H, exch NH
Thiourea), and 11.34 (s, 1H, exch NH Thym). (FAB) m͞z
614.2119, calculated for C33H34N4O6S (M ϩ H)ϩ
615.2277.
N-Acetyl-N-(3-deoxythymidine-3-yl)-N-(5-deoxythymi-
dine-5-yl)guanidine (Compound 7). To solution of 4 (0.400 g,
0.65 mmol) and 5Ј-amino-5Ј-deoxythymidine (6) (0.156 g, 0.65
mmol) in anhydrous dimethylformamide (DMF; 5 ml) was
added anhydrous triethylamine (0.22 ml, 1.62 mmol) and
HgCl2 (0.211 gm, 0.78 mmol). Reaction mixture was allowed to
stir overnight, when black precipitate (ppt) of Hg2S separates
out. Reaction was monitored by TLC and filtered over celite,
and filtrate was evaporated to dryness under reduced pressure
resulting in a oily residue. This oil was then purified by silica
gel column chromatography by using CH2Cl2:MeOH solvent
system. Pure compound elutes at 10% MeOH. Yield was 0.365
g (68%). TLC (CH2Cl2:MeOH, 9:1) Rf ϭ 0.4; 1
H NMR (400
MHz DMSO-d6): ␦ (ppm) 1.4 (s, 3H, CH3 Guan), 1.7 (s, 3H,
CH3 Thym), 1.8 (s, 3H, CH3 Thym), 2–2.4 (m, 4H, 2 ϫ 2Ј-H2),
3.2–3.4 (m, 4H, 2 ϫ 5ЈH2), 3.75 (s, 3
H, OOOCH3 MMTr), 3.8
(m, 1H, 4ЈH), 3.95 (m, 1H, 4ЈH), 4.15 (m, 1H, 3ЈH), 4.8 (m, 1H,
3ЈH), 5.4 (m, 1H, exch, 3Ј-OH), 6.2 (t, 1H, 1ЈH), 6.4(t, 1H,
1ЈH), 6.9 (d, 2H, MMTr), 7.2–7.5 (m, 12H, MMTr), 7.6 (s, 1H,
6-H), 10.3 (s, 1H, exch, NH Thym), 11.4 (2s, 2H, NH Guan).
High-resolution mass spectrometry (FAB) m͞z 822.3384, cal-
culated for C43H48N7O10 (M ϩ H)ϩ
822.3479.
Preparation of Amidite Building Block (Compound 1). To a
solution of 7 (0.6 g, 0.73 mmol) in anhydrous dichloromethane
(10 ml), diisopropylethylamine (0.5 ml, 2.92 mmol) and [Chlo-
ro(diisopropylamino)--cyanoethoxyphosphine] was added.
Reaction mixture was kept stirring at RT for 3 hr. Reaction was
monitored by TLC and evaporated to dryness under vacuum.
This amidite was coevaporated several times over anhydrous
acetonitrile and was used without further purification. 31
P
NMR (DMSO-d6): ␦ (ppm), 147.7 and 148.5. (FAB) m͞z
1022.10, calculated for C52H64N9O11P (M ϩ H)ϩ
1021.4462.
Oligonucleotide Synthesis and Purification. RNA oli-
gomers were obtained from the Integrated DNA Technologies
(Coralville, IA). All of the oligodeoxynucleotides were syn-
thesized on 1.3 mol scale on a Pharmacia GA Plus DNA
synthesizer by using CPG support and base protected 5Ј-O-(4,
4Ј-dimethoxytrityl)deoxyribonucleoside-3[-O-(diisopro-
pylamino)--cyanoethylphosphoramidite] monomers and
phosphoramidite dimer 1. Standard synthesis cycle with an
extended coupling time (15 min) was used during coupling of
modified phosphoramidite dimer 1; coupling efficiency of
Ͼ95% was observed for this step. The final trityl was kept on
for purification purpose, and oligonucleotides were depro-
tected with NH4OH at 60°C. Oligonucleotides containing
guanidinium linkages were subjected to longer NH4OH treat-
ment (48 hr, 60°C) so that guanidinium deprotection is com-
plete. All oligonucleotides were purified by RP-HPLC on
preparative Alltech C-8 RP column. Solvent system used was
solvent A: 0.1 M triethylammoniumacetate, pH 7.0 and solvent
B: CH3CN. The gradient used was 0–2 min, 15% B; 2–12 min,
25% B, and 12–45 min, 25% B at a flow rate of 4 ml͞min. The
HPLC-purified oligonucleotides were then detritylated by
using 1 ml of 80% acetic acid (1 hr, RT), evaporated to dryness
under reduced pressure, redissolved in double distilled water,
and further purified by size exclusion chromatography.
Thermal Denaturation Studies. The concentrations of nu-
cleotide solutions were determined by using the extinction
coefficients (per mol of nucleotide) calculated according to
nearest neighbor approximation for the DNA. (24). 260 MϪ1
cmϪ1
; 9–12, 40880; 13, 50750; 15, 50751, and 16, 50750. All
experiments were conducted in 10 mM Na2HPO4 buffer pH
7.1, and ionic strength was adjusted with NaCl (0, 10, and 100
mM). The final concentration of oligonucleotide strands was 3
M. The solutions were heated to 95°C for 5 min and allowed
to cool to RT slowly before being stored at 4°C overnight.
Spectrophotometric measurements were performed at 260 nm
on a Cary 1E UV͞vis spectrophotometer equipped with
temperature-programming and a thermal-melting software
package, using 1-cm path length quartz cuvettes at a heating
rate of 0.5°C͞min over the range of 10–80°C. Dry nitrogen gas
was flushed in the spectrophotometer chamber to prevent
moisture condensation at temperatures below 15°C. Melting
temperatures were taken as the temperature of half dissocia-
FIG. 1. Structure of internucleoside phosphodiester (DNA) and
guanidium linkage (DNG͞DNA). ء indicates guanidinium linkage in
DNG͞DNA chimeras 9–11.
11048 Chemistry, Biochemistry: Barawkaw and Bruice Proc. Natl. Acad. Sci. USA 95 (1998)
3. tion and were obtained from first derivative plots. The Tm
values are accurate to Ϯ 0.5°C over reported values.
Stoichiometry of Binding. The stoichiometry of binding was
determined by the method of continuous variation (25). So-
lutions containing the different molar ratios of DNG͞DNA 10
and complementary DNA 13 were heated to 90°C and allowed
to cool slowly to 15°C. The total concentration of duplex was
always 3 m. The pH was maintained at 7.1 with 10 mM
Na2HPO4 buffer while the ionic strength was held constant at
10 mM with NaCl. The A at 260 nm of each solution was
measured by Cary 1E UV͞vis spectrophotometer.
Base Composition Analysis. The base composition analysis
of modified oligonucleotides were confirmed by enzymatic
hydrolysis (26). Oligonucleotides 9–12 (0.2 A254 unit) were
dissolved in 20 mM Tris⅐HCl (200 l, pH 8.9) and treated with
snake venom phosphodiesterase (0.02 unit) and alkaline phos-
phatase (0.8 unit) at 37°C for 12 hr. This hydrolyzate (30 l)
was analyzed on analytical C-8 RP-HPLC column and eluted
by using 0.5%͞min gradient of CH3CN in 0.1 M TEAA, pH 7.0
for 50 min at a flow rate of 1 ml͞min. The retention times for
(1) standard nucleosides were dC, 4.59 min; T, 7.12 min; dG,
7.77 min; dA, 11.47 min, and for dinucleoside containing
guanidinium linkage 8, 14.27 min (2) for enzymatic hydroly-
zate of 9 were dC, 4.58 min; T, 7.13 min; dG, 7.77 min; dA,
11.69 min, and TgT 8, 14.13 min. The corresponding peak areas
were used to calculate the base composition of oligonucleo-
tides.
Exonuclease I Digestion Experiment. Oligonucleotides 9–12
(0.2 OD) with phosphodiester or guanidinium linkages were
treated with exonuclease I (500 units) in 10 mM Tris⅐HCl, pH
9.0 (0.2 ml) containing 50 mM KCl. Reaction mixtures were
analyzed on C-8 RP-HPLC column by using 0.5%͞min gradi-
ent of CH3CN in 0.1 M tetraethylammonium acetate, pH 7.0,
for 50 min at a flow rate of 1 ml͞min. Time points were 0, 1,
2, 6, and 12 hr for DNG͞DNA oligonucleotides (9-11) and only
0 and 1 hr for phosphodiester oligomer 12 because this was
completely digested by that time.
RESULTS AND DISCUSSION
Synthesis. For incorporation of guanidinium internucleo-
side linkages into ODN, we synthesized the phosphoramidite
1 (Scheme 1). The guanidinium linkage of 1 remained pro-
tected during ODN synthesis and can be deprotected at the end
of the synthesis to give a positively charged guanidinium
linkage. There are reports of the synthesis of N-substituted
guanidinium internucleoside linkages that are neutral and
cannot be deblocked to give charged guanidinium (27, 28). The
synthesis of 1 involves coupling of the 5Ј-amino group of
5Ј-amino-5Ј-deoxythymidine 6 with in situ-generated carbodi-
imide 5, obtained from reaction of the acetyl-protected thio-
urea 4 with mercury (II) in the presence of Triethylamine (29).
The acetyl protected thiourea 4 was synthesized by using
5Ј-O-monomethoxytrityl-3Ј-amino-3Ј-deoxythymidine 2 and
acetylisothiocyanate 3, in dichloromethane. The electron with-
drawing nature of acetyl group on thiourea of 4 activates the
carbodiimide intermediate 5, facilitating the attack by the
5Ј-amine, and then this acetyl group acts as a protecting group
on the resulting guanidinium linkage of dinucleoside 7.
The acetyl protection of internucleoside guanidinium re-
mains stable to conditions required for DNA solid-phase
synthesis, and the acetyl is removed during usual final depro-
tection conditions of DNA synthesis. To confirm deprotection
of guanidinium linkage, 7 was treated with 35% ammonium
hydroxide at 55°C for 45 hr followed by detritylation to provide
the deprotected guanidinium dinucleoside 8, in quantitative
yield (as determined by HPLC and characterized by FAB MS
[(M ϩ H)ϩ
ϭ 508]). The phosphotylation of 7, using [chloro-
(diisopropylamino)--cyanoethoxyphosphine] provided the fi-
nal phosphoramidite 1, which exhibited the characteristic 31
P
NMR signals at 147.7 and 148.5 ppm. Thus, the ability to
protect the guanidinium linkage with an acetyl function and
remove this function at completion of synthesis of DNG͞DNA
chimera makes it suitable for standard automated solid phase
synthesis.
The guanidinium linkage was incorporated into ODNs 9–11
(Fig. 1) by using Pharmacia GA Plus DNA synthesizer. The
‘‘standard synthesis cycle’’ with an extended time (15 min) was
used during the coupling of phosphoramidite 1. After com-
pletion of the synthesis, ODNs 9–11 were purified by RP-
HPLC and then detritylated and again purified by size exclu-
sion. Electrospray mass spectroscopic analysis for ODN 5Ј-
T*TAGGGT*TA-3Ј (C92H113N39O46P6, 2686.9) indicated the
expected masses for the doubly charged C92H113N39O46P6 (M
ϩ 2H)2ϩ
: 1343.4 and triply charged C92H113N39O46P6 (M ϩ
3H)3ϩ
: 895.4 species confirming the presence of modified
moiety into purified DNG͞DNA chimera.
To further insure that the guanidinium-linked nucleobases
have survived the synthetic chemistry of solid phase synthesis
by phosphoramidite approach, enzymatic hydrolysis of 9–11
were carried out by using snake venom phosphodiesterase
followed by alkaline phosphatase (26). The enzyme digest was
analyzed by RP-HPLC, which indicated the presence of
dinucleoside 8, with guanidinium internucleoside linkage. The
corresponding peak areas showed the correct base composi-
tion of ODNs 9–11.
Duplex Formation by DNG͞DNA Chimeras. The DNG͞
DNA chimeric ODNs 9–11 are 18-mers (Fig. 1) with either
three (at 5Ј, 3Ј, and center), two (at 5Ј and 3Ј) or one (at center)
guanidinium linkages. The duplexes were formed between
DNG͞DNA chimeric ODNs 9–11 with complementary DNA
13 or RNA 15. The 18-mer ODNs containing all four nucleo-
bases are non self-complementary and form only antiparallel
duplexes. Before performing thermal denaturation experi-
ments the stoichiometry of binding of DNG͞DNA chimeric
ODN and DNA was determined by the method of continuous
variation (25) to generate mixing curves of the absorbance vs.
mol fraction of 10 and 13 (Fig. 2). Increasing mol fraction of
13 to the 10 (in 10 mM Na2HPO4, pH 7.1 at 15°C) lowered the
UV A at 260 nm an inflection point at 0.5 mol fraction
indicated the formation of expected 1:1 stoichiometry for
DNG͞DNA:DNA duplex (10:13).
Tm values for unmodified DNA:DNA (12:13) and
DNA:RNA (12:15) duplexes were determined. These Tm
values serve for comparative purposes with the Tm values of
DNG͞DNA chimeras (Table 1). As expected, the Tm was
observed to increase with increase in salt concentration with
DNA:DNA and DNA:RNA. The duplexes of DNG͞DNA
chimera with complementary DNA or RNA showed increase
in A at 260 nm upon increasing temperature and showed the
characteristic sigmoidal melting pattern (Fig. 3). Incorporation
of three or one guanidinium linkages (ODN 9 and 11) has no
effect on hybridization properties with complementary DNA
13, when there is no salt (cf. Exps. 3 and 5 with 1). As salt
concentration was increased (0–100 mM NaCl) stability of
duplexes 9:13 and 11:13 was found to decrease when compared
with DNA:DNA (cf. Exps. 3 and 5 with 1). The duplex of
DNG͞DNA chimeras 9 and 11 with complementary RNA 15
showed destabilization at all salt concentrations (cf. Exps. 6
and 8 with 2). When there are two guanidinium linkages, as in
the duplex of ODN 10 with complementary DNA 13, and in
absence of salt, there is 2°C and 2.8°C increase in Tm compared
with DNA:DNA (12:13) and DNA:RNA (12:15) hybrid, re-
spectively (cf. Exp. 4 with 1 and 2). In absence of salt, the
duplex of DNG͞DNA chimera 10 with complementary RNA
15 exhibits similar stability as does the DNA:RNA duplex
(compare Exp. 7 with 2). With increasing salt concentration
the duplexes of 10 with complementary DNA as well as RNA
exhibit a decrease in Tm (cf. Exp. 4 with 1 and Exp. 7 with 2).
Chemistry, Biochemistry: Barawkaw and Bruice Proc. Natl. Acad. Sci. USA 95 (1998) 11049
4. Our thermal denaturation results show that incorporation of
guanidinium internucleoside linkages into ODNs, either sep-
arated by a few nucleotide units (as in 9) or at the center
position (11), in absence of salt, has no effect on duplex
stability. On the other hand, incorporating guanidinium link-
ages at 5Ј and 3Ј ends (10) shows stabilization of duplex with
complementary DNA as well as RNA (with no salt). This
shows the importance of the position of guanidinium linkages
11050 Chemistry, Biochemistry: Barawkaw and Bruice Proc. Natl. Acad. Sci. USA 95 (1998)
5. in ODN. As expected, increasing salt concentration (0–100
mM NaCl) destabilizes duplexes consisting of DNG͞DNA
chimeras 9–11, with complementary DNA as well as RNA (cf.
Exps. 3–5 with 1 and Exps. 6–8 with 2). This salt effect is
opposite to that seen with DNA:DNA (12:13) and DNA:RNA
(12:15) duplexes, where increasing salt concentration provides
stability by masking the electrostatic opposition of negative
charges. Decreasing salt concentration allows the positively
charged guanidinium of DNG͞DNA chimera to become inti-
mately salt paired with negatively charged phosphodiester of
complementary DNA or RNA. Thus, duplexes of DNG͞DNA
with DNA and RNA are stabilized by lowering of the ionic
strength.
Sequence Specificity. To study the sequence specificity of
binding of DNG͞DNA chimera with complementary DNA or
RNA, DNG͞DNA chimeric ODN was allowed to form duplex
with complementary DNA 14 or RNA 16 containing one base
mismatch (T͞U instead of dA͞A) at the center and then the
stability of the duplex was monitored by thermal denaturation.
The duplexes 12:14 (DNA:DNA) and 12:16 (DNA:RNA)
exhibit (Table 1) 9.1°C and 4.2°C decrease in Tm, respectively,
in comparison to fully complementary 12:13 and 12:15 du-
plexes (cf. Exp. 9 with 1 and 10 with 2). Approximately the
same mismatch discrimination was observed for DNG͞
DNA:DNA hybrid (⌬Tm Ϫ10 to Ϫ11°C, cf. Exps. 11–13 with
3–5) whereas for DNG͞DNA:RNA hybrid base mismatch,
discrimination was almost double (⌬Tm Ϫ9 to Ϫ11°C, cf. Exps.
14–16 with 6–8). This clearly demonstrates that binding of
DNG͞DNA chimera with complementary DNA and RNA is
sequence specific.
Stability of DNG͞DNA Toward Exonuclease. ODNs capped
with guanidinium internucleoside linkages are expected to be
resistant to the cleavage by exonucleases. To investigate this
DNG͞DNA, ODNs 9–11 were subjected to nucleolytic diges-
tion by exonuclease I. The hydrolyzate was then analyzed by
RP-HPLC. The control ODN 12 was found to be completely
hydrolyzed after 1 hr of incubation. The DNG͞DNA chimera
9 and 10 were found to be absolutely stable toward exonuclease
1 digestion even after 12 hr of incubation. The DNG͞DNA
chimera 11, which contains only one guanidinium linkage at
the center of ODN, was found to be partially hydrolyzed after
1 hr (at 0 hr ODN 11, on RP-HPLC gave Rt 32.8 min, and after
1 hr Rt 29.7 min) and there was no further hydrolysis even after
12 hr (Rt 29.7 min). This clearly shows that DNG͞DNA ODNs
9 and 10 having guanidinium linkages at 5Ј and 3Ј ends are
completely stable to exonuclease 1. The partial hydrolysis of
ODN 11, having guanidinium at the center indicates that
phosphodiester linkages around guanidinium are stable to
exonuclease cleavage. Further investigations are in progress.
Conclusions. In this paper, we successfully have demon-
strated the insertion of cationic internucleoside guanidinium
linkage in place of negative phosphodiester linkages in DNA.
For this purpose, we have used standard phosphoramidite
chemistry and automated solid phase synthesis. The DNG͞
DNA chimera 10, with two terminal positive charges, showed
enhanced binding to complementary DNA and RNA in ab-
sence of salt. The binding of 10, with complementary DNA at
10 mM NaCl (close to physiological conditions), was similar to
unmodified DNA:DNA. The binding of DNG͞DNA chimeras
is highly sequence specific. The 5Ј and 3Ј capping of the DNA
with gaunidinium linkage provides protection to hydrolysis by
exonuclease 1. The lower net negative charge of DNG͞DNA
chimera, arising from insertion of positive guanidinium link-
ages, may assist the cellular uptake of these ODNs.
This work was initiated under support by the Office of Naval
Research (N00014-96-1-01232). We gratefully acknowledge continu-
ing support by the National Institutes of Health (3 R37 DK09171–
3451).
1. Uhlmann, E. & Peyman, A. (1990) Chem. Rev. (Washington,
D.C.) 90, 543–584.
2. Cook, P. D. (1991) Anti-Cancer Drug Des. 6, 585–607.
3. De Mesmaeker, A., Haner. R, Martin, P. & Moser, H. E. (1995)
Acc. Chem. Res. 28, 366–374.
4. Sanghvi, Y. S. & Cook, P. D. (1993) in Nucleosides and Nucle-
otides as Antitumor and Antiviral Agents, eds. Chu, C. K. & Baker,
D. C. (Plenum, New York), p. 311.
FIG. 2. Job plot (continuous variation method) of 10 with 13 at 3
m total concentration, at ϭ 260 nm and 15°C. FIG. 3. Plots of A260 vs. t°C., for (E) duplex 12:13 and (‚) 10:13 in
buffer 10 mM Na2HPO4, pH 7.1 with differant NaCl concentration.
(0–100 mM).
Table 1. Duplex melting* temperatures (Tm°C)
Exp no. Duplex No NaCl 10 mM NaCl 100 mM NaCl
1 12:13 34.8 48.6 58.5
2 12:15 34.0 49.7 59.9
3 9:13 34.8 46.6 53.5
4 10:13 36.8 48.6 57.5
5 11:13 34.8 47.6 56.5
6 9:15 30.1 41.8 50.9
7 10:15 34.2 45.9 57.1
8 11:15 31.2 43.9 54.0
9 12:14 — 39.5 —
10 12:16 — 45.5 —
11 9:14 — 35.5 —
12 10:14 — 38.5 —
13 11:14 — 37.6 —
14 9:16 — 30.5 —
15 10:16 — 35.5 —
16 11:16 — 34.6 —
*Buffer used: 10 mM Na2HPO4, pH 7.1 with differant NaCl concen-
tration (0–100 mM).
Chemistry, Biochemistry: Barawkaw and Bruice Proc. Natl. Acad. Sci. USA 95 (1998) 11051
6. 5. Milligan, J. F., Matteucci, M. D. & Martin, C. J. (1993) J. Med.
Chem. 36, 1923–1937.
6. Gryaznov, S. (1997) Nucleosides Nucleotides 16, 899–905.
7. Jones, R. J., Lin, K.-Y., Milligan, J. F., Wadwani, S. & Matteucci,
M. D. (1993) J. Org. Chem. 58, 2983–2991.
8. Rice, J. S. & Gao, X. (1997) Biochemistry 36, 399–411.
9. Sanghvi, Y. S., Swayze, E. E., Peoc’h, D., Bhat, B. & Dimock, S.
(1997) Nucleosides Nucleotides 16, 907–916.
10. De Mesmaeker, A., Lesueur, C., Bevierre, M., Waldner, A.,
Fritsch, V. & Wolf, R. M. (1996) Angew. Chem. Int. Ed. Engl. 35,
2790–2794.
11. Kean, J. M., Kipp, S. A., Miller, P. S., Kulka, M. & Aurelian, L.
(1995) Biochemistry 34, 14617–14620.
12. Nielsen, P., Egholm, M., Berg, R. H. & Buchardt, O. (1991)
Science 254, 1497–1500.
13. Letsinger, R. L., Singman, C. N., Histand, G. & Salunkhe, M.
(1988) J. Am. Chem. Soc. 110, 4470–4471.
14. Fathi, R., Huang, Q., Coppola, G., Delaney, W., Teasdale, R.,
Krieg, A. M. & Cook, A. F. (1994) Nucleic Acids Res. 22,
5416–5424.
15. Dempcy, R. O., Browne, K, A. & Bruice, T. C. (1995) J. Am.
Chem. Soc. 117, 6140–6141.
16. Dempcy, R. O., Luo, J. & Bruice, T. C. (1996) Proc. Natl. Acad.
Sci. USA 93, 4326–4330.
17. Linkletter, B. & Bruice, T. C. (1998) Bioorg. Med. Chem. Lett. 8,
1285–1290.
18. Barawkar, D. A., Rajeev, K. G., Kumar, V. A. & Ganesh, K. N.
(1996) Nucleic Acids Res. 24, 1229–1237.
19. Hashimoto, H., Nelson, M. G. & Switzer, C. J. (1993) J. Am.
Chem. Soc. 115, 7128–7134.
20. Schmid, N. & Behr, J. P. (1995) Tetrahedron Lett. 36, 1447–1450.
21. Blasko, A., Dempcy, R. O., Minyat, E. E. & Bruice, T. C. (1996)
J. Am. Chem. Soc. 118, 7892–7899.
22. Blasko, A., Minyat, E. E., Dempcy, R. O. & Bruice, T. C. (1997)
Biochemistry 36, 7821–7831.
23. Barawkar, D. A., Linkletter, B. & Bruice, T. C. (1998) Bioorg.
Med. Chem. Lett. 8, 1517–1520.
24. Gray, D. M., Hung, S.-H. & Johnson, K. H. (1995) in Methods in
Enzymology (Academic, New York), Vol. 246, pp. 19.
25. Job, P. (1928) Ann. Chim. (Rome) 9, 113–203.
26. Connoly, B. A. (1992) in Methods in Enzymology (Academic, New
York), pp. 36–53.
27. Vandendriessche, F., Aerschot, A. V., Voortmans, M., Janssen, G.,
Busson, R., Overbeke, A. V., Bossche, W. V., Hoogmartens, J. &
Herdewijn, P. (1993) J. Chem. Soc. Perkin Trans. 1, 1567–1575.
28. Pannecouque, C., Wigerinck, P., Areschot, V. A. & Herdewijn,
P. (1992) Tetrahedron Lett. 33, 7609–7612.
29. Robinson, S. & Roskamp, E. J. (1997) Tetrahedron 53, 6697–
6705.
11052 Chemistry, Biochemistry: Barawkaw and Bruice Proc. Natl. Acad. Sci. USA 95 (1998)
![fraction boiling at 90°C contained the pure product. Yield was
5.8 g (40.7%). The IR spectrum had peaks at 1965, 1740, 1230,
and 1145 cmϪ1
. 1
H NMR (CDCl3) 2.37, S, CH3. This singlet in
isothiocyanate moved upfield compared with CH3COCl (2.66
p). 13
C NMR (CDCl3) 165.3, CO; 146.9, NCS, and 27.7, CH3.
HRCI. m͞z 101.12 calculated for C3H3NOS (M ϩ H)ϩ
102.1361.
N-Acetyl-N-(3-deoxythymidine-3-yl)-5-O-(4-methoxy-
phenyl)-diphenyl methyl thiourea (Compound 4). The solution
of 5Ј-O-MMTr-3Ј-amino-3Ј-deoxythymidine 2 (1.02 g, 2 mmol)
in anhydrous dicholoromethane (15 ml) was cooled in ice bath.
To this solution, acetylisothiocyanate 3 (0.185 ml, 2.2 mmol)
was added slowly. The reaction mixture was allowed to attain
RT and stirred for 3 hr. Reaction was monitored by TLC,
evaporated to dryness, and then purified by column chroma-
tography by using CH2Cl2:MeOH solvent system. Yield was
0.760 g (62%). TLC (CH2Cl2:MeOH, 95:5) Rf ϭ 0.35; IR (KBr)
spectrum had peaks at 3330, 3200, 2106, 1678, 1473, 1277, and
1085 cmϪ1
. 1
H NMR (400 MHz DMSO-d6): ␦ (ppm) 1.5 (s, 3H,
CH3 Thiourea), 2.0 (s, 3H, CH3 Thym), 2.25–2.35 (m, 2H, 2Ј
and 2ЈЈ H2), 3.2–3.4 (m, 2H, 5Ј and 5ЈЈ H2), 3.8 (s, 3
H, OOCH3
MMTr), 4.1 (m, 1H, 4ЈH), 5.1 (m, 1H, 3ЈH), 6.2 (t, 1H, 1ЈH),
6.9 (d, 2H, MMTr), 7.2–7.4 (m, 12H, MMTr), 7.6 (s, 1H, 6-H),
10.8 (d, 1H, exch NH Thiourea), 11.28 (s, 1H, exch NH
Thiourea), and 11.34 (s, 1H, exch NH Thym). (FAB) m͞z
614.2119, calculated for C33H34N4O6S (M ϩ H)ϩ
615.2277.
N-Acetyl-N-(3-deoxythymidine-3-yl)-N-(5-deoxythymi-
dine-5-yl)guanidine (Compound 7). To solution of 4 (0.400 g,
0.65 mmol) and 5Ј-amino-5Ј-deoxythymidine (6) (0.156 g, 0.65
mmol) in anhydrous dimethylformamide (DMF; 5 ml) was
added anhydrous triethylamine (0.22 ml, 1.62 mmol) and
HgCl2 (0.211 gm, 0.78 mmol). Reaction mixture was allowed to
stir overnight, when black precipitate (ppt) of Hg2S separates
out. Reaction was monitored by TLC and filtered over celite,
and filtrate was evaporated to dryness under reduced pressure
resulting in a oily residue. This oil was then purified by silica
gel column chromatography by using CH2Cl2:MeOH solvent
system. Pure compound elutes at 10% MeOH. Yield was 0.365
g (68%). TLC (CH2Cl2:MeOH, 9:1) Rf ϭ 0.4; 1
H NMR (400
MHz DMSO-d6): ␦ (ppm) 1.4 (s, 3H, CH3 Guan), 1.7 (s, 3H,
CH3 Thym), 1.8 (s, 3H, CH3 Thym), 2–2.4 (m, 4H, 2 ϫ 2Ј-H2),
3.2–3.4 (m, 4H, 2 ϫ 5ЈH2), 3.75 (s, 3
H, OOOCH3 MMTr), 3.8
(m, 1H, 4ЈH), 3.95 (m, 1H, 4ЈH), 4.15 (m, 1H, 3ЈH), 4.8 (m, 1H,
3ЈH), 5.4 (m, 1H, exch, 3Ј-OH), 6.2 (t, 1H, 1ЈH), 6.4(t, 1H,
1ЈH), 6.9 (d, 2H, MMTr), 7.2–7.5 (m, 12H, MMTr), 7.6 (s, 1H,
6-H), 10.3 (s, 1H, exch, NH Thym), 11.4 (2s, 2H, NH Guan).
High-resolution mass spectrometry (FAB) m͞z 822.3384, cal-
culated for C43H48N7O10 (M ϩ H)ϩ
822.3479.
Preparation of Amidite Building Block (Compound 1). To a
solution of 7 (0.6 g, 0.73 mmol) in anhydrous dichloromethane
(10 ml), diisopropylethylamine (0.5 ml, 2.92 mmol) and [Chlo-
ro(diisopropylamino)--cyanoethoxyphosphine] was added.
Reaction mixture was kept stirring at RT for 3 hr. Reaction was
monitored by TLC and evaporated to dryness under vacuum.
This amidite was coevaporated several times over anhydrous
acetonitrile and was used without further purification. 31
P
NMR (DMSO-d6): ␦ (ppm), 147.7 and 148.5. (FAB) m͞z
1022.10, calculated for C52H64N9O11P (M ϩ H)ϩ
1021.4462.
Oligonucleotide Synthesis and Purification. RNA oli-
gomers were obtained from the Integrated DNA Technologies
(Coralville, IA). All of the oligodeoxynucleotides were syn-
thesized on 1.3 mol scale on a Pharmacia GA Plus DNA
synthesizer by using CPG support and base protected 5Ј-O-(4,
4Ј-dimethoxytrityl)deoxyribonucleoside-3[-O-(diisopro-
pylamino)--cyanoethylphosphoramidite] monomers and
phosphoramidite dimer 1. Standard synthesis cycle with an
extended coupling time (15 min) was used during coupling of
modified phosphoramidite dimer 1; coupling efficiency of
Ͼ95% was observed for this step. The final trityl was kept on
for purification purpose, and oligonucleotides were depro-
tected with NH4OH at 60°C. Oligonucleotides containing
guanidinium linkages were subjected to longer NH4OH treat-
ment (48 hr, 60°C) so that guanidinium deprotection is com-
plete. All oligonucleotides were purified by RP-HPLC on
preparative Alltech C-8 RP column. Solvent system used was
solvent A: 0.1 M triethylammoniumacetate, pH 7.0 and solvent
B: CH3CN. The gradient used was 0–2 min, 15% B; 2–12 min,
25% B, and 12–45 min, 25% B at a flow rate of 4 ml͞min. The
HPLC-purified oligonucleotides were then detritylated by
using 1 ml of 80% acetic acid (1 hr, RT), evaporated to dryness
under reduced pressure, redissolved in double distilled water,
and further purified by size exclusion chromatography.
Thermal Denaturation Studies. The concentrations of nu-
cleotide solutions were determined by using the extinction
coefficients (per mol of nucleotide) calculated according to
nearest neighbor approximation for the DNA. (24). 260 MϪ1
cmϪ1
; 9–12, 40880; 13, 50750; 15, 50751, and 16, 50750. All
experiments were conducted in 10 mM Na2HPO4 buffer pH
7.1, and ionic strength was adjusted with NaCl (0, 10, and 100
mM). The final concentration of oligonucleotide strands was 3
M. The solutions were heated to 95°C for 5 min and allowed
to cool to RT slowly before being stored at 4°C overnight.
Spectrophotometric measurements were performed at 260 nm
on a Cary 1E UV͞vis spectrophotometer equipped with
temperature-programming and a thermal-melting software
package, using 1-cm path length quartz cuvettes at a heating
rate of 0.5°C͞min over the range of 10–80°C. Dry nitrogen gas
was flushed in the spectrophotometer chamber to prevent
moisture condensation at temperatures below 15°C. Melting
temperatures were taken as the temperature of half dissocia-
FIG. 1. Structure of internucleoside phosphodiester (DNA) and
guanidium linkage (DNG͞DNA). ء indicates guanidinium linkage in
DNG͞DNA chimeras 9–11.
11048 Chemistry, Biochemistry: Barawkaw and Bruice Proc. Natl. Acad. Sci. USA 95 (1998)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)