Guanophostin A: Synthesis and evaluation of a high affinity agonist of the D-myo-inositol 1,4,5-trisphosphate receptor

Article abstract of DOI:10.1039/b517911d

Guanophostin A, the guanosine counterpart of the inositol 1,4,5-trisphosphate receptor agonist adenophostin A, has been synthesized and is the first synthetic adenophostin A-like analogue to be equipotent to its parent in stimulating intracellular Ca

Full text of DOI:10.1039/b517911d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Guanophostin A: Synthesis and evaluation of a high affinity agonist of
the D-myo-inositol 1,4,5-trisphosphate receptor
Kana M. Sureshan,a Melanie Trusselle,a Stephen C. Tovey,b Colin W. Taylorb and Barry V. L. Potter*a
Received (in Cambridge, UK) 16th December 2005, Accepted 15th March 2006
First published as an Advance Article on the web 31st March 2006
DOI: 10.1039/b517911d
various analogues of adenophostin and study their structure–
activity relationship (SAR).4 However, none of the synthetic
analogs of adenophostins has achieved the potency of adenophos-
tin itself. Herein, we report the first synthesis and biological
characterisation of ‘guanophostin A’ in which the adenine of
adenophostin A is replaced by guanine, the other naturally
occurring ubiquitous nucleic acid and nucleotide purine base.
Although significant information has been gained from SAR
studies on adenophostin analogs, some of the basic questions
regarding the nature of the molecular interaction with the IP3R
remain unanswered. From SAR studies with synthetic analogues
with and without a purine ring, it was evident that the presence of
the adenine ring (or a surrogate) is crucial for enhanced affinity.5
Two hypotheses have been proposed on the role of adenine to
explain the enhanced potency of adenophostin over Ins(1,4,5)P3.
Either the adenosine moiety disposes the 29-phosphate in a special
spatial arrangement to strengthen the binding interactions with the
receptor as an Ins(1,4,5)P3 1-phosphate group surrogate (indirect
role),6 or the adenine moiety itself is involved in supplementary
binding interactions with a nearby region of the binding site (direct
role).7 We have proposed a model for the interaction of 2 with the
IP3R binding core that involves a cation-p interaction of the
adenine with an arginine residue of the IP3R. To pursue these
possibilities further, purine-modified adenophostin analogs are
necessary. Although some purine-modified (predominantly at the
C-6 position) adenophostin analogs have been synthesized, the
important 2-position of this base is essentially unexplored. Our
adenophostin-docked model of the IP3R binding core also shows
the carboxylate anion of Glu505 near the 2-position of the adenine
base.4a Hence an H-bond donor or cationic residue at the
2-position may interact constructively with this carboxylate group.
To explore this we chose to synthesise guanophostin A.
Guanophostin A, the guanosine counterpart of the inositol
1,4,5-trisphosphate receptor agonist adenophostin A, has been
synthesized and is the first synthetic adenophostin A-like
analogue to be equipotent to its parent in stimulating
intracellular Ca2+ release; its nucleotide moiety is proposed to
interact with the receptor binding core by guanine base cation-p
stacking with Arg504 and hydrogen bonding with Glu505 and
interaction of the ribosyl 29-phosphate group with the helix-
dipole of a6.
Inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] receptors [IP3Rs] are
Ca2+ channels located on the endoplasmic reticulum, the major
intracellular Ca2+ store. Ins(1,4,5)P3 (1, see Fig. 1) is a second
messenger produced by the action of phospholipase C on
phosphatidylinositol 4,5-bisphosphate in response to various
extracellular signals. Ins(1,4,5)P3 then opens the intrinsic calcium
channel of the IP3R and the resultant Ca2+ efflux causes an
increase in cytosolic [Ca2+], which regulates many cellular events.1
To study the interaction of 1 with its receptor and to understand
structure–activity relationships, many synthetic analogues of
Ins(1,4,5)P3 have been synthesized. None of these monomeric
synthetic analogues has proven to be more potent than the natural
ligand. In 1993, Takahashi et al. isolated two unusual agonists of
IP3Rs, adenophostin A (2) and adenophostin B (3) from culture
broths of Penicillium brevicompactum.2 Both were shown to bind
to IP3Rs with much greater affinity than Ins(1,4,5)P3 and to be 10–
100 times more potent than the natural ligand in evoking Ca2+
release via IP3R.3 This has stimulated many attempts to synthesize
Ideally, Vorbru¨ggen condensation of a silylated purine with a
properly protected disaccharide followed by deprotection of
hydroxyl groups (to be later phosphorylated), phosphorylation
and final deprotection are the important steps to be considered for
the synthesis of adenophostin analogues. However, Vorbru¨ggen
condensation of guanine or its derivatives with acylated sugars is
known to be problematic8 due to the formation of N-7 and N-9
regioisomers even under thermodynamic control and hence is not
preferred synthetically. We therefore decided to use 2-amino-6-
chloropurine (7) which is known9 to undergo regiospecific coupling
under Vorbru¨ggen conditions (Scheme 1). Moreover, a variety of
substituted nucleosides could potentially be synthesized by
manipulating the 6-chlorine motif after the glycosylation.
Vorbru¨ggen condensation of 2-amino-6-chloropurine (7) and
disaccharide 6,4a synthesized from penta-O-acetyl-D-glucose (4)
Fig. 1 Structure of Ins(1,4,5)P3 (1), adenophostin
adenophostin B (3).
A (2) and
aWolfson Laboratory of Medicinal Chemistry, Department of Pharmacy
and Pharmacology, University of Bath, Claverton Down, Bath, UK
BA2 7AY. E-mail: b.v.l.potter@bath.ac.uk; Fax: +44 1225 386114;
Tel: +44 1225 386639
bDepartment of Pharmacology, Tennis Court Road, University of
Cambridge, Cambridge, UK CB2 1PD
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2015–2017 | 2015

View Article Online
can compete with hydroxyl groups on the sugar for phosphoryla-
tion, generation of an O-protected guanine moiety is desirable and
we chose an O-benzyl-guanine derivative to reduce the number of
steps for deprotection. Thus, nucleoside 8 on treatment with in
situ-generated sodium benzoxide [benzyl alcohol (6 eq.) and NaH
(5 eq.)] in DMF at 80 uC for 3 h gave a mixture of 12 (43%) and
de-glucosylated 13 (41%) (Scheme 2). Use of only two equivalents
of sodium hydride and BnOH as solvent resulted in the exclusive
formation of 12. In both cases, the formation of 2-N-benzylated
product was not observed.13 Triol 12 on chemoselective phosphi-
tylation followed by in situ oxidation gave the fully protected
trisphosphate 14 in 82% yield. The benzyl protecting groups were
removed by transfer hydrogenolysis and purification by ion
exchange column chromatography provided guanophostin A 15{
quantitatively after assay using total phosphate analysis.
The ability of guanophostin A to stimulate the InsP3R was
measured by using a low-affinity Ca2+ indicator trapped within the
intracellular stores of chicken DT40 cells expressing only
recombinant rat type I IP3R as previously reported.14 The results
relative to adenophostin A and Ins(1,4,5)P3 are shown in Table 1
and demonstrate that guanophostin A is ca. 18 fold more potent
than Ins(1,4,5)P3.
A molecular model of guanophostin A with the 29-endo
conformation of the ribose ring was built in Sybyl 7.1 (Tripos
Associates) and docked into the mouse IP3R1 binding core (PDB
file No.1N4K) using GOLD15 (Version 2.2). For a full account of
the molecular modelling methods refer to Rosenberg et al., 2003.4a
The highest scored binding mode of guanophostin closely
resembled binding mode B of adenophostin A (described in ref.
4a). The interactions of the 4,5-bisphosphate motif of Ins(1,4,5)P3
with the IP3R binding core observed in the crystallographically
determined structure were reproduced by the glucose 30,40-bispho-
sphate group of guanophostin A. Also, a cation-p interaction
Scheme 1 Reagents and conditions: a. 7, BSA, TMSOTf, MeCN, 70 uC;
(b) NH3, MeOH, rt, overnight (90%); (c) 1. (BnO)2PN(iPr)2, ImOTf,
DCM, rt; 2. m-CPBA, 278 uC; (d) 4M NaOH, dioxane, 24 h, rt; (e)
3-hydroxypropiononitrile, DBU, Et3N, DCM, 0 uC, 5d; (f) 3-hydro-
xypropiononitrile, NaH, THF, 40 uC, 4 h; (g) BnONa, DMF, rt, 30 min;
(h) BnOH, K2CO3, 70 uC, 16 h.
and 1,2-O-isopropylidene-D-xylose (5) in ten steps, using TMSOTf
as the Lewis acid was initially investigated. Different silylating
agents (for the persilylated purine) gave different yields for the
subsequent condensation. For instance, the use of TMSOTf as
both silylating agent and Lewis acid gave the expected nucleoside 8
in very low yield (27%) along with a dimeric compound (23%,
possibly glycosylated at N-9 and C2-NH2). The yield and product
distribution could not be improved either by varying the solvent or
the relative amounts of reactants. Similarly, the use of TMSCl–
HMDS as silylating agent followed by the use of TMSOTf as
catalyst also resulted in a low yield of the desired product. Use of
N,O-bis-trimethylsilyl-acetamide (BSA) as silylating agent, how-
ever, followed by TMSOTf as catalyst for the condensation yielded
8 in quantitative yield. Although the exact reason for the
dependence of the glycosylation reaction on the silylating agent
is unclear, interference by the liberated acidic counter ions of the
silylating agents (TfOH or HCl) cannot be ruled out. The high
yield of condensation when BSA was used as the silylating agent,
can be rationalized based on this argument, as neutral acetamide is
the liberated by-product in this case.
Triacetate, 8 on ammonolysis yielded the expected triol 9 in 90%
yield. Chemoselective phosphitylation10 followed by in situ
oxidation yielded the fully protected trisphosphate 10 in 94%
yield. Various attempts (4M NaOH, dioxane; 3-hydroxypropio-
nonitrile, DBU, DCM;11 3-hydroxypropiononitrile, NaH, THF;9
BnOH, K2CO312) to convert 10 to the guanosine derivative 11
failed. In some cases, 10 was inert and in others, depho-
sphorylation occurred. We therefore chose to generate the guanine
moiety before phosphorylation. Since the O6 oxygen of guanine
Scheme 2 Reagents and conditions: (a) BnOH, NaH, rt; (b) 1.
(BnO)2PN(iPr)2, ImOTf, DCM; (2) m-CPBA, 278 uC; (c) Pd(OH)2,
cyclohexene, MeOH, H2O, 80 uC, overnight.
2016 | Chem. Commun., 2006, 2015–2017
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Table 1 Ca2+ release via IP3R evoked by guanophostin Aa
EC50/nM
Hill coefficient Ca2+ release (%)
N-terminus of helix a6 is consistent with a stabilizing interaction
with the helix dipole. This observation adds an extra element to
our original model.4a
Ins(1,4,5)P3
Adenophostin A
Guanophostin A
a
23.7 ¡ 3.7 1.30 ¡ 0.14
0.9 ¡ 0.1 1.20 ¡ 0.06
1.3 ¡ 0.3 0.99 ¡ 0.08
81 ¡ 1
81 ¡ 2
83 ¡ 1
In conclusion, for the first time we have achieved the synthesis
of a full agonist of the IP3R that is equipotent to adenophostin A.
This study reveals that the 2-position of a purine base is worthy of
exploration for further design to the enhance affinity of
Ins(1,4,5)P3R ligands. Further work in this area is underway.
We thank Dr Andrew Riley for useful discussions and the
Wellcome Trust for Programme Grant support (060544 to BVLP
and 072084 to CWT).
Results (means ¡ SEM, n = 3–6) show the concentration of each
ligand required to cause the half-maximal response (EC50), the Hill
coefficient, and the % of the intracellular stores released by a
maximal concentration of each ligand.
between the guanidinium side chain of Arg504 and the guanine
base of guanophostin was observed. An ab initio study of cation-p
interactions in protein-DNA complexes revealed that in general
those involving arginine and guanine are more stable than
arginine–adenine pairs.16 In addition to the base-protein interac-
tions observed in binding mode B of adenophostin A, which
include an N3–Arg269 hydrogen bond, additional potential
hydrogen bonding interactions exist between the guanine 2-amino
group and the binding core amino acid side chain of Glu505. The
only 2-substituted adenophostin derivative so far synthesized on an
adenine base is 2-methoxy-N6-methyl-adenophostin.4a When
evaluated for Ca2+ mobilization activity, this compound was ca.
six-fold less potent than adenophostin A and only two-fold more
potent than Ins(1,4,5)P3.Thus, (with the caveat that the IP3R
binding core is not the whole, much larger, receptor) the higher
potency of guanophostin A may illustrate the advantage of having
a H-bond donor motif at the C-2 position of the purine.
In the X-ray structure17 of the mouse IP3R1 binding core, the
1-phosphate group of Ins(1,4,5)P3 approaches the N-terminal of
alpha helix 6 (a6), comprising of residues 568–585 (a9 in ref. 17). It
is known that the dipole of an alpha helix can stabilise negatively
charged groups such as phosphate at its N-terminal.18 Thus, this is
a likely interaction for the analogous 29-phosphate of adenophos-
tin and its analogues, which may even be better aligned with the
axis of this alpha helix (Fig. 2). The fact that our model of
guanophostin A binding places the 29-phosphate closer to the
Notes and references
{ Data for 15; 1H NMR (400 MHz, D2O): d 3.67–3.85 (m, 6H, H-50,
H-6A0, H-6B0, H-5A9, H-5B9, H-20), 4.04 (dd, 1H, 16.0, 9.11 Hz, H-40), 4.37
(m, 1H, H-49), 4.40-4.45 (m, 1H, H-30), 4.52 (m, 1H, H-39), 5.16 (d, 1H,
3.73 Hz, H-10), 5.20–5.26 (m, 1H, H-29), 6.17 (d, 1H, 6.18 Hz, H-19), 9.00
(s, 1H, H-8). 13C NMR (100 MHz, D2O): d 60.03 (C-59), 60.47 (C-60), 70.21
(C-20, 31P coupled), 71.36 (C-50, 31P coupled), 72.95 (C-40, 31P coupled),
73.42 (C-39, 31P coupled), 75.39 (C-29, 31P coupled), 77.90 (C-30, 31P
coupled), 84.46 (C-49), 88.36 (C-19, 31P coupled), 98.04 (C-10), 108.31 (C-5),
136.53 (C-8), 149.38 (C-4), 154.84 (C-6), 155.18 (C-2). 31P NMR
(161.94 MHz, D2O with excess of TEA): d 3.278, 3.54, 4.297. m/z (ES) =
684.1 [(M 2 H), 100%]; HRMS: Mass calcd for C16H25N5O19P3 [M 2 H],
684.0362; Found, 684.0379.
1 M. J. Berridge, P. Lipp and M. D. Bootman, Nat. Rev. Mol. Cell Biol.,
2000, 1, 11.
2 M. Takahashi, T. Kagasaki, T. Hosoya and S. Takahashi, J. Antibiot.,
1993, 46, 1643.
3 M. Takahashi, K. Tanzawa and S. Takahashi, J. Biol. Chem., 1994, 269,
369.
4 (a) H. J. Rosenberg, A. M. Riley, A. J. Laude, C. W. Taylor and
B. V. L. Potter, J. Med. Chem., 2003, 46, 4860; (b) F. Chretien,
N. Moitessier, F. Roussel, J.-P. Mauger and Y. Chapleur, Current Org.
Chem., 2000, 4, 513.
5 V. A. Correa, A. M. Riley, S. Shuto, G. Horne, E. P. Nerou,
R. Marwood, B. V. L. Potter and C. W. Taylor, Mol. Pharmacol., 2001,
59, 1206.
6 R. A. Wilcox, C. Erneux, W. U. Primrose, R. Gigg and S. R. Nahorski,
Mol. Pharmacol., 1995, 47, 1204.
7 A. M. Riley, V. Correa, M. F. Mahon, C. W. Taylor and B. V. L.
Potter, J. Med. Chem., 2001, 44, 2108.
8 (a) A. A. Koshkin, J. Fensholdt, H. M. Pfundheller and C. Lomholt,
J. Org. Chem., 2001, 66, 8504; (b) M. J. Robins, R. Zou, Z. Guo and
S. F. Wnuk, J. Org. Chem., 1996, 61, 9207; (c) M. Zhong and
M. J. Robins, Tetrahedron Lett., 2003, 44, 9327.
9 C. Rosenbohm, D. S. Pedersen, M. Frieden, F. R. Jensen, S. Arent,
S. Larsen and T. Koch, Bioorg. Med. Chem., 2004, 12, 2385.
10 Y. Hayakawa and M. Kataoka, J. Am. Chem. Soc., 1998, 120, 12395.
11 M. Ashwell, C. Bleasdale, B. T. Golding and I. K. O’Neill, J. Chem.
Soc., Chem. Commun., 1990, 955.
12 X. Chen, E. R. Kern, J. C. Drach, E. Gullen, Y.-C. Cheng and
J. Zemlicka, J. Med. Chem., 2003, 46, 1531.
13 During treatment of sodium benzoxide with 2-amino-6-chlorpurine, in
addition to the expected formation 6-O-benzyl-guanine, the formation
of 2-N-benzyl guanine analogue have been reported. C. R. Frihart and
N. J. Leonard, J. Am. Chem. Soc., 1973, 95, 7174.
14 A. J. Laude, S. C. Tovey, S. G. Dedos, B. V. L. Potter, S. C. R. Lummis
and C. W. Taylor, Cell. Calcium, 2005, 38, 45.
15 G. Jones, P. Willett, R. C. Glen, A. R. Leach and R. Taylor, J. Mol.
Biol., 1997, 267, 727.
16 R. Wintjens, J. Lievin, M. Rooman and E. Buisine, J. Mol. Biol., 2000,
302, 395.
Fig. 2 The highest scored GOLD docked binding mode of guanophostin
A with the 29 endo conformation of the ribose ring shows the potential of
this molecule to interact with Arg269 and Glu505 in addition to the
previously proposed cation-p interaction between Arg504 and the purine
ring. A new 29-phosphate helix–dipole interaction is also illustrated.
17 I. Bosanac, J.-R. Alattia, T. K. Mal, J. Chan, S. Talarico, F. K. Tong,
K. I. Tong, F. Yoshikawa, T. Furuichi, M. Iwai, T. Michikawa,
K. Mikoshiba and I. Mitsuhiko, Nature, 2002, 420, 696.
18 W. G. J. Hol, P. T. Vanduijnen and H. J. C. Berendsen, Nature, 1978,
273, 443.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2015–2017 | 2017