Chemical synthesis and conformational properties of a new cyclouridylic acid having an ethylene bridge between the uracil 5-position and 5'-phosphate group

Full text of DOI:10.1021/jo951756x

1500
J . Org. Chem. 1996, 61, 1500-1504
Notes
Sch em e 1
Ch em ica l Syn th esis a n d Con for m a tion a l
P r op er ties of a New Cyclou r id ylic Acid
Ha vin g a n Eth ylen e Br id ge betw een th e
Ur a cil 5-P osition a n d 5′-P h osp h a te Gr ou p
Kohji Seio,† Takeshi Wada,† Kensaku Sakamoto,‡
Shigeyuki Yokoyama,‡ and Mitsuo Sekine*,†
Department of Life Science, Tokyo Institute of Technology,
Nagatsuta, Midoriku, Yokohama 226, J apan, and
Department of Biophysics and Biochemistry, The University
of Tokyo, Hongo, Bunkyoku, Tokyo 113, J apan
Received September 26, 1995
In tr od u ction
Intramolecularly cyclized nucleosides and nucleotides
have been utilized as useful analogues for studies of the
conformational property or the stereochemistry of nucleic
acids.1 Chemical fixation of the glycosyl or 5′-exo-
methylene torsion angle and the sugar conformation has
been achieved by introducing various cyclic structures
between the base and sugar moieties. However, most of
the cyclonucleosides reported in the literature have
conformations other than those of nucleotide units seen
usually in DNA and RNA2 because of their constrained
structures. Therefore, it is of great importance to ratio-
nally design and synthesize cyclic nucleosides having
fixed sugar conformations which resemble those of natu-
ral DNA and RNA.
point of view.7 By use of NMR spectroscopy and molec-
ular mechanics calculations, Sakamoto et al. investigated
the conformational property of 5-[(methylamino)methyl]-
uridine 5′-monophosphate (pmnm5U), which was also
found in the anticodon first letter of minor tRNAArg of E.
coli,8 and revealed that intrinsic intramolecular interac-
tions between the 5′-phosphate and the 5-[(methylamino)-
methyl] groups considerably stabilize the C3′-endo con-
formation.5 This result suggests the possibility that the
ribose moiety can be fixed in an N-type conformation by
introducing a bridge between the 5′-phosphate group and
the C-5 position of the uracil moiety.9
A number of pyrimidine derivatives having an N-type
sugar conformation stabilized by intramolecular inter-
actions3-6 have been discovered at the anticodon first
letter of tRNAs, and their essential roles in codon
recognition have been studied from a stereochemical
In this paper, we report the synthesis and conforma-
tional analysis of 1, which was designed as an analogue
of pmnm5U cyclized intramolecularly via an ethylene
bridge and expected to have a fixed N-type conformation.
† Tokyo Institute of Technology.
‡ The University of Tokyo.
Resu lts a n d Discu ssion
(1) (a) Maruyama, T.; Adachi, Y.; Honjo, M. J . Org. Chem. 1988,
53, 4552. (b) Yoshimura, Y.; Otter, B. A.; Ueda, T.; Matsuda, A. Chem.
Pharm. Bull. 1992, 40, 1761 and references cited therein. (c) For a
comprehensive review on cyclonucleosides, see: Mizuno, Y. The
Organic Chemistry of Nucleic Acid; Elsevier: Amsterdam, 1986; pp
113-133.
(2) (a) Yamagata, Y.; Tomita, K.; Usui, H.; Sano, T.; Ueda, T. Chem.
Pharm. Bull. 1989, 37, 1971. (b) Saenger, W. In Principles of Nucleic
Acid Structure; Springer Verlag: New York, 1984. (c) Yamagata, Y.;
Tomita, K.; Usui, H.; Matsuda, A.; Ueda, T. Chem. Pharm. Bull. 1992,
40, 6. (d) Bevierre, M. O.; Mesmaeker, A. D.; Wolf, R. M.; Freier, S. M.
Bioorg. Med. Chem. Lett. 1994, 2, 237.
(3) (a) Egert, E.; Lindner, H. J .; Hillen, W.; Bohm, M. C. J . Am.
Chem. Soc. 1980, 102, 3707. (b) Uhl, W.; Reiner, J .; Gassen, H. G.
Nucleic Acids Res. 1983, 11, 1167.
(4) Kawai, G.; Hashizume, T.; Yasuda, M.; Miyazawa, T.; McCloskey,
J . A.; Yokoyama, S. Nucleosides Nucleotides 1992, 11, 759. In this
paper, the authors reported that N-acetyl-2′-O-methylcytidine (ac4Cm)
found in the anticodon first letter of various tRNA species has the
preferred N-type conformer due to through-space and hydrogen bond-
ing interactions.3
To synthesize 1, 5-(cyanomethyl)-2′,3′-O-isopropylide-
neuridine (2)10 was employed as the starting material
(Scheme 1). Treatment of 2 with 1.5 equiv of 4-mono-
methoxytrityl chloride (MMTrCl) in pyridine gave the 5′-
(7) (a) Yokoyama, S.; Yamaizumi, Z.; Nishimura, S.; Miyazawa, T.
Nucl. Acids Res. 1979, 6, 2611. (b) McCloskey, J . A.; Nishimura, S.
Acc. Chem. Res. 1977, 10, 403. (c) Yokoyama, S.; Miyazawa, T. J . Mol.
Struc. 1985, 126, 563. (d) Sierzputowska-Gracz, H.; Sochacka, E.;
Malkiewicz, A.; Kuo, K.; Gehrke, C. W.; Agris, P. F. J . Am. Chem. Soc.
1987, 109, 7171. (e) Agris, P. S.; Sierzputowska-Gracz, H.; Smith, W.;
Malkiewicz, A.; Sochacka, E.; Nawrot, B. J . Am. Chem. Soc. 1992, 114,
2652.
(8) Sakamoto, K.; Kawai, G.; Niimi, T.; Satoh, T.; Sekine, M.;
Yamaizumi, Z.; Nishimura, S.; Miyazawa, T.; Yokoyama, S. Eur. J .
Biochem. 1993, 216, 369.
(9) Maruyama et al.1a reported the synthesis of a cycloadenylic acid
derivative bridged with
a phosphorus atom between the adenine
(5) Sakamoto, K. et al. To be published.
8-position and the 5′-hydroxyl group. This compound exhibited the J 1′,2′
value of 2.93 Hz and no coupling between 3′-H and 4′-H. This result
suggests that this type of compound has an unusual sugar conforma-
tion.
(10) Ikeda, K.; Tanaka, S.; Mizuno, Y. Chem. Pharm. Bull. 1975,
23, 2958.
(6) For papers on other factors which control the equilibrium
between an N- and an S-type conformer see: Uesugi, S.; Miki, H.;
Ikehara, M.; Iwahashi, H.; Kyogoku, Y. Tetrahedron Lett. 1979, 4073.
Thibaudeau, C.; Plavec, J .; Garg, N.; Papchikhin, A.; Chattopadhyaya,
J . J . Am. Chem. Soc. 1994, 116, 4038 and references cited therein.
0022-3263/96/1961-1500$12.00/0 © 1996 American Chemical Society

Notes
J . Org. Chem., Vol. 61, No. 4, 1996 1501
S-type conformers of 1 were calculated by the van’t Hoff’s
plot analysis15 (Figure 1) of the equilibrium constants K
) % S/% N measured at different temperatures, which
was obtained by NMR.
The ∆H and ∆S values were 1.0 kcal/mol and 0.6 cal/
K‚mol, respectively, showing a characteristic combination
of relatively large positive ∆H and relatively small
positive ∆S values compared to those of the other uridine
nucleosides and nucleotides reported.15 From these
thermodynamic parameters, it was concluded that the
N-type conformation of 1 is highly stabilized in an
enthalpic manner due mainly to steric interaction be-
tween the 2-keto and 2′-hydroxyl groups17 and is further
stabilized in an entropic manner which arises from the
cyclic structure.18
The 3D structure model of 1 was constructed by using
MacroModel Version 4.5 (AMBER*)19 with the aid of
molecular mechanics and Monte Carlo algorithm by using
the implicit treatment of solvent water with the GB/SA
model.20 In all simulations, the proton-proton distances
estimated from the 1D-NOE and T1 values21 and a
dihedral angle (60°) around the C3′-C4′-C5′-O5′ bond
from the 1H-NMR J coupling data22 were employed as
the constraints. As shown in Figure 2, the two lowest
energy structures (A, -957.26, and B, -958.76 kJ /mol)
were found out of 1000 conformations. Both structures
have the C3′-endo conformation with anti glycosyl angles
(ø ) -134° (A) and -162° (B)). These results are in
agreement with those obtained by 1H-NMR as mentioned
above.
F igu r e 1. Temperature dependence of the equlibrium con-
stants K[C2′-endo]/[C3′-endo] of 1.
masked product 3 in 77% yield. The cyano group of 3
was hydrolyzed with 1 M NaOH-EtOH to give the
carboxylic acid derivative 4 in 83% yield. Reduction of
4 with 1.7 equiv of borane-methyl sulfide complex11 in
THF gave the 5-(hydroxyethyl)uridine derivative 5 in
71% yield. Treatment of 5 with 1% trifluoroacetic acid
in CH2Cl2 afforded 6 in 79% yield. Compound 6 was
treated with 1.2 equiv of bis(diisopropylamino)(2-cyano-
ethoxy)phosphine in the presence of 1H-tetrazole. Suc-
cessive oxidation using tert-butyl hydroperoxide gave the
cyclized product 7 in 77% yield. Finally, the 12-
membered ring phosphotriester 7 thus obtained was
deprotected by treatment with concentrated ammonia-
pyridine followed by hydrolysis with 60% HCOOH to give
the desired product 1 in 80% yield. The structure of 1
The CD spectrum of 1 in H2O at 25 °C showed an
enhanced positive Cotton effect ([θ]270
) 10 500)
nm
compared with that of uridine as shown in Figure 3. As
seen in Figure 3, the CD spectrum of 1 at a high
temperature of 80 °C was not affected essentially. These
results also reflect the presence of a highly fixed anti-
form.
1
was characterized by UV and H, 13C, and 31P NMR.
The cyclic uridylic acid 1 having a phosphodiester
linkage was completely resistant (24 h) to snake venom
phosphodiesterase, nuclease P1, and calf spleen phos-
phodiesterase, which digest oligonucleotides to give 5′-
or 3′-nucleotides. These results indicated that more
flexible structures or dinucleotide units might be required
for the enzymatic hydrolysis of internucleotidic phos-
phodiester linkages.
In order to study the conformational property of 1, all
3
the J H-H values were determined. A significant confor-
mational feature of 1 is the high population density of
the N-type conformer as suggested by its very small J 1′,2′
value () 1.7 Hz). This J value is comparable to those of
5-nitrouridine (J 1′,2′ ) 1.7 Hz).3a Similar small J 1′,2′ values
were also reported in a series of naturally occurring
5-substituted 2-thiouridines.7a,12 The fractional popula-
tion of the N-type conformer in 1 was estimated to be
It is likely that there are at least three intrinsic factors
which control the predominancy of N-conformers in
modified uridines. One is the steric hindrance around
81% according to the equation of %N ) J 3′,4′/(J 1′,2′
3′,4′).13
To obtain more quantitative information about the
+
J
(15) Yokoyama, S.; Watanabe, T.; Murao, K.; Ishikura, H.; Yamai-
zumi, Z.; Nishimura, S.; Miyazawa, T. Proc. Natl. Acad. Sci. U.S.A.
1985, 82, 4905.
(16) Sierzputowska-Gracz, H.; Guenther, R. H.; Agris, P. F.; Folk-
man, W.; Golankiewicz, B. Magn. Reson. Chem. 1991, 29, 885.
(17) Kawai, G.; Yamamoto, Y.; Kamimura, T.; Masegi, T.; Sekine,
M.; Hata, T.; Iimori, T.; Watanabe, T.; Miyazawa, T.; Yokoyama, S.
Biochemistry 1992, 31, 1040.
(18) Altona, C.; Sundaralingam, M. J . Am. Chem. Soc. 1973, 95,
2333.
(19) Mohamadi, F.; Richards, N. G.; Guida, W. C.; Liskamp, R.;
Caufield, C.; Change, T.; Hendrickson, T.; Still, W. C. J . Comput. Chem.
1990, 11, 440.
(20) Still, W. C.; Tempczyk, A. T.; Hawley, R. C.; Hendrickson, T.
J . Am. Chem. Soc. 1990, 112, 6127.
(21) (a) Rosevear, P. R.; Mildvan, A. S. Methods Enzymol. 1989, 177,
333. (b) Claude, R. J .; Sikakana, T.; Hehire, S.; Kuo, M. C.; Gibbsons,
W. A. Biophys. J . 1978, 24, 815. (c) Kuo, M. C.; Gibbsons, W. A. Ibid.
1980, 32, 807.
(22) (a) Rousse, B.; Sund, C. Glemarec, C.; Sandstrom, A.; Agback,
P.; Chattopadhyaya, J . Tetrahedron 1994, 50, 8711. (b) Rousse, B.;
Puri, N.; Viswanadham, G.; Agback, P.; Glemarec, C.; Sandstrom, A.;
Sund, C. Chattopadhyaya, J . Ibid. 1994, 50, 1777.
sugar puckering of 1, the thermodynamic parameters, ∆H
and ∆S values, in the equilibrium between the N- and
(11) Brown. H. C.; Choi. Y. M.; Narasimham, S. J . Org. Chem. 1982,
47, 3153.
(12) 5-Methoxy-2-thiouridine (mo5s2U, 1.7 Hz),7d 5-[(methoxycarbo-
nyl)methoxy]-2-thiouridine (mcmo5s2U, 1.7 Hz),7d 5-[[(carboxymethyl)-
amino]methyl]-2-thiouridine (cmnm5s2U, 1.8 Hz),7d 5-[(methoxycarbonyl)-
methyl]-2-thiouridine (mcm5s2U, 2.3 Hz),7a 5-[(methylamino)methyl]-
2-thiouridine (mnm5s2U, 2.3 Hz),7a,d 2-thiouridine (s2U, 2.5 Hz).7d
(13) In general, the fractional population of the N- and S-conformers
in a ribonucleoside can be estimated by several equations: (1) % N )
J 3′,4′/(J 1′2′ + J 3′4′),14 (2) % N ) 1-(J 1′2′/10) (if J 3′4′ can not be determined),15
(3) % N ) [(7.5 - J 1′,2′)/6].7d,16 To compare the relative rigidity of 1
with those of known modified uridine derivatives, we employed eq 1
in this study. The % N values of the above 5-substituted 2-thiouridines
are given from the reported data7a,d of J 1′2′ and J 3′4′ as follows: mo5s2U,
82%; mcmo5s2U, 74%; cmnm5s2U, 77%; mcm5s2U, 78%; s2mnm5U, 78%;
s2U, 71%.
(14) Altona, C.; Sundaralingham, M. J . Am. Chem. Soc. 1973, 95,
2333.

1502 J . Org. Chem., Vol. 61, No. 4, 1996
Notes
F igu r e 2. Two lowest energy conformations of 1. Left (A): E ) -957.26 kJ /mol. Right (B): E ) -958.76 kJ /mol.
conformation when incorporated into uridylate dimers.
Further studies are needed to confirm this.
In conclusion, the present study suggests that the
considerable predominance of the C3′ endo conformer in
pmnm5U can be rationalized in terms of intramolecular
hydrogen bond formation. Cyclic uridylic acid 1 should
be useful as a thermodynamically rigid simple model of
A-form nucleotides for conformational studies of nucleic
acids.
Exp er im en ta l Section
Gen er a l Meth od s. TLC was performed by the use of Merck-
Kieselgel 60-F254 (0.25 mm). Column chromatography was
performed with silica gel C-200 purchased from Wako Co., Ltd.,
and a minipump for a goldfish bowl was conveniently used to
attain sufficient pressure for rapid choromatographic separation.
Reversed-phase column chromatography was performed by the
use of µBondapak C-18 silica gel (prep S-500, Waters). Reversed-
phase HPLC was performed on a Waters LC module 1 using a
µBondasphere C-18 column with a linear gradient starting from
0.1 M NH4OAc, pH 7.0 and applying CH3CN at a flow rate of
1.0 mL/min for 30 min. Uridine was purchased from Yamasa
Co., Ltd. Pyridine was distilled two times from p-toluenesulfonyl
chloride and from calcium hydride and then stored over mol-
eculer sieves 4A. Elemental analyses were performed by the
Microanalytical Laboratory, Tokyo Institute of Technology at
Nagatsuta.
5-(Cya n om et h yl)-2′,3′-O-isop r op ylid en e-5′-O-(4-m on o-
m eth oxytr ityl)u r id in e (3). Compound 2 (1.0 g, 3.1 mmol) was
rendered anhydrous by repeated coevaporation with dry pyridine
and finally dissolved in dry pyridine (30 mL). To the solution
was added 4-monomethoxytrityl chloride (1.43 g, 4.6 mmol).
After being stirred at room temperature for 15 h, the mixture
was treated with water and then extracted with ether. The
ethereal extract was washed three times with saturated NaH-
CO3, dried over Na2SO4, filtered, and repeatedly coevaporated
with toluene to dryness. The residue was dissolved in CH2Cl2
and chromatographed on a silica gel column with hexane-ethyl
acetate (7:3, v/v) to give 3 (1.4 g, 77%): 1H-NMR (270 MHz,
CDCl3) δ 1.38 and 1.60 (3H each, s), 2.77 (2H, br), 3.45-3.55
(2H, m), 3.80 (3H, s), 4.36 (1H, m), 4.82-4.89 (2H, m), 6.00 (1H,
d, J 1′,2′ ) 3.0 Hz), 6.85 (2H, d, J ) 8.9 Hz), 7.26-7.41 (10H, m),
7.74 (1H, s); 13C-NMR (68 MHz, CDCl3) δ 15.01, 25.43, 27.24,
55.22, 63.61, 80.77, 84.62, 85.70, 87.14, 92.40, 105.18, 113.30,
114.54, 127.40, 128.03, 128.36, 128.43, 130.39, 134.52, 139.39,
F igu r e 3. CD spectra of 1 and uridine in 10 mM phosphate
buffer (pH 7.0) at 25 and 80 °C.
the 2′-hydroxyl group.7a,17 It is well known that the rigid
structure of s2U is essentially attributable to the steric
interaction between the bigger 2-(thiocarbonyl) group and
the neighboring 2′-hydroxyl group.7a,c,d The second is the
through-space interaction between the 4′-oxygen lone-
paired electrons and the antibonding orbital of the
5-carbon of the 5,6 double bond of the uracil residue, as
discussed in 5-nitorouridine.3a,4 The third is the intramo-
lecular interaction between the 5′-phosphate and the
5-substituent.5 We observed that the cyclic covalent
bonding bridge between the 5′-phosphate and 5-position
of the uracil induces the 3′-endo conformation in 1. This
result implies that intramolecular interactions, such as
hydrogen bonding and van der Waals contact, between
the 5′-phosphate and the 5-[(methylamino)methyl] sub-
stituent are highly plausible as essential factors for
stabilization of the 3′-endo form. Therefore, the com-
pound 1 that we designed is rational for chemical fixation
of the N-type conformation of uridylic acid.
(23) Agris et al. also reported that, in s2UpU and s2mnmUpU, s2U
and s2mnmU exist essentially as the N-conformers: Smith, W. S.;
Sierzutowska-Gracz, H.; Sochacka, E.; Malkiewicz, A.; Agris, P. F. J .
Am. Chem. Soc. 1992, 114, 7989. On the other hand, it is also known
that dihidrouridine is almost 100% S in the crystal (Emerson, J .;
Sundaralingam, M. Acta Crystallogr. 1988, 563, 206) and in solution
(Nawrot, B.; Malkiewicz, A.; Smith, W. S.; Sierzputoska-Gracz, H.
Agris, P. F. Nucleosides Nucleotides 1995, 14, 143).
It will be interesting to synthesize a cyclic U*pU
derivative containing 1 (U* refers to a cyclic bridge
structure) and to see if the cyclonucleotide 1 incorporated
becomes more extremely rigid in the 3′-endo conforma-
tion, since Agris23 reported that some 5-substituted
2-thiouridines are considerably stabilized in the N-

Notes
J . Org. Chem., Vol. 61, No. 4, 1996 1503
143.59, 158.83; FAB HRMS calcd for C34H32O7N3 (M - H)-
594.2241, found 594.2213.
Ta ble 1. T1 Va lu es of th e Ribose P r oton s of Com p ou n d 1
proton
T1 (s)
5-(Ca r boxym eth yl)-2′,3′-O-isop r op ylid en e-5′-O-(4-m on o-
m eth oxytr ityl)u r id in e (4). Compound 3 (2.7 g, 4.4 mmol) was
dissolved in 1 M NaOH-ethanol (50 mL-50 mL), and the
mixture was stirred at 80 °C for 36 h. The solution was
neutralized with Dowex 50W X 8 (pyridinium form, 20 mL),
filtered, and evaporated to dryness. The residue was chromato-
graphed on a silica gel column with CH2Cl2-MeOH (100:4-100:
10, v/v) to give 4 (2.4 g, 83%): 1H-NMR (270 MHz, CDCl3) δ
1.38 and 1.60 (3H each, s), 2.77 (2H, br), 3.51 (2H, m), 3.77 (3H,
s), 4.33 (1H, m), 4.88 (1H, m), 4.93 (1H, dd, J 1′,2′ ) 2.3 Hz, J 2′,3′
) 6.3 Hz), 6.02 (1H, d, J 1′,2′ ) 2.3 Hz), 6.86-6.82 (2H, m), 7.21-
7.42 (13H, m), 7.58 (1H, s, 6-H), 9.20, 9.84 (2H, br); 13C-NMR
(68 MHz, CDCl3) δ 25.43, 27.24, 40.00, 55.15, 63.63, 65.77, 80.74,
84.42, 85.46, 86.96, 91.91, 108.88, 113.23, 114.41, 127.24, 127.92,
128.39, 130.33, 134.79, 139.60, 143.54, 143.68, 150.04, 158.71,
163.61, 174.29. Anal. Calcd for C34H34N2O9‚3H2O: C, 61.63;
H, 6.08; N, 4.23. Found: C, 61.52; H, 5.81; N, 4.17.
5-(H yd r oxyet h yl)-2′,3′-O-isop r op ylid en e-5′-O-(4-m on o-
m eth oxytr ityl)u r id in e (5). Compound 4 (1.6 g, 2.6 mmol) was
dissolved in dry THF (10 mL). A solution of borane-methyl
sulfide (290 µL, 3.1 mmol) in THF (10 mL) was added dropwise
over 30 min and stirred at room temperature. After the solution
was stirred for an additional 1.5 h, borane-methyl sulfide (120
µL, 1.3 mmol) was added, and the mixture was stirred for 3.5 h.
The solution was diluted with CH2Cl2 and washed three times
with saturated NaHCO3. The aqueous layer was extracted with
CH2Cl2, and the CH2Cl2 extract was dried over Na2SO4, filtered,
and evaporated to dryness under reduced pressure. The residue
was dissolved in CH2Cl2 and chromatographed on a silica gel
column with CH2Cl2-MeOH (100:1.5, v/v) to give 5 (1.1 g,
71%): 1H-NMR (270 MHz, CDCl3) δ 1.30 and 1.53 (3H each, s),
2.06 (2H, m), 3.31-3.46 (6H, m), 3.73 (3H, s), 4.24 (1H, m), 4.76
(1H, dd, J 2′,3′ ) 6.3 Hz, J 3′,4′ ) 3.3 Hz), 4.80 (1H, dd, J 1′,2′ ) 3.0
Hz, J 2′,3′ ) 6.3 Hz), 5.93 (1H, d, J 1′,2′ ) 3.0 Hz), 6.78 (2H, d, J )
8.9 Hz), 7.18-7.26 (13H, m), 9.28 (1H, s); 13C-NMR (68 MHz,
CDCl3) δ 25.43, 27.23, 30.32, 53.37, 55.17, 61.30, 63.58, 80.72,
84.15, 85.16, 86.83, 91.68, 112.69, 113.12, 114.50, 127.15, 127.85,
128.39, 128.43, 130.39, 134.79, 138.54, 143.74, 150.01, 158.67,
164.53. Anal. Calcd for C34H36N2O8: C, 67.99; H, 6.04; N, 4.66.
Found: C, 67.53; H, 6.13; N, 4.74.
H6
1.7
3.4
1.9
1.5
0.9
0.4
0.5
H1′
H2′
H3′
H4′
H5′
H6′
(J C,P ) 4.8 Hz), 68.40 (J C,P ) 7.5 Hz), 69.70 (J C,P ) 7.5 Hz), 80.67,
86.74, 95.42, 107.06, 113.01, 116.33, 137.84, 150.0, 163.67; 31P-
NMR (109 MHz, CDCl3-CD3OD, 9:1, v/v) δ -3.09.
One of the diastereoisomers: 1H-NMR (270 MHz, CDCl3-CD3-
OD, 9:1, v/v) δ 1.36 and 1.51 (3H each, s), 2.63 (2H, t, J ) 5.6
Hz), 3.07 (2H, br), 4.06-4.44 (7H, m), 4.84 (2H, m), 6.09 (1H,
s), 7.46 (1H, s); 13C-NMR (68 MHz, CDCl3-CD3OD, 9:1, v/v) δ
19.45 (J C,P ) 7.5 Hz), 24.85, 26.00, 26.54, 61.90 (J C,P ) 4.8 Hz),
67.60 (J C,P ) 6.1 Hz), 68.10 (J C,P ) 7.5 Hz), 79.73, 86.50 (J C,P
)
8.8 Hz), 91.52, 109.26, 113.93, 116.37, 137.32, 150.24, 163.56;
31P-NMR (109 MHz, CDCl3-CD3OD) δ -3.12. Anal. Calcd for
C
17H22N3O9P‚1/2H2O: C, 45.14; H, 5.12; N, 9.29. Found: C,
45.56; H, 5.45; N, 8.29.
Cyclic Ur id ylic Acid Der iva tive 1. Compound 7 (220 mg,
0.5 mmol) was dissolved in concd ammonia-pyridine (25 mL-
25 mL), and the resulting mixture was stirred for 1.5 h. The
solvents were removed under reduced pressure, and the residue
was dissolved in 60% HCOOH (50 mL). After 20 h, the solvent
was removed and the residue was chromatographed on a C-18
reversed-phase silica gel column with water to give 1 (169 mg,
80%). The cyclonucleotide was further purified by reversed-
phase HPLC: HPLC retention time, 3.6 min; UV λmax 264 nm,
λmin 233 nm; 1H-NMR (270 MHz, D2O) δ 2.63 (2H, m), 4.19 (1H,
ddd, J 4′,5′′ ) 2.4 Hz, J 5′,5′′ ) 12.2 Hz, J 5′′,P ) 2.4 Hz), 4.25-4.37
(3H, m), 4.40 (1H, dd, J 1′,2′ ) 1.6 Hz, J 2′,3′ ) 4.5 Hz), 4.46 (1H,
dd, J 2′,3′ ) 4.5 Hz, J 3′,4′ ) 6.9 Hz), 4.50 (1H, ddd, J 4′,5′ ) 2.4 Hz,
J 5′,5′′ ) 12.3 Hz, J 5′,P ) 5.6 Hz), 5.90 (1H, d, J 1′,2′ ) 1.6 Hz), 8.38
(1H, s); 13C-NMR (68 MHz, D2O) δ 22.79 (J C,P ) 7.1 Hz), 60.75
(J C,P ) 6.1 Hz), 61.34 (J C,P ) 6.1 Hz) 65.75, 72.52, 80.45 (J C,P
)
8.6 Hz), 87.87, 108.22, 134.32, 148.48, 163.59; 31P-NMR (109
MHz, D2O) δ 0.11; FAB HRMS calcd for C11H14O9N2P (M - H)-
349.0437, found 349.0425.
5-(Hydr oxyeth yl)-2′,3′-O-isopr opyliden eu r idin e (6). Com-
pound 5 (1.0 g, 1.7 mmol) was dissolved in 1% trifluoroacetic
acid-CHCl3 (100 mL) and stirred for 2 h. The solution was
partitioned between CH2Cl2-H2O, and the organic layer was
washed three times with saturated NaHCO3, dried over Na2-
SO4, filtered, and repeatedly coevaporated with toluene to
dryness. The residue was chromatographed on a silica gel
column with CH2Cl2 to give 6 (432 mg, 79%): 1H-NMR (270
MHz, CDCl3-CD3OD, 9:1, v/v) δ 1.36 and 1.58 (3H each, s), 2.52
(2H, t, J ) 5.6 Hz), 3.69-3.77 (3H, m, 5′′-H), 3.87 (1H, d, J )
12.2 Hz), 4.30 (1H, m), 4.88 (2H, m), 5.85 (1H, s), 7.59 (1H, s);
13C-NMR (68 MHz, CDCl3-CD3OD, 9:1, v/v) δ 24.98, 26.90,
29.71, 60.04, 61.83, 80.32, 84.37, 86.58, 93.08, 111.27, 113.98,
139.05, 150.42, 164.37. Anal. Calcd for C14H20N2O7: C, 51.22;
H, 6.14; N, 8.53. Found: C, 51.36; H, 6.43; N, 7.93.
F u lly P r otected Cyclon u cleotid e 7. Compound 6 (126 mg,
0.38 mmol) and 1H-tetrazole (106 mg, 0.46 mmol) were rendered
anhydrous by repeated coevaporation with dry pyridine and
successively with dry toluene and finally dissolved in CH3CN
(30 mL). Bis(diisopropylamino)(2-cyanoethoxy)phosphine (146
µL, 0.46 mmol) was added dropwise over 10 min. After the
mixture was stirred for 30 min, tert-butyl hydroperoxide (380
µL, 3.8 mmol) was added, and the resulting mixture was stirred
for an additional 1.5 h. The solvent was removed under reduced
pressure, and the residue was partitioned between CH2Cl2-H2O.
The organic layer was washed three times with saturated
NaHCO3, dried over Na2SO4, filtered, and coevaporated repeat-
edly with toluene to dryness. The residue was chromatographed
on a silica gel column with CH2Cl2-MeOH (100:1.5, v/v) to give
7 (169 mg, 77%).
Deter m in a tion of Th er m od yn a m ic P a r a m eter s of th e
Ribose P u ck er in g Usin g 1H-NMR sp ectr a . Compound 1 (50
A
260) was dissolved in 10 mM sodium phosphate buffer (700 µL),
and the pH value was adjusted to 7.0 by addition of 1 M NaOH
and lyophilized. The remaining H2O was removed by repeated
lyophylization three times with D2O. The residue was finally
dissolved in D2O (700 µL). The vicinal coupling constants J 1′,2′
,
J
2′,3′, and J 3′,4′ at 20, 25, 30, 35, 40, 45, 50, 60, and 65 °C were
measured quantitatively by using the Deconvolusion program
implemented in the same spectrometer. The digital resolution
after the Deconvolusion was 0.15 Hz. The fractional population
of the S and N conformations, % S and % N, respectively, were
obtained from the formula % N ) J 3′,4′/(J 1′,2′+J 3′,4′) and % S ) 1
- % N. The equilibrium constants, % S/% N, at each temper-
ature were subjected to the van’t Hoff’s plot analysis with the
aid of the method of least -squares analysis to give the enthalpy
difference (∆H) and the entropy difference (∆S).
T1 a n d NOE Mea su r em en t. Compound 1 (50A260) was
dissolved in 10 mM sodium phosphate buffer (700 µL), and the
pH value was adjusted to 7.0 by addition of 1 M NaOH and
lyophilized. The remaining H2O was removed by repeated
lyophilization three times with D2O. The remaining oxgen gas
was removed by repeated displacement of the air in the sample
tube with argon. The T1 values were measured according to the
(selective saturation)-t1-(observation) sequence where the t1
values were set to 50-500 ms with the increment of 50 ms. The
time dependence of the recovery ratio of saturated protons was
quantified by comparison of the peak intensity with those of the
nonirradiated peaks. The 1D-differential NOE spectra were
measured using the noedif program implemented in the spec-
trometer where the presaturation time was set to 8 s. The
quantity of the magnetization transfer to proton i from proton
j, ηi(j), were measured relative to the saturated peaks in the 1D-
differential NOE spectra.
One of the diastereoisomers: 1H-NMR (270 MHz, CDCl3-CD3-
OD, 9:1, v/v) δ 1.28 and 1.48 (3H each, s), 2.40 (1H, m), 2.71
(2H, t, J ) 6.3 Hz), 2.82 (1H, d, J ) 13 Hz), 4.11-4.33 (6H, m),
4.66 (1H, br), 4.74 (1H, d, J 2′,3′ ) 5.7 Hz), 4.85 (1H, d, J 2′,3′ ) 5.9
Hz), 5.68 (1H, s), 7.42 (1H, s); 13C-NMR (68 MHz, CDCl3-CD3-
OD, 9:1, v/v) δ 19.30 (J C,P ) 7.5 Hz), 26.24, 26.79, 29.42, 61.60
Deter m in a tion of th e In ter p r oton Dista n ces a n d th e
Con for m a tion a r ou n d th e C4′-C5′ Bon d . The cross relax-

1504 J . Org. Chem., Vol. 61, No. 4, 1996
Notes
Ta ble 2. NOE En h a n cem en t (%) of th e Su ga r P r oton s of
Com p ou n d 1a
Ta ble 4. In ter p r oton Dista n ce (Å) Der ived fr om th e
Cr oss-Rela xa tion Ra tes Listed in Ta ble 3a
H6
H1′
H2′
H3′
H4′
H5′
H5′′
H6
H1′
H2′
H3′
H4′
H5′
H5′′
H6
*
3.4
*
4.9
6.0
*
-
0
9.1
0
-
*
0
-
3.0
0
1.4
0
0
*
0
0
0
0
0
*
34.4
0
0
H6
*
3.7
*
3.2
3.1
*
2.8
-
-
-
-
-
-
-
*
-
H1′
H2′
H3′
H4′
H5′
H5′′
1.2
5.8
8.8
3.5
-
H1′
H2′
H3′
H4′
H5′
H5′′
4.0
3.1
2.8
3.3
-
3.5
-
-
14.5
2.8
2.9
-
-
1.6
1.3
-
*
3.1
3.9
3.9
-
-
-
-
*
-
*
-
2.9
2.3
2.3
-
-
-
-
0
-
4.5
-
-
1.2
0
4.0
-
-
3.3
1.8
*
a
a
The first column represents the irradiated protons and the
first row the observed protons. - indicates that the corresponding
NOE was not observed quantitatively because of the nonselective
irradiation of the neighboring proton signals.
The distance between H5′ and H5′′, 1.8 Å, was used as a
reference. The cells where the correponding NOE were not
observed or not quantitative are indicated with a -.
ployed as the constraints. A potential energy term representing
the interproton distance restraints was added to the total energy
of the system according to the formula E ) 0 (if (r - r0)2 < 0.04
Å) or E ) k(r - r0)2 (if (r - r0)2 > 0.04 Å),27 where k ) 100 kJ /
mol and r0 is the interproton distance derived from the NOE
data described above. The torsion around the C4′-C5′ bond was
also constrained to 60 ( 40° with the force constant of 1000 kJ /
mol‚rad2 in order to reproduce the g+ conformation which was
suggested from the 1H-NMR spectrum of 1 as described above.
The initial structure was built from the uridine residue imple-
mented in the MacroModel Version 4.5 structure file. In total,
1000 conformers were generated by using the MCMM option,
and the structures were energy-minimized 500 times. Conse-
quently, the two lowest energy structures (A, -957.26, and B,
-958.76 kJ /mol) were obtained as shown in Figure 2. Both
structures A and B have C3′-endo conformations with anti
glycosyl angles ø (O4′-C1-N1-C2) ) -134° and -162°, respec-
tively.
Ta ble 3. Cr oss-Rela xa tion Ra tes (σ) betw een Tw o
P r oton s Ca lcu la ted Usin g th e NOE En h a n cem en t Va lu e
(η) a n d th e T1 Va lu e of Ea ch P r oton Accor d in g to th e
i
F or m u la σij ) ηi(j)/T1
H6
H1′
H2′
H3′
H4′
H5′
H5′′
H6
*
1.1
*
2.5
3.0
*
-
0
6.2
0
-
*
0
1.6
0
0
*
0
0
0
-
0
*
81.9
0
0
-
H1′
H2′
H3′
H4′
H5′
H5′′
0.71
3.5
5.2
2.0
-
3.4
0.81
0.84
-
20.2
2.8
-
-
-
0
-
-
5.1
0.71
0
2.0
*
ation rate between protons i and j, σi,j, was estimated by the
i
formula σi,j ) ηi(j)/T1 . The interproton distance Rij was esti-
mated according to the formula Ri,j ) R5′,5′′(σ5′,5′′/σi,j)1/6 24 using
σi,j thus obtained and R5′,5′′ ) 1.8 Å.25 These results are
summarized in Tables 1-4. The interproton distance between
two protons, i and j, which was used in the computer simulation
described below was the averaged value of Ri,j and Rj,i thus
obtained.
CD Sp ectr a . The sample (0.47A260) was dissolved in 10 mM
sodium phosphate buffer (pH 7.0). The spectra were recorded
at 25 and 80 °C in a 0.5 cm pathlength cell.
Ack n ow led gm en t. This work was supported by the
Ciba-Geigy Foundation for the Promotion of Science and
a Grant-in-Aid for Scientific Research from the Ministry
of Education, Science and Culture, J apan.
The conformation around the C4′-C5′ bond was analyzed by
using the equation % g+ ) [13.3 - (J 4′,5′ + J 4′,5′′)]/9.7.26 The % g+
of 1 was calculated to be 89% at 25 °C.
3D Mod el Bu ild in g of 1 Usin g Molecu la r Mech a n ics.
The NMR refined structure of 1 was built using MacroModel
Version 4.5 with the aid of molecular mechanics and Monte Carlo
calculation using AMBER* force field.19 The effect of solvation
was included in these simulations by using the implicit treat-
ment of solvent water with GB/SA model.20 In all simulations,
the proton-proton distances estimated from the 1D-NOE and
T1 values21 and a dihedral angle (60°) around the C3′-C4′-
C5′-O5′ bond from the 1H-NMR J coupling data22 were em-
Su p p or tin g In for m a tion Ava ila ble: 1H, 13C, and 31P
NMR and mass spectra of the synthetic intermediates and the
final product 1 (20 pages). This material is contained in
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
J O951756X
(24) Wagner, G.; Wu¨thrich, K. J . Magn. Reson. 1979, 33, 675.
(25) Wu¨thrich, K. In NMR of Proteins and Nucleic Acids; J ohn Wiley
& Sons, Inc.: New York, 1986.
(27) Fesik, S. W.; Bolis, G.; Sham, H. L.; Olejniczak, E. T. Biochem-
istry 1987, 26, 1851.
(26) Altona C. Recl. Trav. Chim. Pays-Bas 1982, 101, 413.