Practical synthesis of cyclic bis(3′-5′)diadenylic acid (c-di-AMP)

Article abstract of DOI:10.1246/cl.2011.1113

Cyclic bis(3′-5′)diadenylic acid (c-di-AMP) (Chart 1), recently identified as a second messenger monitoring DNA integrity during sporulation in the soil bacterium Bacillus subtilis, was synthesized on an 80 μmol scale by a combination of the phosphoramidi

Full text of DOI:10.1246/cl.2011.1113

doi:10.1246/cl.2011.1113
Published on the web September 23, 2011
1113
Practical Synthesis of Cyclic Bis(3¤-5¤)diadenylic Acid (c-di-AMP)
Noritaka Suzuki,1 Kin-ichi Oyama,2 and Masaki Tsukamoto*1
1Graduate School of Information Science, Nagoya University, Chikusa-ku, Nagoya, Aichi 464-8601
2Chemical Instrumentation Facility, Research Center for Materials Science, Nagoya University,
Chikusa-ku, Nagoya, Aichi 464-8602
(Received June 20, 2011; CL-110515; E-mail: tsukamoto@is.nagoya-u.ac.jp)
Cyclic bis(3¤-5¤)diadenylic acid (c-di-AMP) (Chart 1),
recently identified as a second messenger monitoring DNA
integrity during sporulation in the soil bacterium Bacillus
subtilis, was synthesized on an 80 ¯mol scale by a combination
of the phosphoramidite and phosphotriester methods using a
commercially available adenosine phosphoramidite as starting
material. An artificial analog 2¤-bis(tert-butyldimethylsilyl)-c-di-
AMP was also obtained by our procedure.
Although c-di-AMP was suggested to be formed as a side
product in some reactions12-14 such as homopolymerization
of adenosine 3¤-monophosphate,13,14 reports on target-oriented
synthesis have been limited. So far, c-di-AMP has been
synthesized by the phosphotriester method,15,16 and by the
construction of a cyclic sugar backbone and subsequent
introduction of the protected adenine.17 However, these methods
consist of multistep reactions, which cannot provide a sufficient
amount of c-di-AMP. We therefore synthesized c-di-AMP by a
method employing phosphoramidite and phosphotriester ap-
proaches, which had already been applied to a large-scale
synthesis of c-di-GMP by Hyodo and Hayakawa.18 In addition,
we identified and characterized the hydrophobic analog 2¤-O-
bis(tert-butyldimethylsilyl) (TBDMS)-c-di-AMP (5) (Scheme 1),
which it is hoped will promote the biological activities of
c-di-AMP, based on the finding that 2¤-O-bis-TBDMS-c-di-
GMP showed a stronger inhibitory effect than c-di-GMP in
Recently, cyclic bis(3¤-5¤)diadenylic acid (c-di-AMP) was
discovered from the soil bacterium Bacillus subtilis1,2 and was
identified as a second messenger monitoring DNA integrity
during sporulation.3 B. subtilis possesses DNA integrity scan-
ning protein (DisA), which regulates generation of c-di-AMP: at
the onset of sporulation, DisA produces c-di-AMP from ATP,
whereas generation of c-di-AMP is stopped with interruption
of sporulation when DisA detects branched DNA. DisA has a
diadenylate cyclase (DAC) domain able to produce c-di-AMP.
Because the DAC domain is widely present in bacteria and
archaea even in some nonsporulating species, c-di-AMP might
act as a second messenger in various ways. For example, c-di-
AMP induces a type I interferon.4 Moreover, c-di-AMP is
structurally related to cyclic bis(3¤-5¤)diguanylic acid (c-di-
GMP),5 which controls cellulose synthesis,5 biofilm formation,
motility, and virulence factor production in a variety of
bacteria.6-8 Thus, comparison of the biological activities of c-
di-AMP with those of c-di-GMP is a subject of interest, and has
been performed to some degree in studies of riboswitch
recognition9 and adjuvant activities.10 To investigate the
unknown biological activities of c-di-AMP, a sufficient amount
(ideally, 10 ¯mol for a series of experiments) of this molecule is
necessary. However, the amount of c-di-AMP obtained from
bacteria is limited. Furthermore, c-di-AMP is commercially
available in units of 0.1 ¯mol, although, it is fairly expensive.11
Therefore, it is very important to establish a practical synthesis
of c-di-AMP.
Scheme 1. Reagents and conditions: a) 1) allyl alcohol, imidazo-
lium perchlorate (IMP), molecular sieves 3 ¡ (MS 3 ¡), CH3CN, rt, 2)
t-C4H9OOH/toluene, rt; b) CHCl2COOH, CH2Cl2, 0 °C; c) 1) 1, IMP,
MS 3 ¡, CH3CN, rt, 2) t-C4H9OOH/toluene, rt; d) CHCl2COOH,
CH2Cl2, 0 °C; e) NaI, acetone, reflux; f) 2,4,6-triisopropylbenzene-
sulfonyl chloride, N-methylimidazole, 28 °C; g) concd aq. NH3-
CH3OH (1:1 v/v), 50 °C; h) (C2H5)3N¢3HF, rt.
Chart 1.
Chem. Lett. 2011, 40, 1113-1114
© 2011 The Chemical Society of Japan
www.csj.jp/journals/chem-lett/

1114
shows the HPLC profile of c-di-AMP obtained after purification
by reverse-phase preparative HPLC. The retention time of
synthetic c-di-AMP was found to be 11.3 min, which was
identical to that of the authentic sample (Figure S1 in Support-
ing Information (SI)26). Moreover, the synthetic and authentic
samples showed the same fragmentation patterns in ESI tandem
mass measurements (Figure S2 in SI26).
In summary, we synthesized c-di-AMP and its hydrophobic
analog 2¤-O-bis-TBDMS-c-di-AMP from the adenosine phos-
phoramidite in 40% and 44% overall yields, respectively. Our
method can provide sufficient amounts of c-di-AMP for
implementation of a variety of biological activities, and the
results will be reported in due course.
This study was supported by a Grant-in-Aid for Scientific
Research (No. 22750148) from the Japan Society for the
Promotion of Science (JSPS). We are grateful to Professor Y.
Hayakawa (Aichi Institute of Technology) and Dr. M. Hyodo
(Hokkaido University) for their encouragement and helpful
input.
References and Notes
Figure 1. Reverse-phase HPLC profiles of (a) a mixture of 5 and 6,
and (b) purified c-di-AMP. For details of HPLC conditions, see
Figure S1.26
1
G. Witte, S. Hartung, K. Büttner, K.-P. Hopfner, Mol. Cell 2008, 30,
167.
2
3
U. Römling, Sci. Signaling 2008, 1, pe39.
Y. Oppenheimer-Shaanan, E. Wexselblatt, J. Katzhendler, E. Yavin,
S. Ben-Yehuda, EMBO Rep. 2011, 12, 594.
Anaplasma phagocytophilum infection of human myelocytic
leukemia HL-60 cells.19
4
5
J. J. Woodward, A. T. Iavarone, D. A. Portnoy, Science 2010, 328,
1703.
P. Ross, H. Weinhouse, Y. Aloni, D. Michaeli, P. Weinberger-Ohana,
R. Mayer, S. Braun, E. de Vroom, G. A. van der Marel, J. H.
van Boom, M. Benziman, Nature 1987, 325, 279.
M. Hyodo, Y. Hayakawa, in Modified Nucleosides: In Biochemistry,
Biotechnology and Medicine, ed. by P. Herdewijn, Wiley-VCH,
Weinheim, 2008, Chap. 13, pp. 343-363.
We synthesized c-di-AMP according to Scheme 1 by use of
the commercially available adenosine cyanoethyl (CE) phos-
phoramidite 1 as starting material. The amidite 1 was condensed
with allyl alcohol by the aid of imidazolium perchlorate (IMP)20
in the presence of molecular sieves 3 ¡ (MS 3 ¡)21 in aceto-
nitrile. The resulting phosphite triester was oxidized with tert-
butyl hydroperoxide22,23 to give the corresponding phospho-
triester, which was converted into the nucleotide 2 by depro-
tection of the 5¤-O-p,p¤-dimethoxytrityl (DMTr) group under
acidic conditions. The overall yield of 2 from 1 was 96%. This
three-step procedure consisting of condensation, oxidation, and
detritylation was safely applied to a combination of the amidite
1 and the nucleotide 2 to give the protected linear dimer 3 in
82% yield as shown in steps c and d. Thereafter, the allyl group
on the 3¤-terminal phosphate of 3 was deprotected by sodium
iodide.18 The linear dinucleotide 3¤-phosphodiester thus ob-
tained was cyclized by a mixture of 2,4,6-triisopropylbenzene-
sulfonyl chloride and N-methylimidazole in THF under high-
dilution conditions (substrate concentration: ca. 6 mM) to give
the fully protected c-di-AMP 4 in 57% overall yield from 3.24
This yield was in the range of reported values (40-87%) in the
synthesis of related compounds.18,25
Subsequently, benzoyl (Bz) and cyanoethyl (CE) protecting
groups of 4 were removed by a mixture of concentrated aqueous
ammonia and methanol to afford the bis TBDMS-protected c-di-
AMP 5 and the mono TBDMS-protected c-di-AMP 6 in a ratio
of 96:4 in 97% yield. Figure 1a shows the reverse-phase HPLC
profile of the crude sample, indicating that partial deprotection
of the TBDMS group occurs in step g (Scheme 1).
Finally, removal of the TBDMS groups of 5 and 6 was
conducted in an 89 ¯mol scale by treatment of triethylamine
trihydrofluoride to give c-di-AMP in 91% yield. Figure 1b
6
7
8
9
R. Hengge, Nat. Rev. Microbiol. 2009, 7, 263.
H. Yan, W. Chen, Chem. Soc. Rev. 2010, 39, 2914.
K. D. Smith, S. V. Lipchock, T. D. Ames, J. Wang, R. R. Breaker,
S. A. Strobel, Nat. Struct. Mol. Biol. 2009, 16, 1218.
10 T. Ebensen, R. Libanova, K. Schulze, T. Yevsa, M. Morr, C. A.
Guzmán, Vaccine, 2011, 29, 5210.
11 c-di-AMP is available from BIOLOG.
12 W. H. Cook, D. Lipkin, R. Markham, J. Am. Chem. Soc. 1957, 79,
3607.
13 Y. Lapidot, H. G. Khorana, J. Am. Chem. Soc. 1963, 85, 3857.
14 E. Ohtsuka, H. Tsuji, M. Ikehara, Chem. Pharm. Bull. 1974, 22, 1022.
15 H. C.-Y. Jenny, D. Don, J. Oger, Nucleosides Nucleotides 1985, 4, 377.
16 P. Ross, R. Mayer, H. Weinhouse, D. Amikam, Y. Huggirat, M.
Benziman, E. de Vroom, A. Fidder, P. de Paus, L. A. J. M. Sliedregt,
G. A. van der Marel, J. H. van Boom, J. Biol. Chem. 1990, 265, 18933.
17 N. Amiot, K. Heintz, B. Giese, Synthesis 2006, 4230.
18 M. Hyodo, Y. Hayakawa, Bull. Chem. Soc. Jpn. 2004, 77, 2089.
19 T.-H. Lai, Y. Kumagai, M. Hyodo, Y. Hayakawa, Y. Rikihisa,
J. Bacteriol. 2009, 191, 693.
20 Y. Hayakawa, R. Kawai, A. Hirata, J.-i. Sugimoto, M. Kataoka, A.
Sakakura, M. Hirose, R. Noyori, J. Am. Chem. Soc. 2001, 123, 8165.
21 Y. Hayakawa, A. Hirata, J.-i. Sugimoto, R. Kawai, A. Sakakura, M.
Kataoka, Tetrahedron 2001, 57, 8823.
22 Y. Hayakawa, M. Uchiyama, R. Noyori, Tetrahedron Lett. 1986, 27,
4191.
23 J. G. Hill, B. E. Rossiter, K. B. Sharpless, J. Org. Chem. 1983, 48,
3607.
24 V. A. Efimov, S. V. Reverdatto, O. G. Chakhmakcheva, Tetrahedron
Lett. 1982, 23, 961.
25 M. Hyodo, Y. Sato, Y. Hayakawa, Tetrahedron 2006, 62, 3089.
26 Supporting Information is available electronically on the CSJ-Journal
Web site, http://www.csj.jp/journals/chem-lett/index.html.
Chem. Lett. 2011, 40, 1113-1114
© 2011 The Chemical Society of Japan
www.csj.jp/journals/chem-lett/