Synthesis and biochemical evaluation of O-acetyl-ADP-ribose and N-acetyl analogs

Article abstract of DOI:10.1039/b710231c

Synthetic routes for the preparation of O-acetyl-ADP-ribose and two novel non-hydrolyzable analogs containing an N-acetyl are described and shown to interact with the macro domain of histone protein H2A1.1. The Royal Society of Chemistry.

Full text of DOI:10.1039/b710231c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Synthesis and biochemical evaluation of O-acetyl-ADP-ribose and
N-acetyl analogs†
Lindsay R. Comstock and John M. Denu*
Received 5th July 2007, Accepted 21st August 2007
First published as an Advance Article on the web 4th September 2007
DOI: 10.1039/b710231c
Synthetic routes for the preparation of O-acetyl-ADP-ribose
and two novel non-hydrolyzable analogs containing an N-
acetyl are described and shown to interact with the macro
domain of histone protein H2A1.1.
is capable of interacting with the macro domain of H2A1.1. An
atypical histone variant containing an additional 20 kDa domain,
macroH2A, has been implicated in transcriptional regulation and
proliferation.7 Additional findings indicate that OAADPr may
induce the assembly of the yeast Sir complex8 and modulate
cellular responses by regulating the opening of the Ca2+-permeable
long transient receptor potential channel 2 (TRPM2).9
Post-translational modifications of histone proteins, specifically
acetylation, regulate transcription, DNA synthesis and repair.1
Dynamic acetylation is controlled by histone acetyltransferases
(HATs) and histone deacetylases (HDACs). Evaluation of HDACs
is of high interest due to their direct link to cancer and aging.2
To date, three classes of HDACs have been identified. Class I
and II share a conserved catalytic domain and function via a
zinc-mediated hydrolysis to afford the deacetylated substrate and
acetate.3 Class III HDACs, the sirtuin family of histone–protein
deacetylases, play a role in several cellular processes ranging
from silencing gene expression, metabolism and apoptosis.2b,4
Unlike the Class I and II HDACs, the sirtuins require NAD+
and deacetylation is coupled with the formation of a novel
metabolite, O-acetyl-ADP-ribose (OAADPr) (Fig. 1).5 Mecha-
nistic investigations support the 2ꢀ-O-acetyl regioisomer as the
enzyme product and that solution equilibrium via intra-molecular
trans-esterification affords a 1 : 1 mixture (at pH 7.5) of the 2ꢀ- and
3ꢀ-O-acetyl regioisomer.6
The identification of additional targets of OAADPr is cur-
rently restricted by both its inherent instability and the limited
quantity available from enzyme derived products.4b Hydrolysis
occurs at either the acetyl functionality on the ribose sugar
or the pyrophosphate linkage by ARH3 (a poly(ADP-ribose)
glycohydrolase),10 Nudix hydrolases,11 esterases and unidentified
metabolizing enzymes.12 Thus, to effectively evaluate the cellular
function of OAADPr and to serve as biochemical tools to
overcome these limitations, a synthetic pathway to authentic
OAADPr and two non-hydrolyzable analogs have been developed
(Fig. 2). Inthedesignofthenon-hydrolyzableanalogs, substitution
of the O-acetyl moiety with an N-acetyl is anticipated to greatly
stabilize this functionality to both spontaneous and enzyme-
dependent hydrolysis, in addition to preventing acyl migration
that is typically observed with OAADPr. Generation of N-acetyl
analogs at the 2ꢀ- and 3ꢀ-positions of the ribose sugar (2 and 3) will
also prove to be valuable in the identification of the biologically
active isomer for each protein target.
Retrosynthetic analysis of OAADPr and its N-acetyl analogs,
indicated that a coupling between AMP and the functionalized
ribose sugars would afford the desired products (Fig. 2). Thus,
the development of novel synthetic pathways to OAADPr and its
N-acetyl analogs would require the production of three different
phosphorylated sugars containing the appropriate substitution
at the 2- and 3-positions via orthogonal protection strategies.
Although it would have been ideal to develop synthetic routes
containing common intermediates for introducing substitution
on the ribose sugar, literature precedence dictated that different
starting materials were necessary.
OAADPr is hypothesized to be a substrate for other linked
enzymatic processes or to serve as an allosteric regulator or second
messenger. Early efforts to identify the role of OAADPr in vivo
using microinjections indicated that it caused a delay/block in
maturation and cell division in starfish oocytes and blastomeres,
respectively.5b More recent work has demonstrated that OAADPr
Department of Biomolecular Chemistry, University of Wisconsin-Madison,
Madison, WI, 53706, USA. E-mail: jmdenu@wisc.edu; Fax: +1 (608) 262-
5253; Tel: +1 (608)265-1859
† Electronic supplementary information (ESI) available: Complete experi-
mental procedures for all compounds described are provided. Details and
results of HPLC stability studies and macroH2A binding experimentals
are presented. See DOI: 10.1039/b710231c
Fig. 1 Sirtuin-mediated deacetylation.
The Royal Society of Chemistry 2007
This journal is
Org. Biomol. Chem., 2007, 5, 3087–3091 | 3087
©

View Article Online
Previous syntheses of nitrogen-bearing ribose analogs substi-
tuted at the 2-position have traditionally relied upon azide instal-
lation via an activated uridine and require subsequent hydrolysis
of the uridine base.19 Due to difficulties in removing the uridine
base from the substituted ribose, an alternative synthetic pathway
was developed and utilized chemistry previously described by
Oppenheimer and Bobek.20 Instead of directly installing the
nitrogen at the desired 2-position of the ribose, the alternative
methodology involved its incorporation at the 3-position of a
substituted glucofuranose and relocation to the 2-position via a
NaIO4-mediated oxidative rearrangement.
As depicted in Scheme 2A, conversion of commercially available
1,2:5,6-di-O-isopropylidene-a-D-glucofuranose to the azide sugar
9 was accomplished as previously described.21 Following selective
benzylation of the 6-position22 (10), acid hydrolysis of the 1,2-O-
isopropylidene afforded the requisite sugar for the rearrangement.
Oxidation of the 1,2-diol with NaIO4 yielded the correspond-
ing di-aldehyde, which underwent spontaneous ring closure to
generate 2-azido ribose sugar 11.20 Silyl ether protection of the
resulting 1,3-diol afforded 12 as a 1 : 6 mixture of a- and b-
anomers in 60% yield.23 Following reduction of the azide using
Staudinger conditions24 and subsequent acylation of the resulting
primary amine,25 benzyl deprotection was affected using standard
hydrogenolysis at 50 psi18 to afford 13 in 71% yield over three steps.
At this stage, the a- and b-anomers were separable using flash
chromatography. Direct phosphorylation of either anomer using
Fig. 2 Retrosynthetic analysis of OAADPr and its N-acetyl derivatives.
Generation of the functionalized O-acetyl ribose is depicted in
Scheme 1 and begins with the previously described 5-O-benzyl-
1,2-O-isopropylidene-a-D-xylofuranose (4).13 Inversion of the 3-
alcohol was necessary to afford a ribofuranose and accomplished
via a two-step process involving a PDC oxidation followed by
subsequent reduction with NaBH4. Benzylation of the resulting
alcohol provided sugar 5 in 52% yield over 3 steps.13b Selective
protection at the 1-position was initiated with isopropylidene
cleavage with Dowex-H+, followed by a dibutyl stannylene-based
benzylation to afford the tri-benzylated ribose, 6, in 74% yield.14
Acylation of the resulting alcohol,15 followed by a selective benzyl
deprotection,16 rendered 7 in 48% yield. Direct phosphorylation
using POCl3 in THF,17 followed by a hydrogenolysis,18 afforded
the TEA salt of sugar 8 (trans-esterification mixture of 2ꢀ- and
3ꢀ-O-acetyl regioisomers).
17
POCl3 of 13 resulted in activated sugar 14 (as the triethylamine
salt) for subsequent coupling chemistry in 93% yield.
Installation of the N-acetyl at the 3-position of the ribose is
depicted in Scheme 2B and bears a resemblance to chemistry
developed for substitution at the 2-position. Beginning with 4,
azide installation occurred via a SN2 displacement of the triflate
26
with LiN3 to provide sugar 15 in 38% yield. Upon subjection to
Dowex-H+19c to generate the 1,2-diol, subsequent reprotection as
the silyl ethers22 rendered 16 (b anomer only) in a yield of 64%.
Following Staudinger reduction of the azide24 and acylation of the
amine,25 benzyl deprotection produced 17 in 84% yield over three
steps.18 Phosphorylation17 resulted in the activated sugar 18 (as the
triethylamine salt).
To complete the synthesis of OAADPr and its N-acetyl analogs
(1–3), the chemistry condensing the adenosine and the function-
alized ribose sugar was investigated. Preliminary condensations
of sugar phosphates 14 and 18 with AMP failed to yield the
anticipated TBS-protected products. It was initially hypothesized
that TBS functionalities were interfering sterically with the desired
chemistry. Thus, silyl ether deprotection of 14 and 18 was
carried out utilizing Dowex-50 (H+) in high yeilds.27 The resulting
phosphomonoesters 19 and 20 (as the TEA salt), along with O-
acetyl phosphate 8, were successfully coupled directly to AMP
activated as the diphenyl pyrophosphate (21) using procedures
established by Michelson (Scheme 3).28 The resulting products (1,
2, and 3) were first purified by anion-exchange chromatography
(DEAE cellulose) to remove the resulting by-product diadenosine
5ꢀ-pyrophosphate (AppA) and followed by preparative HPLC to
afford authentic OAADPr and its two N-acetyl analogs. The
yields of the coupling reactions were low (typically 5–20%), but
comparable to similar couplings previously carried out.29
Scheme 1 Synthesis of 2-O-acetyl sugar. (a) PDC, AcOH, CH2Cl2 (65%);
(b) NaBH4, MeOH; (c) BnBr, NaH, DMF (80%, 2 steps); (d) Dowex-H+,
dioxane; (e) Bu2SnO, MeOH; (f) BnBr, K2CO3, DMF (74%, 3 steps); (g)
Ac2O, pyridine (97%); (h) H2 (50 psi), 10% Pd/C, MeOH–AcOH, 0.08%
pyridine (40–50%, 5–15% recovered); (i) POCl3, TEA, THF (85%); (j) H2
(50 psi), 10% Pd/C, EtOH (88%).
To validate the installation of an N-acetyl as a non-hydrolyzable
substitution, the stabilities of both 2ꢀ- and 3ꢀ-NAADPr were
3088 | Org. Biomol. Chem., 2007, 5, 3087–3091
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Scheme 2 Synthesis of 2-N- and 3-N-acetyl sugars. (a) 70% AcOH; (b) Bu2SnO, toluene, then BnBr and Bu4NBr (65%, 2 steps); (c) Dowex-H+, dioxane;
(d) NaIO4, NaHCO3 dioxane; (e) TBSCl, imidazole, DMF (12: 60%, 3 steps; 16: 64%, 2 steps); (f) PPh3, aqueous NH3, pyridine; (g) Ac2O, pyridine,
CH2Cl2; (h) H2 (50 psi), 10% Pd/C, EtOH (13: 71%, 3 steps; 17: 84%, 3 steps); (i) POCl3, TEA, THF (14: 93%; 18: 90%); (j) Tf2O, pyridine, CH2Cl2 (87%);
(k) LiN3, DMF (44%).
constants and binding enthalpies produced from the binding
curves indicate comparable values to those previously determined
for ADPr and OAADPr (panel A).7 Both 2ꢀ- and 3ꢀ-NAADPr
yielded simililar Kd values (16 versus 19 lM, respectively),
although there was a small four-fold reduction in values when
compared to OAADPr (panel B). Similar DH values (∼13–15 kcal
mol−1) among the four suggests a similar mode of binding, likely
within the same ligand pocket. Most notable, the nearly identical
Kd values of 2ꢀ- and 3ꢀ-NAADPr indicate that macroH2A1.1
does not discriminate between the two regioisomers. Although the
NAADPr analogs are capable of interacting with macroH2A1.1 in
a manner that mimics the natural ligand, further evaluation of the
ability of the NAADPr analogs to bind competitively with other
OAADPr-binding proteins/enzymes will be addressed in future
studies.
Molecular modeling was performed to evaluate the interactions
Scheme 3 Phosphate couplings to AMP. (a) Dowex-H+, ACN (19: 97%;
of macroH2A1.1 with the N-acetyl analogs. Utilizing the crystal
20: 94%); (b) 8, 19, or 20, DMF, pyridine (5–20%).
structure previously determined for ADPr bound to Af152132 and
macroH2A1.1,7 both 2ꢀ- and 3ꢀ-NAADPr were modeled into the
binding pocket.33 As shown in Fig. 4, the pocket is capable of
evaluated at physiological temperature and pH. Results previously
accommodating both 2ꢀ- (panel A) and 3ꢀ-NAADPr (panel B)
and supports the observed similar dissociation constants for both
analogs. Although no obvious structural differences were observed
for macroH2A1.1 when modeled with NAADPr compared to that
obtained with OAADPr (data not shown), it is hypothesized that
the observed four-fold reduction in binding for both 2ꢀ- and 3ꢀ-
NAADPr may be due to minor alterations in the hydrogen bond
network.34 The substitution of the O-acetyl moiety with an N-
acetyl reverses the nature of the hydrogen bond (from an acceptor
to a donor).
obtained with OAADPr indicated that approximately 18% was
hydrolyzed spontaneously to ADPr within 3 h.30 HPLC analysis
of both 2 and 3 using the exact conditions indicated negligible
decomposition (<1%) over 3 days and confirmed the stability of
the acetyl functionality to spontaneous hydrolysis.31
Upon completing the synthesis of the N-acetyl analogs and
confirming their stability, we sought to validate the ability of both
2ꢀ- and 3ꢀ-NAADPr to mimic OAADPr. Among several reported
protein targets, recent work demonstrated in vitro that OAADPr
binds to macroH2A1.1 histone protein.7 To establish NAADPr
binding to macroH2A1.1, isothermal titration calorimetry (ITC)
was performed and the titration of ADPr, OAADPr, and the N-
acetyl analogs individually with ligand-free macroH2A1.1 resulted
in the binding curves shown in Fig. 3. Resulting dissociation
In summary, the synthetic generation of OAADPr and two
non-hydrolyzable analogs containing an N-acetyl is reported.
Subsequent experiments indicated that both 2ꢀ- and 3ꢀ-NAADPr
are mimics of OAADPr in binding with macroH2A1.1 and
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3087–3091 | 3089
©

View Article Online
Acknowledgements
This work was supported by grants from the NIH
(R01GM065386) and a post-doctoral fellowship (LRC) from the
American Heart Association (0625727Z). We thank Dr Alvan
Hengge and Dr James Slama for helpful discussions and S. Slauson
for work in preparing the O-acetyl phosphate sugar. We also thank
A. G. Ladurner for providing the macroH2A1.1 plasmid, L. Tong
for aiding in the expression of macroH2A1.1 and performing ITC,
and B. Smith for performing the SYBYL modeling.
Notes and references
1 S. L. Berger, Curr. Opin. Genet. Dev., 2002, 12, 142.
2 (a) P. A. Marks and M. Dokmanovic, Expert Opin. Investig. Drugs,
2005, 14, 1497; (b) S. Hekimi and L. Guarente, Science, 2003, 299,
1351.
3 C. M. Grozinger and S. L. Screiber, Chem. Biol., 2002, 9, 3.
4 (a) J. M. Denu, Curr. Opin. Chem. Biol., 2005, 9, 431; (b) D. Moazed,
Curr. Opin. Cell Biol., 2001, 13, 232.
5 (a) M. D. Jackson and J. M. Denu, J. Biol. Chem., 2002, 277, 18535;
(b) M. T. Borra, F. J. O’Neill, M. D. Jackson, B. Marshall, E. Verdin,
K. R. Foltz and J. M. Denu, J. Biol. Chem., 2002, 277, 12632.
6 (a) A. A. Sauve, I. Celic, J. Avalos, H. Deng, J. D. Boeke and V. L.
Schramm, Biochemistry, 2001, 40, 15456; (b) B. C. Smith and J. M.
Denu, Biochemistry, 2006, 45, 272.
7 G. Kustatscher, M. Hothorn, C. Pugieux, K. Scheffzek and A. G.
Ladurner, Nat. Struct. Mol. Biol., 2005, 12, 624.
8 G. G. Liou, J. C. Tanny, R. G. Kruger, T. Walz and D. Moazed, Cell,
2005, 121, 515.
9 O. Grubisha, L. A. Rafty, C. L. Takanishi, X. Xu, L. Tong, L. A. L.
Perraud, A. M. Scharenberg and J. M. Denu, J. Biol. Chem., 2006, 281,
14057.
10 T. Ono, A. Kasamatsu, S. Oka and J. Moss, Proc. Natl. Acad. Sci.
U. S. A., 2006, 103, 16687.
11 A. S. Mildvan, Z. Xia, H. F. Azurmendi, V. Saraswat, P. M. Legler,
M. A. Massiah, S. B. Gabelli, M. A. Bianchet, L. W. Kang and L. M.
Amzel, Arch. Biochem. Biophys., 2005, 433, 129.
12 L. A. Rafty, M. T. Schmidt, A. L. Perraud, A. M. Scharenberg and
J. M. Denu, J. Biol. Chem., 2002, 277, 47114.
13 (a) R. D. Marwood, S. Shuto, D. J. Jenkins and B. V. L. Potter, Chem.
Commun., 2000, 219; (b) P. B. Alper, M. Hendrix, P. Sears and C. H.
Wong, J. Am. Chem. Soc., 1998, 120, 1965.
14 T. L. Su, R. S. Klein and J. J. Fox, J. Org. Chem., 1982, 47, 1506.
15 J. Ning and F. Kong, Carbohydr. Res., 1997, 300, 355.
16 B. E. Maryanoff, A. B. Reitz, G. F. Tutwiler, S. J. Benkovic, P. A.
Benkovic and S. J. Pilkis, J. Am. Chem. Soc., 1984, 106, 7851.
17 J. T. Slama, N. Aboul-Ela, D. M. Goli, B. V. Cheesman, A. M. Simmons
and M. K. Jacobson, J. Med. Chem., 1995, 38, 389.
18 C. H. Heathcock and R. Ratcliffe, J. Am. Chem. Soc., 1971, 93,
1746.
19 (a) G. P. Kirschenheuter, Y. Zhai and W. A. Pieken, Tetrahedron Lett.,
1994, 35, 8517; (b) J. B. Hobbs and F. Eckstein, J. Org. Chem., 1977,
42, 714.
20 (a) P. R. Sleath, A. L. Handlon and N. J. Oppenheimer, J. Org. Chem.,
1991, 56, 3608; (b) M. Bobek and V. Martin, Tetrahedron Lett., 1978,
22, 1919; (c) M. Bobek, Carbohydr. Res., 1979, 70, 263.
21 M. P. Watterson, L. Pickering, M. D. Smith, S. J. Hudson, P. R.
Marsh, J. E. Mordaunt, D. J. Watkin, C. J. Newman and G. W. Fleet,
Tetrahedron: Asymmetry, 1999, 10, 1855.
22 (a) A. Lubineau and Y. Queneau, Tetrahedron, 1989, 45, 6697; (b) R. V.
Stick, K. A. Stubbs and A. G. Watts, Aust. J. Chem., 2004, 57, 779.
23 E. J. Corey and A. Venkateswarlu, J. Am. Chem. Soc., 1972, 94,
6191.
Fig. 3 ITC curves with macroH2A1.1 for (A) ADPr (open circles)
and OAADPr (open triangles) and (B) 2ꢀ-NAADPr (open circles) and
3ꢀ-NAADPr (closed triangles). Experiments were performed at 25 ◦C in
50 mM KH2PO4, pH 6.5 and 1 mM DTT. Values previously obtained for
7
OAADPr are 2.6 lM and −17.8 kcal mol−1
.
24 M. R. Webb, G. P. Reid, V. R. Munasinghe and J. E. Corrie,
Biochemistry, 2004, 43, 14463.
25 H. B. Mereyala, M. Baseeruddin and S. R. Koduru, Tetrahedron:
Asymmetry, 2004, 15, 3457.
26 (a) O. Botta, E. Moyroud, C. Lobato and P. Strazewski, Tetrahedron,
1998, 54, 13529; (b) A. V. Azhayev, A. M. Ozols, A. S. Bushnev, N. B.
Dyatkina, S. V. Kochetova, L. S. Victorova, M. K. Kikhanova, A. A.
Krayevsky and B. P. Gottokh, Nucleic Acids Res., 1979, 6, 625.
demonstrate that they will be valuable chemical tools for the future
evaluation of the physiological role of OAADPr. Additionally,
these analogs are deemed useful for future studies to uncover
unidentified targets of OAADPr.
3090 | Org. Biomol. Chem., 2007, 5, 3087–3091
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Fig. 4 Molecular modeling of (A) 2ꢀ-NAADPr and (B) 3ꢀ-NAADPr with macroH2A1.1 (PDB accession code 2FXK). Color coding: green
(macroH2A1.1), grey (carbon), oxygen (red), nitrogen (blue), phosphorus (orange).
27 E. J. Corey, J. W. Ponder and P. Ulrich, Tetrahedron Lett., 1980,
21, 137.
32 G. I. Karras, G. Kustatscher, H. R. Buhecha, M. D. Allen, C. Pugieux,
F. Sait, M. Bycroft and A. G. Ladurner, EMBO J., 2005, 24, 1911.
33 The structure of macroH2A1.1 was not co-crystallized with OAADPr.
Thus, the crystal structure of ADPr bound to the Af1521 macro domain
(PDB accession code 2BFQ) was overlayed with macroH2A1.1 (carbon
backbone was matched based upon homology) and the resulting
structure of ADPr was utilized as a starting point for modeling the N-
acetyl analogs into macroH2A1.1. The structure of ADPr was modified
in SYBYL, torsionally constrained about the new bonds, and energies
minimized across the new structure. PyMol was utilized to produce the
images in Fig. 4.
28 (a) A. M. Michelson, Biochim. Biophys. Acta, 1964, 91, 1; (b) Q. F.
Ma, M. A. Reynolds and G. L. Kenyon, Bioorg. Chem., 1989, 17, 194.
Note: additional methodologies to activate AMP have been described
to perform such couplings. See ref. 17 and; D. W. Koh, D. L. Coyle,
N. Mehta, S. Ramsinghani, H. Kim, J. T. Slama and M. K. Jacobson,
J. Med. Chem., 2003, 46, 4322.
29 (a) S. M. Graham, D. J. Macaya, R. N. Sengupta and K. B. Turner, Org.
Lett., 2004, 6, 233; (b) T. Fukui and K. Tanizawa, Methods Enzymol.,
1997, 280, 41.
30 M. T. Borra, F. J. O’Neill, M. D. Jackson, B. Marhsall, E. Verdin, K. R.
Foltz and J. M. Denu, J. Biol. Chem., 2002, 277, 12632.
31 See ESI for experimental details and HPLC traces.
34 Previous studies of the Af1521 macro domain bound to ADPr indicated
that mutation of Asn 34 to Ala reduced binding affinity three-fold. See
ref. 32.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3087–3091 | 3091
©