A new approach to the 2,5-diamino-5,6-dihydro-1H-pyrimidine-4-one derivatives: Synthesis of TAN-1057A/B and analogs

Article abstract of DOI:10.1016/S0040-4039(03)01381-9

TAN-1057A/B has shown potent activity against MRSA. A new approach to the 2,5-diamino-5,6-dihydro-1H-pyrimidine-4-one derivatives has been developed. TAN-1057A/B and its analogs were efficiently synthesized using this new approach.

Full text of DOI:10.1016/S0040-4039(03)01381-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 5871–5873
A new approach to the
2,5-diamino-5,6-dihydro-1H-pyrimidine-4-one derivatives:
synthesis of TAN-1057A/B and analogs
Lijun Zhang, Lianhong Xu* and Choung U. Kim
Department of Medicinal Chemistry, Gilead Sciences, Inc., 333 Lakeside Drive, Foster City, CA 94404, USA
Received 7 May 2003; revised 3 June 2003; accepted 3 June 2003
Abstract—TAN-1057A/B has shown potent activity against MRSA. A new approach to the 2,5-diamino-5,6-dihydro-1H-pyrim-
idine-4-one derivatives has been developed. TAN-1057A/B and its analogs were efficiently synthesized using this new approach.
© 2003 Published by Elsevier Ltd.
TAN-1057A/B, isolated a few years ago from Flexibac-
ter in Japan,1 exhibits potent activity against methi-
cillin-resistant Staphylococcus aureus (MRSA) (Fig. 1).
As part of our structure–activity relationship (SAR)
study on TAN-1057A/B, methylated analogs 2 and 3
were targeted (Scheme 1).
iodomethane in the presence of cesium carbonate
yielded multiple products, and only a small amount of
the N-3 methylated product 5a could be isolated. In
order to synthesize 2 and 3 successfully, a new route
needed to be developed for the efficient construction of
heterocyclic compound 6. In this paper, we report a
new approach for the construction of the 2,5-diamino-
5,6-1H-dihydropyrimidine-4-one scaffold and the
regioselective syntheses of 2 and 3 using this new
approach.
The total synthesis of TAN-1057A/B has been reported
by the groups of Williams and de Meijere.2 Both syn-
theses were based on the coupling of an a,b-diamino
acid derivative with an S-methylisothiourea derivative
to construct the 2,5-diamino-5,6-dihydro-1H-pyrim-
idine-4-one scaffold. Recently, construction of the dihy-
dropyrimidinone derivative using a tandem reaction
reported by Ganesan et al. is an extension of the above
strategy.3 We also reported a concise construction of
this scaffold that allows for synthesis of TAN-1057A/B
analogs with alteration at the 2-position.4 Unfortu-
nately, the existing approaches are not suitable for
regioselective synthesis of analogs 2 and/or 3. Not
surprisingly, direct methylation of the Boc-protected
heterocycle core 5 (R1=R2=H) (Scheme 1) with
As shown in Scheme 1, we envisioned that both 2 and
3 could be prepared through the coupling of diazoke-
tone 4 with heterocycle core 5 following the strategy
developed by de Meijere group.
Figure 1.
* Corresponding author. E-mail: lianhong xu@gilead.com
–
Scheme 1.
0040-4039/$ - see front matter © 2003 Published by Elsevier Ltd.
doi:10.1016/S0040-4039(03)01381-9

5872
L. Zhang et al. / Tetrahedron Letters 44 (2003) 5871–5873
product 13b was isolated in 50% yield, along with 13%
of regioisomer 6 (R1=H, R2=CH2Ph). The lower
regioselectivity compared to the methyl analog may be
due to the higher steric hindrance of the benzyl group
over the methyl group. Compound 6 (R1=R2=H),
which was not accessible by direct reaction of guanidine
with 7, could be readily obtained by hydrogenation of
compound 13b.
With the key intermediate 13 in hand, we turned our
attention to the methylation of 2-amino functional
group of 13. It was reported that direct methylation of
compounds resembling 13 favored the amide nitrogen
(N-3) over the 2-NH2 nitrogen.7 When compound 13b
was treated with iodomethane and potassium carbonate
in DMF, the regioisomer 14b was indeed formed as a
major product. However, the desired product 15b can
be obtained in high yield by additional treatment of 14b
with potassium carbonate in methanol. Base treatment
of 14b induced the ring-opening of 14b followed by a
ring-closure to yield the thermodynamicaly more stable
isomer.8
Scheme 2.
It was observed2b that N-(tert-butoxycarbonyl)-N-
methyldehydroalanine methyl ester 7 failed to react
with N-guanylurea 9 for the synthesis of heterocycle 10
possibly due to insufficient reactivity of 9 (Scheme 2).
On the other hand, reaction of dehydroalanine 7 with
guanidine 8 (R1=H) proceeded all the way to the
bicycle compound 11 through an ensuing Michael addi-
tion of intermediate 6 (R1=R2=H) with a second
molecule of 7 and subsequent cyclization.
Final steps of syntheses of methylated analogs 2 and 3
are illustrated in Schemes 4 and 5, starting from 13a
and 15b, respectively. Compound 13a was treated with
benzyloxycarbonyl isocyanate to give compound 16.
Similarly, acetylation of 15b afforded 17. TAN-1057
A/B analogs 2 and 3 were thus obtained respectively by
deprotection of the Boc group, coupling of the resulting
amine with side chain 4,2b and hydrogenolysis removal
of the Cbz and benzyl groups.9
We considered that if 1-alkylguanidine 8 (R1=alkyl)
was used, compound 6 should be a stable intermediate
because it could no longer undergo the second round of
Michael addition and subsequent cyclization with 7. In
addition, compound 6 (R2=H) should be formed as a
regioselective major product because Michael addition
of 8 should favor at the alkylated N-1 position due to
the anticipated higher nucliphilicity of the alkylated
nitrogen over the other two nitrogens.
Indeed, as outlined in Scheme 3, when 1-methyl-
guanidine hydrochloride 8 (R1=Me) and compound 7,
prepared from serine methyl ester 12,5 was treated with
potassium carbonate in acetonitrile at room tempera-
ture, desired pyrimidone 13a was isolated in 85% yield.
When 1-benzylguanidine 8 (R1=CH2Ph)6 was used for
the Michael addition–cyclization reaction, desired
Scheme 4.
Scheme 3.
Scheme 5.

L. Zhang et al. / Tetrahedron Letters 44 (2003) 5871–5873
5873
Ono, H. J. Antibiot. 1993, 46, 606; (b) Funabashi, Y.;
Tsubotani, S.; Koyama, K.; Katayama, N.; Harada, S.
Tetrahedron 1993, 49, 13.
2. (a) Yuan, C.; Williams, R. J. Am. Chem. Soc. 1997, 119,
11777; (b) Sokolov, V. V.; Kozhushkov, S. I.; Nikolskaya,
S.; Belov, V. N.; Es-Sayed, M.; Meijere, A. D. Eur. J. Org.
Chem. 1998, 777.
3. Lin, P.; Ganesan, A. Synthesis 2000, 2127.
4. Xu, L.; Zhang, L.; Bryant, C. M.; Kim, C. U. Tetrahedron
Lett. 2003, 44, 2601.
5. Torisawa, Y.; Nakagawa, M.; Hosaka, T.; Tanabe, K.;
Lai, Z.; Ogata, K.; Nakata, T.; Oishi, T.; Hino, T. J. Org.
Chem. 1992, 57, 5741.
Scheme 6.
6. Yong, Y. F.; Kowalski, J. A.; Lipton, M. A. J. Org. Chem.
1997, 62, 1540.
With the availability of 13b, an alternative route for the
synthesis of TAN-1057A/B can be achieved (Scheme 6).
Treatment of 13b with benzyloxycarbonyl isocyanate
provided 18. TAN-1057A/B was then obtained apply-
ing the same protocol as was used for synthesis 2 and 3.
It is worth noting here that TAN-1057A/B is synthe-
sized in 7 steps with 12% overall yield from serine
analog 12. This is an improvement over the existing
routes for synthesis of TAN-1057A/B and the method
developed here is complimentary to the reported routes.
7. Greenhill, J. V.; Ismail, J. M.; Edwards, P. N.; Taylor, P.
J. J. Chem. Soc., Perkin Trans. 2 1985, 1255.
8. Experimental procedures for 13b and 15b:
13b: To a solution of 7 (795 mg, 3.7 mmol) and 1-benzyl-
guanidine TFA salt (3.7 mmol) in isopropanol (15 mL)
was added potassium carbonate (1380 mg, 10 mmol).
After stirred at room temperature for 16 h and at 50°C for
4 h, the reaction mixture was filtered and concentrated
under reduced pressure. The residue was purified by chro-
matography on silica gel (eluted with 10% MeOH/CH2Cl2)
to afford desired product 13b (610 mg, 50%). 1H NMR
(CDCl3) l: 1.4 (two brs (2:1), 9H), 2.8 (brs, 3H), 3.35 (m,
1H), 3.7 (m, 1H), 4.6 (m, 2H), 4.9 (m, 1H), 7.2–7.4 (m,
5H). MS (M+1): 333.
15b: To a solution of 13b (300 mg, 0.9 mmol) in DMF (4
mL) was added iodomethane (155 mg, 1.09 mmol) and
potassium carbonate (280 mg, 2 mmol). The reaction
mixture was stirred at room temperature for 4 h, diluted
with ethyl acetate and filtered. The filtrate was concen-
trated under reduced pressure to give crude product 14b
[1H NMR (CDCl3) l: 1.4 (two brs (2:1), 9H), 2.8 (two brs,
3H), 3.2 (m, 1H), 3.3 (s, 3H), 3.6 (m, 1H), 4.4–4.9 (m, 3H),
7.2–7.4 (m, 5H). MS (M+1): 347]. The crude product 14b
was dissolved in methanol (20 mL) and potassium carbon-
ate (28 mg) was added. The resulting mixture was stirred
at room temperature for 3 h, filtered and concentrated
under reduced pressure. The residue was purified by chro-
matography on silica gel and precipitation from hexane–
ethyl acetate to yield compound 15b (275 mg, 88%). 1H
NMR (CDCl3) l: 1.4 (brs, 9H), 2.8 (brs, 3H), 2.9 (brs,
3H), 3.35 (m, 1H), 3.7 (m, 1H), 4.5 (m, 2H), 5.05 (m, 1H),
7.2–7.4 (m, 5H). MS (M+1): 347.
In summary, based on a Michael addition and subse-
quent cyclization of 1-alkylguanidine 8 with dehy-
droalanine methyl ester 7, a new approach to the
2,5-diamino-5,6-1H-dihydropyrimidine-4-one class of
compounds was developed. TAN-1057A/B and its
methylated analogs 2 and 3 were synthesized in good
yield and with a high degree of regioselectivity using
this new approach. This method allows for rapid prepa-
ration of TAN-1057A/B and its analogs for biological
activity evaluation. More detailed SAR studies of this
class of compounds will be discussed in future
publications.
Acknowledgements
We thank Dr. Ke-Yu Wang for performing NMR
studies on the key intermediates and compounds, and
Dr. Christopher Lee for his assistance in manuscript
preparation.
9. Both analogs
2 and 3 displayed reduced minimal
inhibitory concentration (MIC) against Staphylococci,
data are listed as follows: MIC, S. aureus 29213 (mg/mL):
TAN-1057, 8; analog 2, >256; analog 3, 64.
References
1. (a) Katayama, N.; Fukusumi, S.; Funabashi, Y.; Iwahi, T.;