6
226
R. W. Humble et al. / Tetrahedron Letters 52 (2011) 6223–6227
with the presence of the neighbouring amide NH. TLC plate stain-
ing using ferric chloride indicated that only one of the by-products
reacted, which is consistent with the presence of the aromatic alco-
However, extending the reaction time to three hours resulted in
deprotection of both the acetyl and isopropylidene groups yielding
bredinin. A comparison was made of this material with an authen-
1
hol group in 14b, and its absence in 15b. The synthesis of
a
-bredi-
tic specimen of bredinin (Sigma) and the H NMR spectroscopic
nin did not demonstrate a similar selectivity since the reaction
using either a molar equivalent or slight excess of ethyl formimi-
date resulted in a mixture. These nucleosides resisted separation
but TLC again showed two characteristic UV- and ferric chloride-
data were found to be identical.
Conclusions
positive spots (indicative of 11
positive spot, indicative of oxazole 15
both the - and b-series proved difficult to separate or purify
due to their polarity. Bredinin itself exists predominantly in a zwit-
terionic form (see Fig. 1). Attempts at derivatising the 5-OH group
to facilitate separation of the mixtures were problematic; acylation
provided labile esters which did not survive chromatography, and
alkylation was hampered by competitive reaction at N-3.
a
and 14
a
) along with one UV-only
We have developed a convenient synthesis of bredinin (1) from
a
. Nucleoside derivatives in
0
a simple sugar amide precursor. Cyclisation of 2-amino-N-(5 -O-
a
0
0
acetyl-2 ,3 -O-isopropylidene-
D-ribofuranosyl)
malondiamide
(
10b) is best achieved by treatment with ethyl formimidate, in
place of the more commonly used triethyl orthoformate. We have
found that this strategy can equally be applied to the synthesis of
the a-anomer of bredinin though competing modes of cyclisation,
even using the conditions optimised for the b-anomer, leading to
the production of alternatively cyclised compounds which could
not be separated. In addition, cyanoamide 16b, produced in an ano-
mer-selective reaction, could be converted into the cyano analogue
0b of bredinin in a sequence involving direct amination of 16b
using O-mesitylsulfonylhydroxylamine.
With the success of making a nucleoside by this approach, we
turned our attention to the preparation of the 5-cyano analogue
which we envisaged could be made by the same route using cyano-
acetyl chloride in the place of ethyl malonyl chloride. This route
was of interest as the nitrile group should be amenable to modifi-
cation leading to a range of analogues in the 4-position of the imid-
azole ring (hydrolysis of 4-carboxamide in bredinin is not possible;
the 5-hydroxy-4-carboxamide arrangement is formally a b-keto
acid which decarboxylates). Thus, freshly prepared cyanoacetyl
2
Acknowledgments
chloride1 was reacted with 2,3-O-isopropylidene-
6
We thank Toyo Jozo Co., Ltd. (now part of the Asahi Chemical
Industry Co., Ltd.) (R.W.H.) and the EPSRC (D.F.M.) for funding,
and also Dr. Kiyofumi Fukukawa (Toyo Jozo Co., Ltd.) for useful
discussions.
D-ribofuranosyl-
amine at ꢀ10 °C to give 16, in which only the crystalline b anomer
was isolated (Scheme 2). The temperature of this reaction was
important as at +10 °C the
a anomer predominated (4:1), and at
temperatures higher still the product (mp 155–6 °C) from N- and
0
References and notes
5
-O bis-acylation was formed. This by-product was readily crystal-
lised from the reaction mixture, but proved to be exclusively the
a
1
.
Mizuno, K.; Tsujino, M.; Takada, M.; Hayashi, M.; Atsumi, K.; Asano, K.;
Matsuda, T. J. Antibiot. 1974, 27, 775.
0
0
0
anomer [d 5.81 dd, 1H, J (H1 –NH) 8.7 Hz and J (H1 –H2 ) 4.7 Hz (H-
0
1
)]. Similar reactions of 16b to that described for the carboxamide
2. (a) Inou, T.; Kusaba, R.; Takahashi, I.; Sugimoto, H.; Kuzuhara, K.; Yamada, Y.;
Ohtsuka, Y.; Yamakawa, Y.; Ohtani, K.; Kudo, T.; Ohtomo, Y.; Nagata, S.;
Shimizu, T. Pediatr. Int. 2010, 52, e57–e59.
series were investigated, but we found that yields were lower and
the products proved particularly sensitive to glycosidic cleavage.
However it was discovered that a direct amination of 16b could
be achieved using O-mesitylsulfonylhydroxylamine,17 thus by-
3.
Ichinose, K.; Origuchi, T.; Kawashiri, S.-Y.; Iwamoto, N.; Fujikawa, K.; Aramaki,
T.; Kamachi, M.; Arima, K.; Tamai, M.; Nakamura, N.; Ida, H.; Kawakami, A.;
Tsukada, T.; Ueki, Y.; Eguchi, K. Intern. Med. 2010, 49(20), 2211.
passing the oxime stage and its reduction. This transformation
0
could be applied to the 5 -O unprotected derivative so avoiding
4. Franchetti, P.; Cappellacci, L.; Grifantini, M. Farmaco 1996, 51(7), 457.
5
6
.
.
Webster, H. K.; Whaun, J. M. J. Clin. Invest. 1982, 70, 461.
(a) Hayashi, M.; Hirano, T.; Yaso, M.; Mizuno, K.; Ueda, T. Chem. Pharm. Bull.
three steps. This method was unsuccessfully applied to the C-2
amination of substrates in the 4-carboxamide series. Thus treat-
ment of 16b with either NaH, or preferably LDA, followed by
quenching with O-mesitylsulfonyl hydroxylamine gave a mixture
of epimers of 17b. With NaH the conversion seemed to stop at
about 50% as judged by TLC. Fractional crystallisation of the crude
material from hot ethanol separated the mesitylsulfonic acid from
the unreacted starting material and product, but chromatography
to separate the latter was not trivial. Typically 20–25% of the pure
amine 17b could be isolated, but with significant fractions contam-
inated with the starting material which needed to be re-chromato-
graphed. With LDA the conversion could be made to go to
completion but only by the addition of 4 equiv of LDA and O-mes-
itylsulfonylhydroxylamine. We presume there is competition be-
1975, 23, 245; (b) Horton, J. K.; Stevens, M. F. G. J. Pharm. Pharmacol. 1981, 33,
808; (c) Tarumi, Y.; Atsumi, T. J. Heterocycl. Chem. 1983, 20, 875; (d) Tarumi, Y.;
Takebayashi, Y.; Atsumi, T. J. Heterocycl. Chem. 1984, 21, 849; (e) Tarumi, Y.;
Moriguchi, K.; Atsumi, T. J. Heterocycl. Chem. 1984, 21, 529.
Franchetti, P.; Pasqualini, M.; Cappellacci, L.; Petrelli, R.; Vita, P.; Grifantini, M.;
Jayaram, H. N. Nucleosides Nucleotides Nucleic Acids 2005, 24(10–12), 2023.
7
.
8. (a) Fukukawa, K.; Shuto, S.; Hirano, T.; Ueda, T. Chem. Pharm. Bull. 1984, 32,
644; (b) Fukukawa, K.; Shuto, S.; Hirano, T.; Ueda, T. Chem. Pharm. Bull. 1986,
4, 3653; (c) Shuto, S.; Haramuishi, K.; Fukuoka, M.; Matsuda, A. J. Chem. Soc.,
1
3
Perkin Trans. 1 2000, 3603.
9. Tipson, R. S. J. Am. Chem. Soc. 1961, 26, 2462.
10. Cusack, N. J.; Shaw, G. Chem. Commun. 1970, 1114.
11. HPLC semi-preparative reverse-phase ODS-HPLC column (Dynamax,
A
3
0 ꢁ 2.5 cm) was employed using a stepwise water–methanol gradient 0–
3
100% over a period of 1 h (flow rate 7 cm /min). A Gilson Holochrome UV–vis
detector was set at 250 nm for 18b (retention time 32 min), 280 nm for 11b
(
29 min) and for 1 (17 min).
0
tween deprotonation at the 5 -OH, the amide NH as well as the
12. Ewing, D. F.; Humble, R. W.; Mackenzie, G. J. Carbohydr. Chem. 1991, 10, 387.
active methylene protons, and only addition of excess base ensures
full deprotonation at the carbon a- to the nitrile.
13. (a) Harnden, M. R.; Parkin, A.; Wyatt, P. G. Tetrahedron Lett. 1988, 29, 701; (b)
Montgomery, J. A.; Temple, C. J. Am. Chem. Soc. 1957, 79, 5238.
1
1
4. Ohme, R.; Schmitz, E. Angew. Chem., Int. Ed. Engl. 1967, 6, 566.
Cyclisation of 17b was conducted as described previously, using
ethyl formimidate in DMF, and this gave the nucleoside 18b along
with a by-product tentatively assigned as 19 (Fig. 2) based on the
same observations made in the bredinin series. Finally, the treat-
ment of 18b with aqueous TFA for 30 minutes at room temperature
0
0
0
5. Synthesis of 11b. To dry 2-amino-N-(5 -O-acetyl-2 ,3 -O-isopropylidene-b-
ribofuranosyl)-malondiamide (410 mg, 1.24 mmol) in anhydrous
dimethylformamide (5 cm ) was added ethyl formimidate hydrochloride
136 mg, 1.24 mmol), under a nitrogen atmosphere. The reaction mixture
D-
3
(
was heated in an oil bath at 110 °C for 5 min. The flask was allowed to cool, and
ammonium chloride was removed by filtration through a cotton wool plug in a
Pasteur pipette. The solvent was removed using an oil pump vacuum (20 °C).
The pale green gum was dissolved in the minimal volume of dichloromethane
and purified by radial chromatography (Chromatotron, 4 mm silica disc). The
silica gel disc was equilibrated with dichloromethane, followed by a stepwise
methanol gradient (3–20%). Fractions containing the desired nucleoside were
1
8
gave cleanly the deprotected cyano analogue of bredinin, 20b.
Returning to the synthesis of bredinin (1), we found that the sim-
ilar treatment of 11b with aqueous TFA at room temperature re-
sulted in selective cleavage of the isopropylidene group.