Synthesis and Ca2+-mobilizing activity of purine-modified mimics of adenophostin A: a model for the adenophostin-Ins(1,4,5)P3 receptor interaction.

Article abstract of DOI:10.1021/jm030883f

The synthesis of a series of adenophostin A analogues modified at C-6 and C-2 of adenine is described. The target compounds were synthesized by a convergent route involving a modified Vorbrueggen condensation of either 6-chloropurine or 2,6-dichloropurine

Full text of DOI:10.1021/jm030883f

4860
J . Med. Chem. 2003, 46, 4860-4871
Syn th esis a n d Ca 2+-Mobilizin g Activity of P u r in e-Mod ified Mim ics of
Ad en op h ostin A: A Mod el for th e Ad en op h ostin -In s(1,4,5)P 3 Recep tor
In ter a ction
Heidi J . Rosenberg,‡ Andrew M. Riley,‡ Alex J . Laude,§ Colin W. Taylor,§ and Barry V. L. Potter*,‡
Wolfson Laboratory of Medicinal Chemistry, Department of Pharmacy and Pharmacology, University of Bath,
Claverton Down, Bath BA2 7AY, UK, and Department of Pharmacology, Tennis Court Road, University of Cambridge,
Cambridge, CB2 1PD, UK
Received April 28, 2003
The synthesis of a series of adenophostin A analogues modified at C-6 and C-2 of adenine is
described. The target compounds were synthesized by a convergent route involving a modified
Vorbru¨ggen condensation of either 6-chloropurine or 2,6-dichloropurine with a protected
disaccharide, yielding two versatile intermediates capable of undergoing substitution with a
range of nucleophiles. The new analogues showed a range of abilities to mobilize Ca2+ from
the intracellular stores of permeabilized hepatocytes and are among the first totally synthetic
compounds to approach the activity of adenophostin A. In agreement with the biological results,
docking studies of adenophostin A using the recently reported X-ray crystal structure of the
type 1 Ins(1,4,5)P3 receptor binding core suggested that, in likely binding modes of adenophostin
A, the area around N6 may be relatively open, identifying this region of the adenophostin A
molecule as a promising target for further elaboration. The docking results also point to specific
interactions involving residues within the binding domain of the Ins(1,4,5)P3 receptor that
may be involved in the molecular recognition of the adenophostins.
In tr od u ction
To investigate which features of the aromatic com-
ponent of adenophostin A are important for high affinity
interactions with InsP3Rs, we have recently synthesized
several adenophostin A analogues in which the nucleo-
base motif has been replaced by various surrogates.29-31
Although this approach provided several novel com-
pounds (e.g., 4-6, Figure 1) with a range of potencies
greater than that of Ins(1,4,5)P3, none was more potent
than adenophostin A.11 To explore this base-modification
approach further, and to test steric tolerability around
the nucleobase, we have now turned our attention to
elaboration of the adenine. Several observations make
N6 of adenophostin A an attractive target for modifica-
tion. First, early results for the activity of the hypox-
anthine analogue7 of adenophostin B, prepared by
deamination of natural material, indicated that it had
activity similar to that of adenophostin A. Second, the
most potent totally synthetic adenophostin A analogue
so far identified is purinophostin (4)11,30 in which the
6-amino group is simply deleted. This suggests that
although a purine ring, or equivalent, is necessary for
high potency, the 6-amino group does not itself contrib-
ute greatly to the activity of adenophostin A. Here we
report the synthesis of a series of N6-modified analogues
(7-13, Figure 2), which are elaborated versions of
adenophostin A designed to explore the introduction of
a range of different sized hydrophobes. The route to
analogues 7-13 is versatile and also provides an entry
to C-2 modified ligands. Such compounds could have
potential in the design of high affinity adenophostin-
based tools for exploring the mechanism of action of
InsP3Rs. We also report the results of docking experi-
ments using adenophostin A and the recently reported
X-ray crystal structure of the type 1 InsP3R binding
core.
D-myo-Inositol 1,4,5-trisphosphate [Ins(1,4,5)P3, 1]
is an intracellular messenger that mediates the
release of Ca2+ from intracellular stores by binding to
Ins(1,4,5)P3-gated Ca2+ channels [Ins(1,4,5)P3 receptors,
InsP3Rs] located on the endoplasmic reticulum.1,2
With the synthesis of many inositol-based analogues of
Ins(1,4,5)P3, a sound understanding of structure-
activity relationships at the Ins(1,4,5)P3 receptor has
been established;3,4 although with the exception of
recently reported dimeric Ins(1,4,5)P3 analogues,5 none
of these ligands has proven to be more potent than
Ins(1,4,5)P3 itself. In 1993 adenophostins A (2) and its
6′′-acetylated derivative adenophostin B were isolated
from a culture broth of Penicillium brevicompactum6
and were shown to be 10 to 100 times more potent than
Ins(1,4,5)P3 at releasing Ca2+ from permeabilized cells,
and to also bind to InsP3Rs with much greater affinity
than Ins(1,4,5)P3 in equilibrium competition binding
assays.6-11 Three total syntheses of adenophostin A
have since been reported,12-15 and various novel com-
pounds related to the adenophostins have also been
synthesized in efforts to clarify the structural basis
for the high affinity interaction of the adenophostins
with InsP3Rs.16-28 Synthetic disaccharide-like analogues
such as 3,26 for example, are slightly less potent than
Ins(1,4,5)P3 in Ca2+ release assays, and no adenophostin
analogue lacking the adenine base (or an aromatic base-
substitute) has so far attained a potency greater than
that of Ins(1,4,5)P3.9-11
* Author for correspondence. Tel: +44 1225 386639. Fax +44 1225
386114. E-mail: b.v.l.potter@bath.ac.uk.
‡ University of Bath.
§ University of Cambridge.
10.1021/jm030883f CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/02/2003

Purine-Modified Mimics of Adenophostin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 4861
Sch em e 1a
a
Reagents and conditions: (i) TMSOTf, DBU, MeCN, 0 °C to
60 °C; 89% (15), 67% (16).
similar reaction between 14 and 2,6-dichloropurine
afforded the 2,6-dichloro derivative 16 in 67% yield.
With 15 and 16 in hand, attention now turned to the
reaction of these key intermediates with a range of
nucleophiles. The 6-methoxy analogue of adenophostin
A (7, Figure 2) was chosen as the first target. Reaction
of 15 with 2 equiv of sodium methoxide in methanol at
65 °C gave the 6-methoxy derivative 17 (Scheme 2).
Since these conditions also cleaved the acetate protect-
ing groups, no additional deprotection step was re-
quired, and the required triol 17 was isolated in 74%
yield. Phosphate groups were then introduced using the
phosphoramidite method. Thus, triol 17 was reacted
with a preformed complex between 1H-tetrazole and
bis(benzyloxy)(diisopropylamino)phosphine in dichloro-
methane. The resulting trisphosphite was oxidized in
situ with 3-chloroperoxybenzoic acid (mCPBA) to give
the protected trisphosphate 18. Deprotection of 18 to
give the 6-methoxy analogue of adenophostin A was
achieved with catalytic hydrogen transfer using Pd(OH)2
as catalyst and cyclohexene as transfer reagent, provid-
ing 7 in 83% yield.
F igu r e 1. Structures of D-myo-inositol 1,4,5-trisphosphate (1),
adenophostin A (2), and examples of synthetic adenophostin
analogues (3-6).
Attention now turned to the introduction of substi-
tuted amines at the N6-position. To avoid possible
complications from competing aminolysis of acetate
esters, the acetate groups of 15 were first removed
(Scheme 2) to give triol 19. Reaction of 19 with cyclo-
pentylamine at 80 °C for 3 h proceeded smoothly to
produce the N6-cyclopentyl derivative 20, isolated in
54% yield. Phosphate units were introduced as previ-
ously described for 18, and the removal of all benzyl
protecting groups was achieved by catalytic transfer
hydrogenation as before to give N6-cyclopentyl adeno-
phostin A (8).
A similar strategy was now used synthesize a series
of N6-substituted adenophostins with groups of varying
size and hydrophobicity (Scheme 3). In each case the
nucleophilic substitution at C-6 was carried out directly
on the acetate-protected 6-chloro dervative 15, whose
higher solubility in most organic solvents made it easier
to purify and to use than triol 19. Reaction of 15 with
F igu r e 2. Structures of synthetic purine-modified adenophos-
tin analogues synthesized in the present work.
Resu lts a n d Discu ssion
Syn th esis. The synthesis of all the adenophostin
analogues 7-13 begins with the disaccharide precursor
14 (Scheme 1).29-32 The target compounds were syn-
thesized by a convergent route involving a modified
Vorbru¨ggen condensation of either 6-chloropurine or 2,6-
dichloropurine with disaccharide 14 to yield two versa-
tile intermediates (15 and 16) capable of undergoing
substitution with a range of nucleophiles. Treatment of
14 with 6-chloropurine, trimethylsilyl triflate (TMSOTf),
and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) in aceto-
nitrile gave the 6-chloro derivative 15 in 89% yield. A

4862 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Rosenberg et al.
Sch em e 2a
Sch em e 3a
a
Reagents and conditions: (i) NaOMe, MeOH, reflux, 30 min,
74%; (ii) (a) (BnO)2PNPri2, 1H-tetrazole, CH2Cl2; (b) mCPBA, -78
°C, 69% (18), 25% (21); (iii) Pd(OH)2/C, MeOH/cyclohexene/H2O,
reflux, 83% (7), 46% (8); (iv) NaOMe, MeOH, rt, 55%; (v) cyclo-
pentylamine, EtOH, 80 °C, 3 h, 54%.
methylamine or dimethylamine proceeded smoothly in
each case to give N6-alkyl compounds 22 and 25,
respectively. The acetyl groups of 22 and 25 were then
removed using sodium methoxide in methanol to give
crystalline triols 23 and 26. When the reaction of 15
with cyclohexylamine was carried out, longer reaction
times were required, resulting in some acetate depro-
tection. The mixture of products was therefore deacety-
lated using sodium methoxide in methanol before
purification, giving triol 28. Phosphitylation/oxidation
of triols 23, 26, and 28 as described for 18, followed by
removal of all benzyl protecting groups by catalytic
hydrogen transfer as before gave the corresponding
trisphosphates 9-11, respectively. Attempts to apply
the same synthetic strategy to the synthesis of an N6-
noradamantyl analogue of adenophostin A, an attractive
target with a classical bulky hydrophobe, met with
problems. Reaction of 15 with 3-aminonoradamantane
was extremely sluggish, and none of the desired product
was isolated. These difficulties were eventually over-
come by using a modified approach (see below).
To explore further elaboration of the adenine, a route
to a 2,6-disubstituted adenophostin A from 16 was now
explored (Scheme 4). Prolonged reaction of 16 with a
large excess of methylamine hydrochloride and triethy-
lamine in dichloromethane and ethanol at reflux gave
rise to several inseparable products, however, and FAB
mass spectrometry of the crude product showed no peak
corresponding to the required disubstituted product. It
is known that C-2 in 2,6-dichloroadenosine is less
reactive to nucleophilic attack than C-6,33 and this
a
Reagents and conditions: (i) methylamine hydrochloride,
CH2Cl2/EtOH, Et3N, 60 °C, 67%; (ii) NaOMe, MeOH, rt, 94% (23),
57% (26); (iii) (a) (BnO)2PNPri2, 1H-tetrazole, CH2Cl2; (b) mCPBA,
-78 °C, 36% (24), 54% (27), 68% (29); (iv) Pd(OH)2/C, MeOH/
cyclohexene/H2O, reflux, 80% (9), 31% (10), 46% (11); (v) dimethyl-
amine hydrochloride, CH2Cl2/EtOH, Et3N, 60 °C, 2 h, 96%; (vi)
cyclohexylamine, CH2Cl2/EtOH, Et3N, 60 °C, 16 h, 50%.
difference in reactivity suggested that an initial reaction
of 16 with a nucleophile under controlled conditions
should yield a 6-substituted product. This should then
be capable of undergoing further substitution at C-2
with a stronger nucleophile. Accordingly, the reaction
time was reduced to 4 h, and one major product was
isolated and identified as the N6-methyl-2-chloro deriva-
1
tive 30 (Scheme 4) on the basis of its H NMR and FAB
mass spectra. Substitution of the 2-chlorine atom with
a methoxy group required forcing conditions. When 30
was heated at 65 °C with 2 equiv of sodium methoxide
in methanol for 16 h and the product was isolated, the
1
lack of any O-methyl signal in the H NMR spectrum
and the characteristic isotope pattern for chlorine in the
FAB mass spectrum confirmed that only acetate depro-
tection had occurred, with no substitution at C-2.
However, when the product (triol 31) was refluxed with
a large excess of sodium methoxide in methanol for a
further 16 h, the required triol 32 was successfully
isolated in 71% yield after purification by flash chro-
matography and crystallization. Phosphitylation/oxida-
tion of 32, as described for 17 gave fully protected 33
which, after deprotection, yielded trisphosphate 13.

Purine-Modified Mimics of Adenophostin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 4863
Ta ble 1. 45Ca2+ Release Data for 2, 7-13 from Permeabilized
Hepatocytes
Sch em e 4a
max.
EC50
nMa
,
response, EC50 with
a
ha
%
respect to 2b na
Ins(1,4,5)P3 185 ( 18
2.3 ( 0.3
35 ( 1
35 ( 1
34 ( 3
35 ( 2
37 ( 4
39 ( 2
45 ( 1
49 ( 2
34 ( 1
0.08 ( 0.01 11
2
14 ( 1.0 2.4 ( 0.3
19 ( 1.9 3.8 ( 1.5
13 ( 1.6 2.2 ( 0.2
13 ( 2.3 2.1 ( 0.4
19 ( 2.1 2.1 ( 0.1
25 ( 7.0 2.3 ( 0.1
34 ( 2.0 2.1 ( 0.2
82 ( 3.5 8.8 ( 1.6
1
13
5
7
0.57 ( 0.07
0.77 ( 0.06
0.80 ( 0.08
0.82 ( 0.08
0.56 ( 0.16
0.47 ( 0.06
0.17 ( 0.01
8
5
9
5
10
11
12
13
4
4
6
5
a
The EC50 values and Hill coefficients (h) were separately
determined for n independent experiments by fitting results to
four-parameter logistic equations. The maximum response denotes
the % of actively accumulated Ca2+ released by a maximally
effective concentration of the analogue. b The EC50 for each agonist
is expressed relative to the EC50 for 2 in paired comparisons.
Results are shown as means ( SEM.
absolute EC50 values. The N6-cyclopentyl (8), N6-methyl
(9), and N6-dimethyl (10) analogues were the most
active, being almost as potent as adenophostin A: each
had an EC50 ratio of about 0.8 (<1 is less active than
adenophostin A). The N6-cyclohexyl analogue (11), N6-
noradamantyl derivative (12), and 6-methoxy analogue
(7) were weaker (EC50 ratios of about 0.5). All these
compounds are nevertheless considerably more potent
than Ins(1,4,5)P3, and the first three have activities very
close to that of adenophostin A. Interestingly, the N6-
methyl-2-methoxy analogue (13) was significantly less
potent than all the other compounds (0.18) and only
some 2-fold more potent than Ins(1,4,5)P3.
These results showed that the InsP3R is surprisingly
tolerant of hydrophobic additions to the N6-position of
adenophostin A, suggesting that in the bound conforma-
tion of adenophostin A, this part of the adenine is either
directed into a large binding pocket or outward from
the Ins(1,4,5)P3-binding site. However, the addition of
a methoxy group at C-2 of adenine (compound 13)
caused a marked reduction in activity. This may indi-
cate either that the area of the receptor around C-2 is
more restricted, or that this modification to C-2 of
adenine has more subtle effects on adenophostin bind-
ing, perhaps by an influence on overall conformation or
on electron density in the adenine ring.
a
Reagents and conditions: (i) methylamine hydrochloride,
CH2Cl2/EtOH, Et3N, 60 °C, 4h, 67%; (ii) NaOMe, MeOH, rt 57%
(31), 84% (35); (iii) NaOMe, MeOH, reflux; 16 h, 71%; (iv) (a)
(BnO)2PNPri , 1H-tetrazole, CH2Cl2 (b) mCPBA, -78 °C, 57% (33),
2
46% (36); (v) Pd(OH)2/C, MeOH/cyclohexene/H2O, reflux, 59% (12),
47% (13); vi) 3-aminonoradamantane hydrochloride, CH2Cl2/EtOH,
Et3N, 60 °C, 48 h, 80%.
Finally, we returned to the synthesis of N6-norada-
mantyl adenophostin (12). Results from earlier experi-
ments had indicated that C-6 in 16 was considerably
more reactive than the corresponding position in 15,
prompting us to make a second attempt at the synthesis
of the N6-noradamantyl analogue via 16. When 16 was
reacted with 3-aminonoradamantane hydrochloride and
triethylamine in dichloromethane and ethanol at 60 °C
overnight (Scheme 6), the N6-monosubstituted com-
pound (34) was isolated in good yield (80%). The acetate
esters of 34 were removed using sodium methoxide in
methanol, and subsequent phosphitylation followed by
oxidation yielded fully protected 36. Catalytic transfer
hydrogenation as before successfully removed all benzyl
protecting groups with concomitant reduction at C-2 to
yield N6-noradamantyl adenophostin A (12).
Ca 2+ F lu x Exp er im en ts. The ability of compounds
7 to 13 to mobilize Ca2+ from the intracellular stores of
permeabilized rat hepatocytes was examined. The re-
sults were compared with those obtained for adenophos-
tin A (2) and Ins(1,4,5)P3. Because the analogues were
tested in experiments spanning a considerable period,
during which there was some variation in the absolute
sensitivity to Ins(1,4,5)IP3 and 2, the EC50 value (i.e.,
the concentration causing half-maximal Ca2+ release)
for each analogue is expressed relative to that for adeno-
phostin A measured in parallel experiments. Table 1
reports both these relative potencies (which provide the
most reliable comparison with adenophostin A) and the
Molecu la r Dock in g of In s(1,4,5)P 3 a n d Ad en o-
p h ostin A. While this work was nearing completion,
the X-ray crystal structure of the binding core of the
mouse type 1 receptor (InsP3R1) in complex with Ins-
(1,4,5)P3 was published.34 Hepatocytes, which were used
in the Ca2+ flux experiments described above, express
more InsP3R2 than InsP3R1, but the residues forming
the active site are conserved with the exception that
Gly268 in InsP3R1 is replaced by Leu in InsP3R2. We
have previously demonstrated that full length InsP3R1

4864 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Rosenberg et al.
F igu r e 3. Crystallographically observed binding mode of Ins(1,4,5)P3 (purple) at the receptor binding core34 and GOLD38 docked
binding mode of Ins(1,4,5)P3 (mode A, red) in docking experiments carried out with six molecules of water (W1 to W6) placed in
the binding site. When W2 was omitted, similar results were obtained for the docked position of Ins(1,4,5)P3 in the site, but the
crystallographically observed conformation of the 1-phosphate group was not reproduced (mode B, green). For clarity, hydrogen
atoms are not shown.
and InsP3R2 both bind adenophostin A with seven to
8-fold greater affinity than Ins(1,4,5)P3.35 The receptor
binding core includes only residues 224-604 of the full
length (2700 residue) receptor, and it is possible that
residues outside the binding core domain could influence
the access of Ins(1,4,5)P3 and adenophostin A to the
binding site.35 However, studies agree that the 224-
604 InsP3R1 binding core does bind adenophostin A with
higher affinity than Ins(1,4,5)P3.35-37 We therefore
reasoned that docking experiments using the InsP3R1
binding core structure might yield some clues to the
origin of the enhanced affinity of adenophostin A for
InsP3Rs.
water molecules in the site, solutions of type B, with
the 2′endo conformation were more highly scored.
The sensitivity of the docking results to the placing
of a single water molecule illustrates the difficulty of
establishing a definitive binding mode for adenophostin
A to this highly hydrophilic site by molecular docking.
However, some useful conclusions can be drawn. First,
it is clear that, as previously suggested,7,40 the glucose
3′′,4′′-bisphosphate of adenophostin A can mimic the 4,5-
bisphosphate of Ins(1,4,5)P3 at the Ins(1,4,5)P3 binding
site. The crystallographically observed interactions of
the 4,5-bisphosphate of Ins(1,4,5)P3 with Arg265, Thr267,
Gly268, Arg269, Arg508, Arg511, and Tyr567 and with
the water molecules are all reproduced by the glucose
3′′,4′′-bisphosphate of adenophostin A in the binding
modes scored most highly by GOLD. Second, and also
as previously suggested7,40 it appears that the 2′-
phosphate group of adenophostin A could interact with
the binding site in a different way to the 1-phosphate
group of Ins(1,4,5)P3. In the X-ray structure, the 1-phos-
phate of Ins(1,4,5)P3 has two contacts with Arg568, but
only an indirect interaction with the amide NH of
Lys569 via a molecule of water (W2). In many highly
scored docked conformations of adenophostin A, how-
ever, its 2′-phosphate group shows modified interactions
with the binding site. The particular combination of
residues involved is highly sensitive to the docking
procedure, but it appears that the 2′-phosphate of
adenophostin A can, at least potentially, interact with
the amide NH of Lys569 and the side chain of Arg504
in addition to, or instead of, Arg568. Third, none of the
highly scored docked conformations of adenophostin A
shows any interaction of N6 with the site, consistent
with the potent activity of purinophostin (4, Figure 1).11
The space around N6 is relatively open, consistent with
the tolerance for even quite large N6-substituents found
in the present work. In binding mode A (Figure 4) the
adenine simply projects outward into solvent and has
no interactions with the binding core, while in mode B
The docking studies were carried out using GOLD,38
an automated docking program that uses a genetic
algorithm to explore possible bound conformations of
flexible ligands to protein binding sites. It should be
noted that this and other docking approaches take only
limited account of protein flexibility, which may be an
important factor in receptor-ligand interactions.39 The
X-ray crystal structure showed that six water molecules
were directly involved in the formation of hydrogen
bonds that bridge Ins(1,4,5)P3 and the receptor,34 and
it was necessary to include some of these water mol-
ecules in the docking experiments in order to reproduce
the observed binding mode of Ins(1,4,5)P3 (see Experi-
mental Section for details). Figure 3 compares the
crystallographically observed position of Ins(1,4,5)P3 in
the binding site with docking results for Ins(1,4,5)P3
using either six or five molecules of water, while Figure
4 shows the highest scored binding modes of adeno-
phostin A when docked using the same conditions.
While the inclusion of water molecule W2 in the docking
experiments has a relatively minor effect on the results
for Ins(1,4,5)P3, the effect is greatly enhanced for
adenophostin A. In binding mode A (Figure 4), which
was highly scored in dockings using six water molecules,
the adenophostin molecule has the 3′endo conformation.
In contrast, when dockings were performed with five

Purine-Modified Mimics of Adenophostin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 4865
F igu r e 4. Possible binding modes for adenophostin A identified by GOLD38 docking experiments. Mode A (red) was highly scored
by GOLD in docking experiments with six molecules of water (W1 to W6 in Figure 3) in the site, while mode B (green) was more
highly scored when W2 was omitted. In mode B, the side chain of Arg504 shows a potential cation-π interaction with the adenine
of adenophostin A. The crystallographically observed34 binding mode of Ins(1,4,5)P3 (purple) is shown for comparison. For clarity,
water molecules and hydrogen atoms are not shown.
the N3 atom of adenine accepts a hydrogen bond from
Arg269. A hydrogen bonding interaction involving N3
could contribute to the greater potency of adenophostin
A and purinophostin (4, Figure 1) relative to the
benzimidazole analogue (5). Interestingly, in mode B
there is also a potential cation-π interaction between
the guanidinium side chain of Arg504 and the adenine
of adenophostin A. This kind of interaction would not
be sensed by GOLD, and is therefore not optimized in
the docking experiments. However, planar stacking of
arginine and aromatic side-chains within proteins has
previously been observed,41 and such cation-π interac-
tions are increasingly recognized as being important in
molecular recognition.42 Comparable interactions be-
tween arginine and adenine were readily identifiable
in RNA binding proteins such as poly A binding protein
(1CVJ ) using the program ENTANGLE,43 and a similar
cation-π interaction between an arginine side chain and
the adenine of NADP+ in the X-ray crystal structure of
aldehyde dehydrogenase in complex with NADP+ has
been reported.44 A cation-π interaction could also
explain why the naphthalenyl analogue 6 (Figure 1) is
more potent than Ins(1,4,5)P3, even though its aromatic
component has no groups that can engage in hydrogen
bonding with the binding site. On the basis of these
considerations regarding a possible role for the adenine,
we propose mode B (Figure 4) as a working model for
the interaction of adenophostin A with InsP3Rs.
lular stores of permeabilized hepatocytes showed that
the introduction of hydrophobes at N6 was well tolerated
by the receptor, giving a series of new compounds with
greater potency than the natural ligand, Ins(1,4,5)P3.
N6-methyl, N6-dimethyl, and N6-cyclopentyl analogues
showed potencies approaching that of adenophostin A
itself while the introduction of larger hydrophobes (N6-
cyclohexyl and N6-noradamantyl) or a methoxy group
at C-6 appeared to have a greater impact on activity.
Notably, addition of a methoxy group to C-2 of 9, as in
the 2,6-disubstituted analogue 13, led to a marked
reduction in potency. These results identify N6 as a
promising target for further modifications to adeno-
phostin A. Docking experiments with adenophostin A
pointed to specific interactions involving residues within
the binding domain of the Ins(1,4,5)P3 receptor that may
be involved in the molecular recognition of the adeno-
phostins. The docking results broadly support the
biological findings with respect to 7-13 and are con-
sistent with the activities of other nucleobase-modified
analogues. A model is proposed for the binding of adeno-
phostin A to Ins(1,4,5)P3 receptors, featuring a cation-π
interaction between an arginine residue and the adenine
of adenophostin A.
Exp er im en ta l Section
Chemicals were purchased from Acros, Aldrich, Sigma, and
Fluka. Dimethylformamide (DMF) was distilled from barium
oxide under reduced pressure and then stored over 4 Å
molecular sieves. Pyridine was dried over potassium hydroxide
pellets, distilled, and then stored over potassium hydroxide
pellets. Dichloromethane and acetonitrile were dried over
calcium hydride, distilled, and then stored over 4 Å molecular
sieves. 1,4-Dioxane, dimethyl sulfoxide (DMSO), ethanol,
diethyl ether, THF, and toluene were purchased in anhydrous
form. Ins(1,4,5)P3 was purchased from American Radiolabeled
Chemicals. TLC was performed on precoated plates (Merck
aluminum sheets silica 60 F254, Art. No. 5554). Products were
visualized under UV light and by dipping into phosphomolyb-
dic acid in MeOH, followed by heating or by dipping into
anisaldehyde in ethanol followed by heating. Flash chroma-
Con clu sion s
We have described a versatile strategy for the syn-
thesis of C-6 modified analogues of adenophostin A and
illustrated its application to the synthesis of 6-methoxy
adenophostin A (7) and a series of N6-alkyl analogues
(8-11) having a range of different sized hydrophobes.
The strategy was further developed to give the first 2,6-
modified adenophostin analogue (13) and also N6-nor-
adamantyl adenophostin A (12). A comparison of the
abilities of 7 to 13 to mobilize Ca2+ from the intracel-

4866 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Rosenberg et al.
tography was carried out on silica gel (particle size 40-63 µm).
1H and 13C NMR spectra were recorded on J EOL J MN GX-
270 or EX-400 NMR spectrometers. Unless otherwise stated,
chemical shifts were measured in ppm relative to internal
tetramethylsilane. 31P NMR chemical shifts were measured
in ppm and denoted positive downfield from external 85% H3-
PO4. Melting points were determined using a Reichert-J ung
hot stage microscope apparatus and are uncorrected. Mi-
croanalysis was carried out at the University of Bath Mi-
croanalysis Service. Low resolution mass spectra were recorded
at the University of Bath Mass Spectrometry Service using
+ve and -ve ion fast atom bombardment (FAB) with m-
nitrobenzyl alcohol as the matrix. High-resolution accurate
mass spectra were recorded at the University of Bath Mass
Spectrometry Service. Optical rotations were measured at
ambient temperature using an Optical Activity Ltd AA-10
polarimeter in a cell volume of 1 mL or 5 mL, and specific
rotation are given in 10-1 deg mL g.-1 HPLC analysis was
carried out on a Dynamax model SD-200 with reverse phase
column: APEX ODS II 5 µm S/N 7121103. A gradient of 0.05
M triethylammonium acetate buffer containing 0.1% tetrabu-
tylammoniumhydrogen sulfate and acetonitrile was used as
eluent at 0.9 mL/min, with a UV detector set at 259 nm. HPLC
purification was carried out on a Hewlett-Packard series
chromatograph with a strong anion-exchange resin (AGMP1,
column size 3 mm × 150 mm). A gradient of 150 mM
trifluoroacetic acid was used as eluent at 1 mL/min, with the
UV detector set at 259 nm. Synthetic phosphates were assayed
by an adaptation of the Briggs phosphate test.45
45Ca 2+ Relea se fr om P er m ea bilized Ra t Hep a tocytes.
The methods were similar to those reported previously.10
Briefly, permeabilized hepatocytes were loaded to steady state
(5 min at 37 °C) with 45Ca2+ in a cytosol-like medium (CLM:
KCl, 140 mM, NaCl, 20 mM, 2 mM MgCl2, 1 mM EGTA, 300
µM CaCl2, free [Ca2+] ) 200 nM, 20 mM Pipes, pH7.0)
containing ATP (1.5 mM), creatine phosphate (5 mM) creatine
phosphokinase (5 units/mL), and FCCP (10 µM). After 5 min,
thapsigargin (1.25 µM) was added to the cells to inhibit further
Ca2+ uptake, 30 s later the cells were added to appropriate
concentrations of the agonists, and after a further 60 s the
45Ca2+ contents of the stores were determined by rapid
filtration. Concentration-response relationships were fitted
to a four-parameter logistic equation using Kaleidegraph
software (Synergy Software, PA) from which the maximal
response, half-maximally effective agonist concentration (EC50)
and Hill slope (h) were determined. All results are expressed
as means ( SEM. Ins(1,4,5)P3 was from American Radiola-
beled Chemicals.
A molecular model of Ins(1,4,5)P3 with fully ionized phos-
phate groups was built in Sybyl and charges were calculated
by the Gasteiger Hu¨ckel method. The structure was then
minimized using the Tripos force field and a distance-depend-
ent dielectric of 1. Docking studies were then carried out using
GOLD38 (Version 2.0) running under Microsoft Windows XP
on a 2.0 GHz Pentium 4 PC. The model of Ins(1,4,5)P3 was
docked 50 times to the protein using GOLD. An active site
radius of 15 Å and automatic cavity detection were used, with
six molecules of water (W1 to W6)34 in the binding site, as
described above. The GoldScore38 fitness function was em-
ployed. The three highest scoring (“fittest”) solutions found by
Gold were almost identical, and each reproduced the crystal-
lographically determined position and conformation of Ins-
(1,4,5)P3 in the binding site with <0.6 Å rms deviation (Figure
3, mode A). To determine the minimum number and combina-
tion of water molecules required, water molecules were then
systematically removed, giving active sites with various
combinations of 5, 4, 3, 2, 1, and zero water molecules in place,
and Ins(1,4,5)P3 was redocked each time. These experiments
showed that the position of Ins(1,4,5)P3 in the active site could
be reliably reproduced by docking to a protein with five water
molecules (W1, W3, W4, W5, and W6)34 in the site (Figure 3,
mode B). These five water molecules participate in a network
of hydrogen bonds involving the 4,5-bisphosphate of Ins(1,4,5)-
P3. The fact that an analogous structure (the glucose 3′′,4′′-
bisphosphate) exists in adenophostin A justified the inclusion
of these water molecules in subsequent docking experiments
with adenophostin A. Similar results could be obtained using
as few as three deeply buried water molecules in the site (W3,
W4, and W5) but many more cycles of docking were required
before the cluster of fittest solutions was identified. In both
cases an additional water molecule (W2) was required to
consistently reproduce the crystallographically determined
orientation of the 1-phosphate group and its interaction with
Arg568. Dockings using less than three water molecules were
much less successful, giving rise to solutions in which the Ins-
(1,4,5)P3 was docked in various inverted, rotated, and distorted
conformations. In general, the more water molecules (up to
the maximum of 6) that were placed in the site, the fewer
cycles of docking were required to give reproducible and
concordant results.
Models of adenophostin A were built within Sybyl 6.8 with
either 2′endo or 3′endo conformations of the ribose and then
charged and minimized as for Ins(1,4,5)P3. The resulting
structures were used in GOLD docking experiments exactly
as described for Ins(1,4,5)P3 above using five (W1, W3, W4,
W5, and W6) or six (W1 to W6) water molecules in the active
site. The fittest solutions identified fell into two clusters,
represented by structures A and B in Figure 4. Whether type
A (3′endo) or type B (2′endo) solutions were more highly scored
by GOLD depended on whether the sixth water molecule (W2)
was included in the docking experiments. Docking with as few
as three deeply buried water molecules (W3, W4, and W5) in
the site gave the same results as for five water molecules, but
many more cycles of docking were required before the cluster
of fittest solutions was identified. The fittest solution obtained
when adenophostin A was docked with either three or five
water molecules in the site (mode B, Figure 4) showed a
potential cation-π interaction between a surface-exposed
residue, Arg504, and the adenine of adenophostin A.
Molecu la r Dock in g Exp er im en ts. Modeling was carried
out on a Silicon Graphics Octane2 workstation using Sybyl
6.8 (Tripos Associates) molecular modeling software. The
crystal structure of the Ins(1,4,5)P3 binding core of mouse
InsP3R1 in complex with Ins(1,4,5)P3 (1N4K34) was retrieved
from the Brookhaven Protein Database and prepared for
docking experiments using Sybyl 6.8 (Tripos Associates) mo-
lecular modeling software as follows. Atom types for the
phosphate oxygen atoms in the bound molecule were first
corrected to the O.co2 type, and hydrogen atoms were then
added to Ins(1,4,5)P3, water molecules, and protein. Atomic
charges for Ins(1,4,5)P3 and for the 140 crystallographically
determined water molecules were calculated by the Gasteiger
Hu¨ckel method, and dictionary charges (Kollmann all atom)
were assigned to the protein. All heavy atoms were then fixed
by designating them as a single aggregate within Sybyl, and
hydrogen atoms were minimized using the Tripos force field
and a constant dielectric of 1, terminating at a gradient of 0.01
kcal mol-1A-1. The bound Ins(1,4,5)P3 molecule and 134 water
molecules were then removed, leaving in the active site only
the six water molecules (W1-W6)34 that interact directly with
Ins(1,4,5)P3. In other experiments, the orientations of the 140
water molecules were first optimized by cycles of molecular
dynamics and minimizations, but this approach gave almost
identical orientations for W1-W6 as the simple minimization
strategy described.
2′,3′′,4′′-Tr i-O-a cet yl-2′′,5′,6′′-t r i-O-b en zyl-3′O-r-D-glu -
cop yr a n osyl-6-ch lor o-9-â-D-r ibofu r a n osylp u r in e (15). To
a stirred solution of 14 (500 mg, 0.66 mmol), 6-chloropurine
(113 mg, 0.73 mmol), and DBU (0.3 mL, 2.00 mmol) in
acetonitrile (5 mL) at 0 °C was added TMSOTf (0.48 mL, 2.65
mmol) dropwise. The mixture was heated at 60 °C for 1 h, after
which time it was cooled and quenched by careful addition of
saturated aqueous NaHCO3 solution (25 mL). The mixture was
extracted with CH2Cl2 (3 × 30 mL), and the combined extracts
were dried (MgSO4), filtered, and concentrated under reduced
pressure to leave a yellow oil. Purification by flash chroma-
tography on silica using CH2Cl2-acetone (40:1) as eluent gave
the title compound (500 mg, 89%) as a colorless oil; [R]20D +71.0

Purine-Modified Mimics of Adenophostin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 4867
1
(c 1.14, CHCl3); H NMR (400 MHz; CDCl3) δH 8.73 (s, 1 H,
then cooled to -78 °C, and mCPBA (360 mg, 1.25 mmol) was
added. After 10 min, 10% w/v aq Na2SO3 solution (15 mL) and
EtOAc (20 mL) were added, and the mixture was allowed to
warm to room temperature. The organic layer was separated
and washed with saturated aqueous NaHCO3 solution (15 mL)
and brine (15 mL). The organic layer was dried (MgSO4),
filtered, and concentrated under reduced pressure to leave an
oil, which was purified by flash chromatography on silica using
EtOAc-hexane (1:1 then 4:1) to give the title compound (200
mg, 69%) as a colorless oil; 1H NMR (400 MHz; CDCl3) δH 8.40
(s, 1 H, H-8), 8.07 (s, 1 H, H-2), 7.40-6.94 (m, 45 H, ArCH),
6.40 (d, 1 H, J 1′,2 6.2 Hz, H-1′), 5.62-5.57 (m, 1 H, H-2′), 5.33
(d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 5.06-4.87 (m, 8 H, 3.5 × OCH2Ar,
H-3′′), 4.80-4.35 (m, 13 H, 5 × OCH2Ar, H-3′, H-4′, H-4), 4.30
(AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2Ar), 4.14 (s, 3 H, OCH3),
3.86-3.83 (m, 1 H, H-5′′), 3.71-3.57 (m, 4 H, H-5′a, H-5′b,
H-6′′a, H-6′′b) and 3.43 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.7 Hz,
H-2′′); 31P NMR (162 MHz; CDCl3; 1H decoupled) δP -0.37,
-0.97, -1.14; MS: (FAB) m/z 1496.0 [(M + H)+, 78%], m/z
calcd for C80H81N4O19P3 [M + H]+, 1495.4786 found m/z
1495.4815.
H-8), 8.48 (s, 1 H, H-2), 7.37-7.23 (m, 15 H, ArCH), 6.41 (d, 1
H, J 1′,2 5.5 Hz, H-1′), 5.69 (t, 1 H, J 2′,1′ ) J 2′,3 5.5 Hz, H-2′),
5.44 (t, 1 H, J 3′′4′′ ) J
9.8 Hz, H-3′′), 5.02 (t, 1 H, J 4′′,3′′ )
3′′,2′′
J 4′′,5′′ 9.8 Hz, H-4′′), 4.96 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.71 (dd,
1 H, J 3′,4′ 3.9 Hz, J 3′,2 5.3 Hz, H-3′), 4.62 (AB, 1 H, J AB 12.1 Hz,
0.5 × OCH2Ar), 4.56-4.48 (m, 5 H, 2 × OCH2Ar, H-4′), 4.34
(AB, 1 H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.00-3.96 (m, 1 H,
H-5′′), 3.70 (dd, 1 H, J 5′a,4 2.7 Hz, J 5′a,5′ 10.9 Hz, H-5′a), 3.60
(dd, 1 H, J 5′b,4 2.7 Hz, J 5′b,5′ 10.9 Hz, H-5′b), 3.56 (dd, 1 H, J 2′′,1′′
3.5 Hz, J 2′′,3 9.8 Hz, H-2′′), 3.40-3.36 (m, 2 H, H-6′′a, H-6′′b),
1.99 (s, 3 H, CH3CO), 1.94 (s, 3 H, CH3CO) and 1.87 (s, 3 H,
CH3CO); MS: (FAB) m/z 845.3 [(M), 65%], m/z calcd for
C43H44N4O12Cl [M + H]+ 37Cl, 847.2771 found m/z 847.2738,
[M + H]+ 35Cl, 846.2834 found m/z 846.2820; Anal. C43H44N4O12
C, H, N.
2′,3′′,4′′-Tr i-O-a cet yl-2′′,5′,6′′-t r i-O-b en zyl-3′O-r-D-glu -
cop yr a n osyl-2,6-d ich lor o-9-â-D-r ibofu r a n osylp u r in e (16).
To a stirred solution of 14 (500 mg, 0.66 mmol), 2,6-
dichloropurine (138 mg, 0.73 mmol), and DBU (0.4 mL, 2.67
mmol) in acetonitrile (5 mL) at 0 °C was added TMSOTf (0.48
mL, 2.65 mmol), dropwise. The mixture was heated at 60 °C
for 1 h, after which time it was cooled and quenched by careful
addition of saturated aqueous NaHCO3 solution (25 mL). The
mixture was extracted with CH2Cl2 (3 × 30 mL), and the
combined extracts were dried (MgSO4), filtered, and concen-
trated under reduced pressure to leave a yellow oil. Purifica-
tion by flash chromatography on silica using EtOAc-hexane
(3:7) as eluent gave the title compound (390 mg, 67%) as a
3′O-r-D-Glu cop yr a n osyl-6-m et h oxy-9-â-D-r ib ofu r a n o-
sylp u r in e 2′,3′′,4′′-Tr isp h osp h a te (Na + sa lt) (7). A mixture
of 18 (35 mg, 0.023 mmol) and Pd(OH)2 on carbon (10%, 110
mg) in cyclohexene (3 mL), MeOH (6 mL) and water (0.5 mL)
was heated at 65 °C for 31 h. The catalyst was filtered off and
washed well with deionized water and MeOH. The filtrate and
washings were concentrated under reduced pressure to leave
a glassy solid. The residue was dissolved in deionized water
(1 mL), applied to a Diaion WK-20 resin column (Na+ form),
and eluted with deionized water. The eluent was concentrated
under reduced pressure to give the title trisphosphate (0.019
colorless oil. [R]20 +63.3 (c 0.44, CHCl3); 1H NMR (400 MHz;
D
CDCl3) δH 8.45 (s, 1 H, H-8), 7.37-7.16 (m, 15 H, ArCH), 6.37
(d, 1 H, J 1′,2 5.6 Hz, H-1′), 5.61 (t, 1 H, J 2′,1′ ) J 2′,3 5.6 Hz, H-2′),
5.45 (t, 1 H, J 3′′,4′′ ) J 3′′,2′′ 9.8 Hz, H-3′′), 5.02 (t, 1 H, J 4′′,3′′ ) J
1
mmol, 83%) as its sodium salt; H NMR (400 MHz; D2O) δH
4
9.8 Hz, H-4′′), 4.97 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′, 4.69 (dd, 1
8.32 and 8.28 (2 s, 2 H, H-2, H-8), 6.14 (d, 1 H, J 1′,2 6.2 Hz,
H-1′), 5.19 (d, 1 H, J 1′′,2 3.9 Hz, H-1′′), 5.10-5.04 (m, 1 H, H-2′),
4.55-4.48 (1 H, m, H-3′), 4.38-4.23 (m, 2 H, H-4′, H-4′′), 3.97
(s, 3 H, OCH3) and 3.83-3.52 (m, 7 H, H-5′a, H-5′b, H-2′′, H-3′′,
H-5′′, H-6′′a, H-6′′b); 31P NMR (162 MHz; D2O; 1H decoupled)
δP 5.64, 4.93 and 4.87; MS: (FAB) m/z 684.1 [(M - H)-, 28%],
m/z calcd for C17H27N4O19P3 [M - H]-, 684.0437 found m/z
684.0447.
′′,5′′
H, J 3′,4′ 3.5 Hz, J 3′,2 5.6 Hz, H-3′), 4.63 (AB, 1 H, J AB 12.1 Hz,
0.5 × OCH2Ar), 4.56-4.42 (m, 5 H, 2 × OCH2Ar, H-4′), 4.36
(AB, 1 H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.02-3.98 (m, 1 H,
H-5′′), 3.70 (dd, 1 H, J 5′a,4 2.4 Hz, J 5′a,5′ 10.9 Hz, H-5′a), 3.60-
3.55 (m, 2 H, H-2′′, H-5′b), 3.45-3.38 (m, 2 H, H-6′′a, H-6′′b),
1.99 (s, 3 H, CH3CO), 1.96 (s, 3 H, CH3CO) and 1.89 (s, 3 H,
CH3CO); MS: (FAB) m/z 879.1 [(M + H)+, 83%], m/z calcd for
C43H43N4O12Cl2 [M + H]+ 37Cl2 883.2352 found m/z 883.2394,
[M + H]+ 35Cl,37Cl 881.2381 found m/z 881.2391, [M + H]+ 35Cl2
879.2411 found m/z 879.2400; Anal. Calcd for C43H43N4O12 Cl2
C, H, N.
2′′,5′,6′′-Tr i-O-ben zyl-3′O-r-D-glu copyr an osyl-6-m eth oxy-
9-â-D-r ibofu r a n osylp u r in e (17). A solution of 1 M NaOMe
(0.8 mL, 0.8 mmol) in MeOH was added to a stirred solution
of 15 (400 mg, 0.4 mmol) in MeOH. The mixture was heated
at reflux for 30 min after which it was cooled, neutralized with
Dowex 50WX4-50 ion-exchange resin, and filtered. The
filtrate was concentrated under reduced pressure to leave an
oil, which was purified by flash chromatography on silica using
EtOAc-hexane (4:1) as eluent to give the title compound as a
colorless oil (250 mg, 74%); [R]20D +16.7 (c 0.6, CHCl3); 1H NMR
(400 MHz; CDCl3) δH 8.44 (s, 1 H, H-8), 8.22 (s, 1 H, H-2),
7.37-7.24 (m, 15 H, ArCH), 6.27 (d, 1 H, J 1′,2 6.6 Hz, H-1′),
4.79 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.69 (AB, 1 H, J AB 11.7 Hz,
0.5 × OCH2Ar), 4.62-4.56 (m, 1 H, H-2′), 4.53-4.41 (m, 6 H,
2.5 × OCH2Ar, H-4′), 4.21-4.19 (m, 4 H, H-3′, OCH3), 4.10 (t,
1 H, J 3′′,2′′ ) J 3′′,4′′9.4 Hz, H-3′′), 3.96-3.91 (m, 1 H, H-5′′), 3.71-
3.65 (m, 3 H, H-4′′, H-6′′a, H-6′′b), 3.60 (dd, 1 H, J 5′a,4 2.7 Hz,
J 5′a,5′ 10.5 Hz, H-5′a), 3.54 (dd, 1 H, J 5′b,4 2.7 Hz, J 5′b,5′ 10.5 Hz,
H-5′b) and 3.43 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.4 Hz, H-2′′); MS:
(FAB) m/z 715.3 [(M + H)+, 11%], m/z calcd for C38H42N4O10
[M + H]+, 715.2979 found m/z 715.2983.
2′′,5′,6′′-Tr i-O-ben zyl-3′-O-r-D-glu cop yr a n osyl-6-ch lor o-
9-â-D-r ibofu r a n osylp u r in e (19). To a stirred solution of 15
(300 mg, 0.35 mmol) in MeOH was added NaOMe (30 mg, 0.56
mmol). The mixture was stirred for 1.5 h at room temperature,
after which time it was neutralized with Dowex 50WX4-50
ion-exchange resin and filtered. The filtrate was concentrated
under reduced pressure to leave an oil, which was purified by
flash chromatography on silica using EtOAc-hexane (7:3) as
eluent to give the title compound as a colorless oil (140 mg,
1
55%); H NMR (400 MHz; CDCl3) δH 8.65 (s, 1 H, H-8), 8.44
(s, 1 H, H-2), 7.37-7.23 (m, 15 H, ArCH), 6.23 (d, 1 H, J 1′,2 6.2
Hz, H-1′), 4.81 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.70 (AB, 1 H, 0.5
× OCH2Ar), 4.68-4.64 (m, 1 H, H-2′), 4.56-4.40 (m, 7 H, 2.5
× OCH2Ar, H-2′, H-4′, OH), 4.26 (br s, 1 H, OH), 4.22 (dd, 1
H, J 3′,4 2.7 Hz, J 3′,2 5.5 Hz, H-3′), 4.06 (t covered by s, J 3′′,2
)
J 3′′,4′′ 9.7 Hz, H-3′′, OH), 3.92-3.88 (m, 1 H, H-5′′), 3.68-3.65
(m, 3 H, H-4′′, H-6′′a, H-6′′b), 3.62 (dd, 1 H, J 5′a,4′2.7 Hz, J 5′a,5′b
10.9 Hz, H-5′a), 3.55 (dd, 1 H, J 5′b,4′2.3 Hz, J 5′b,5′a 10.9 Hz, H-5′b)
and 3.43 (dd, 1 H, J 2′′,1′′ 3.5 Hz, J 2′′,3′′ 9.7 Hz, H-2′′); MS: (FAB)
m/z 719.1 [(M + H)+, 32%], m/z calcd for C37H39N4O9Cl [M +
H]+ 37Cl, 721.2454 found m/z 721.2481, [M + H]+ 35Cl, 719.2483
found m/z 719.2478.
2′′,5′,6′′-Tr i-O-b en zyl-3′O-r-D-glu cop yr a n osyl-6-cyclo-
p en tyla m in o-9-â-D-r ibofu r a n osylp u r in e (20). To a stirred
solution of 19 (140 mg, 0.19 mmol) in ethanol (10 mL) were
added cyclopentylamine (0.04 mL, 0.38 mmol) and Et3N (0.03
mL, 0.23 mmol). The mixture was heated at 80 °C for 3 h and
then cooled and concentrated under reduced pressure. Water
(20 mL) was added to the residue, and the resulting solution
was extracted with EtOAc (3 × 25 mL). The combined extracts
were dried (MgSO4), filtered, and concentrated under reduced
pressure to leave an oil, which was purified by flash chroma-
tography on silica using EtOAc-hexane (8:2) and then EtOAc-
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4′′-tr is-O-[d i(ben zyloxy)p h os-
ph or yl]-3′O-r-D-glu copyr an osyl-6-m eth oxy-9-â-D-r ibofu r a-
n osylp u r in e (18). Bis(benzyloxy)(diisopropylamino)phosphine
(405 mg, 1.17 mmol) and 1H-tetrazole (123 mg, 1.76 mmol)
were stirred together in CH2Cl2 (3 mL) for 30 min, and the
mixture thus obtained was then added to triol 17. After a
further 20 min, TLC (EtOAc:MeOH, 95:5) indicated conversion
of 17 into a single trisphosphite. The reaction mixture was

4868 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Rosenberg et al.
MeOH (10:1) to give the title compound (80 mg, 54%) as a
colorless oil; 1H NMR (400 MHz; CDCl3) δH 8.44 (s, 1 H, H-8),
8.03 (s, 1 H, H-2), 7.37-7.21 (m, 15 H, ArCH), 6.28 (d, 1 H,
J 1′,2 5.9 Hz, H-1′), 6.18 (broad s, 1 H, NH), 5.92 (broad s, 1 H,
2′′,5′,6′′-Tr i-O-ben zyl-3′O-r-D-glu cop yr a n osyl-6-m eth yl-
a m in o-9-â-D-r ibofu r a n osylp u r in e (23). To a solution of 22
(100 mg, 0.12 mmol) was added NaOMe (5 mg, 0.09 mmol),
and the mixture was stirred for 30 min. It was then concen-
trated under reduced pressure to leave a white solid, which
was washed with water (10 mL) and extracted with CHCl3 (3
× 20 mL). The combined extracts were dried (MgSO4), filtered,
and concentrated under reduced pressure to give an oil (80
mg, 94%) which was crystallized from MeOH; mp 174-175
OH), 4.77 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.66 (AB, 1 H, J
11.7
AB
Hz, 0.5 × OCH2Ar), 4.58-4.42 (m, 7 H, 2.5 × OCH2Ar, H-2′,
H-3′), 4.19-4.06 (m, 2 H, H-4′, H-3′′), 3.99-3.96 (m, 1 H, H-5′′),
3.78-3.63 (m, 3 H, H-4′′, H-6′′a, H-6′′b), 3.61-3.50 (m, 2 H,
H-5′a, H-5′b), 3.40 (dd, 1 H, J 2,1 3.5 Hz, J 2,3 9.8 Hz, H-2′′), 2.3
(broad s, 2 H, OH), 2.17-2.09 (m, 2 H, cyclopentyl ring), 1.79-
1.63 (m, 4 H, cyclopentyl ring) and 1.59-1.25 (m, 3 H,
cyclopentyl ring); MS: (FAB) m/z 768.2 [(M + H)+, 88%], m/z
calcd for C42H49N5O9 [M + H]+, 768.3608 found m/z 768.3612.
°C; [R]20 0 ( 1 (c 0.4, CHCl3); 1H NMR (400 MHz; CDCl3) δH
D
8.31 (s, 1 H, H-8), 8.02 (s, 1 H, H-2), 7.49-7.22 (m, 15 H,
ArCH), 6.25 (d, 1 H, J 1′,2 5.9 Hz, H-1′), 6.09 (broad s, 1 H, NH),
5.67 (broad s, 1 H, OH), 4.80 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.70
(AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2Ar), 4.68-4.38 (m, 7 H,
H-2′, H-3′, 2.5 × OCH2Ar), 4.19-4.18 (m, 1 H, H-4′), 4.11 (t, 1
H, J 3′′,2′′ ) J 3′′,4′′ 9.7 Hz, H-3′′), 3.96-3.90 (m, 1 H, H-5′′), 3.78-
3.60 (m, 3 H, H-4′′, H-6′′a, H-6′′b), 3.58-3.47 (m, 2 H, H-5′a,
H-5′b), 3.42 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.7 Hz, H-2′′) and 3.18
(broad s, 3 H, NCH3); MS: (FAB) m/z calcd for C38H43N5O9 [M
+ H]+ 714.3139 found m/z 714.3139. Anal. Calcd for C38H43N5O9
C, H, N.
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4′′tr is-O-[d i(ben zyloxy)p h os-
p h or yl]-3′O-r-D-glu cop yr a n osyl-6-cyclop en tyla m in o-9-â-
D-r ibofu r a n osylp u r in e (21). Phosphorylation of 20 (80 mg,
0.10 mmol) was performed as described for the synthesis of
18. Purification was achieved by flash chromatography on
silica using EtOAc-hexane (1:1 then 4:1 then 1:0) to give the
1
title compound (40 mg, 25%) as a colorless oil; H NMR (400
MHz; CDCl3) δH 8.48 (s, 1 H, H-8), 8.00 (s, 1 H, H-2), 7.63 (d,
1 H, J 8.9 Hz, NH), 7.38-6.95 (m, 45 H, ArCH), 6.29 (d, 1 H,
J 1′,2 6.3 Hz, H-1′), 5.57-5.52 (m, 1 H, H-2′) 5.30 (d, 1 H, J 1′′,2
3.5 Hz, H-1′′), 5.00-4.85 (m, 8 H, H-3′′, 3.5 × OCH2Ar), 4.76-
4.26 (m, 14 H, H-3′, H-4′, H-4′′, 5.5 × OCH2Ar) 3.82-3.80 (m,
1 H, H-5′′), 3.68-3.48 (m, 5 H, H-5′a, H-5′b, H-2′′, H-6′′a,
H-6′′b) and 2.19-1.20 (m, 9 H, cyclopentyl ring); 31P NMR (162
MHz; CDCl3; 1H decoupled) δP 0.15, -0.66, -0.79; MS: (FAB)
m/z 1564.8 [(M + H)+, 44%], m/z calcd for C84H89N5O18P3 [M
+ H]+ 1564.5364, found m/z 1564.5345.
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4′′tr is-O-[d i(ben zyloxy)p h os-
p h or yl]-3′O-r-D-glu cop yr a n osyl-6-m eth yla m in o-9-â-D-r i-
bofu r a n osylp u r in e (24). Phosphorylation of 23 (80 mg, 0.10
mmol) was performed as described for the synthesis of 18.
Purification by flash chromatography on silica using EtOAc-
hexane (1:1 then 4:1) gave the title compound (60 mg, 36%)
1
as a colorless oil; H NMR (400 MHz; CDCl3) δH 8.31 (s, 1 H,
H-8), 7.83 (s, 1 H, H-2), 7.52-6.95 (m, 45 H, ArCH), 6.32 (d, 1
H, J 1′,2 6.2 Hz, H-1′), 5.73 (br s, 1 H, NH), 5.65-5.60 (m, 1 H,
H-2′), 5.32 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 5.05-4.86 (m, 8 H, H-3′′,
3.5 × OCH2Ar), 4.76-4.36 (m, 13 H, H-3′, H-4′, H-4′′, 5 ×
OCH2Ar), 4.30 (AB, 1 H, 0.5 × OCH2Ar), 3.84-3.82 (m, 1 H,
H-5′′), 3.66-3.53 (m, 5 H, H-5′a, H-5′b, H-2′′, H-6′′a, H-6′′b)
and 3.16 (s, 3 H, NCH3); 31P NMR (162 MHz; CDCl3; 1H
decoupled) δP -0.24, -0.91, -1.05; MS: (FAB) m/z 1494.8 [(M
+ H)+, 84%], m/z calcd for C80H82N5O18P3 [M + H]+ 1494.4946
found m/z 1494.4942.
3′O-r-D-Glu cop yr a n osyl-6-cyclop en t yla m in o-9-â-D-r i-
bofu r a n osylp u r in e 2′,3′′,4′′-Tr isp h osp h a te (Na + sa lt) (8).
Deprotection of 21 (40 mg, 0.025 mmol) and purification were
performed as described for the synthesis of 7 to give the target
compound 8 (0.012 mmol, 46%) as a highly hygroscopic solid;
1H NMR (400 MHz; D2O) δH 8.02 and 8.00 (2 s, 2 H, H-2, H-8),
6.04 (d, 1 H, J 1′,2 6.6 Hz, H-1′), 5.15 (d, 1 H, J 1′′,2 3.9 Hz, H-1′′),
5.08-5.03 (m, 1 H, H-2′), 4.55-4.41 (m, 1 H, H-3′), 4.32-4.21
(m, 3 H, H-4′, H-4′′, NH), 3.83-3.44 (m, 7 H, H-5′a, H-5′b, H-2′′,
H-3′′, H-5′′, H-6′′a, H-6′′b) and 1.99-1.42 (m, 9 H, cyclopentyl
ring); 31P NMR (162 MHz; D2O; 1H decoupled) δP 3.28, 2.56
and 1.59; MS: (FAB) m/z 736.2 [(M + H)+, 44%], m/z calcd for
3′O-r-D-Glu cop yr a n osyl-6-m eth yla m in o-9-â-D-r ibofu r a -
n osylp u r in e 2′,3′′,4′′-Tr isp h osp h a te (Na + sa lt) (9). Depro-
tection of 24 (28 mg, 0.018 mmol) and purification were
performed as described for the synthesis of 7 to give the target
compound 9 (0.016 mmol, 80%) as a highly hygroscopic solid;
1H NMR (400 MHz; D2O) δH 8.09 (s, 2 H, H-2, H-8), 6.12 (d, 1
H, J 1′,2 6.6 Hz, H-1′), 5.20 (d, 1 H, J 1′′,2 3.9 Hz, H-1′′), 5.17-
5.11 (m, 1 H, H-2′), 4.49-4.48 (m, 1 H, H-3′), 4.38-4.30 (m, 3
H, H-4′, H-3′′, NH), 3.87 (m, 1 H, H-4′′), 3.78-3.59 (m, 6 H,
H-5′a, H-5′b, H-2′′, H-5′′, H-6′′a, H-6′′b) and 2.92 (s, 3 H,
C
21H34N5O18P3 [M - H]- 736.1033, found m/z 736.1019.
Gen er a l P r oced u r e for th e Am in a tion of 15. To a
solution of 15 in CH2Cl2 (5 mL) and EtOH (1 mL) were added
the corresponding amine/amine salt (7 equiv) and dry Et3N
(14 equiv). The mixture was heated at 60 °C for the required
time, then cooled and concentrated under reduced pressure.
The residue was taken up in EtOAc (30 mL) and washed with
water (20 mL). The extract was dried (MgSO4), filtered, and
concentrated under reduced pressure to leave an oil, which
was purified by flash chromatography on silica.
1
NCH3); 31P NMR (162 MHz; D2O; H decoupled) δP 2.45, 2.10
and 0.86; MS: (FAB) m/z 682.1 [(M - H)-, 84%], m/z calcd
for C17H28N5O18P3 [M - H]- 682.0564 found m/z 682.0567.
2′,3′′,4′′-Tr i-O-a cet yl-2′′,5′,6′′-t r i-O-b en zyl-3′O-r-D-glu -
cop yr a n osyl-6-(d im et h yla m in o)-9-â-D-r ibofu r a n osylp u -
r in e (25). The reaction was carried out with dimethylamine
hydrochloride (180 mg, 2.20 mmol) for 2 h as described for 22,
and purification by flash chromatography on silica using CH2-
Cl2-acetone (50:1 then 9:1) as eluent gave the title compound
2′,3′′,4′′-Tr i-O-a cet yl-2′′,5′,6-t r i-O-b en zyl-3′O-r-d -glu -
cop yr a n osyl-6-m et h yla m in o-9-â-D-r ib ofu r a n osylp u r in e
(22). The amination was carried out with methylamine
hydrochloride (218 mg, 3.2 mmol) as described above, and the
product was purified by flash chromatography on silica using
CH2Cl2-acetone (40:1) as eluent to give 22 (260 mg, 67%) as
a colorless oil; [R]20D +40.0 (c 0.60, CHCl3); 1H NMR (400 MHz;
CDCl3) δH 8.41 (s, 1 H, H-8), 8.04 (s, 1 H, H-2), 7.40-7.23 (m,
15 H, ArCH), 6.30 (d, 1 H, J 1′,2′ 5.1 Hz, H-1′), 5.87 (broad s, 1
H, NH), 5.71 (t, 1 H, J 2′,1′ ) J 2′,3′ 5.1 Hz, H-2′), 5.42 (t, 1 H,
J 3′′,2 ) J 3′′,4 9.8 Hz, H-3′′), 5.03 (t, 1 H, J 4′′,3 ) J 4′′,5′′ 9.8 Hz,
H-4′′), 4.97 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.75 (t, 1 H, J 3′,2 ) J 3′,4
5.1 Hz, H-3′), 4.61 (AB, 1 H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.56-
4.44 (m, 5 H, H-4′, 2 × OCH2Ar), 4.41 (AB, 1 H, J AB 12.1 Hz,
0.5 × OCH2Ar), 3.99-3.95 (m, 1 H, H-5′′), 3.74 (dd, 1 H, J 5′a,4
2.7 Hz, J 5′a,5′b 10.9 Hz, H-5′a), 3.64 (dd, 1 H, J 5′b,4 3.1 Hz, J 5′b,5′
10.9 Hz, H-5′b), 3.54 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.8 Hz, H-2′′),
3.39-3.33 (m, 2 H, H-6′′a, H-6′′b), 3.19 (broad s, 3 H, NCH3),
1.97 (s, 3 H, CH3CO), 1.93 (s, 3 H, CH3CO) and 1.88 (s, 3 H,
CH3CO); MS: (FAB) m/z 840.3 [(M + H)+, 42%], m/z calcd for
(290 mg, 96%) as a colorless oil; [R]20 +52.9 (c 0.55, CHCl3);
D
1H NMR (400 MHz; CDCl3) δH 8.46 (s, 1 H, H-8), 8.17 (s, 1 H,
H-2), 7.47-7.17 (m, 15 H, ArCH), 6.44 (d, 1 H, J 1′,2 5.1 Hz,
H-1′), 5.82 (t, 1 H, J 2′,1′ ) J 2′,3′ 5.1 Hz, H-2′), 5.53 (t, 1 H, J 3′′,2′′
) J 3′′,4′′ 9.8 Hz, H-3′′), 5.16 (t, 1 H, J 4′′,3 ) J 4′′,5′′ 9.8 Hz, H-4′′),
5.08 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.85 (t, 1 H, J 3′,2′ ) J 3′,4 5.1
Hz, H-3′), 4.74 (AB, 1 H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.70-
4.57 (m, 5 H, H-4′, 2 × OCH2Ar), 4.43 (AB, 1 H, J AB 12.1 Hz,
0.5 × OCH2Ar), 4.11-4.06 (m, 1 H, H-5′′), 3.85 (dd, 1 H, J 5′a,4
2.7 Hz, J 5′a,5′b 10.9 Hz, H-5′a), 3.75 (dd, 1 H, J 5′b,4 3.1 Hz, J 5′b,5′
10.9 Hz, H-5′b), 3.67-3.64 (m, 7 H, H-2′′, N(CH3)2), 3.52-3.44
(m, 2 H, H-6′′a, H-6′′b), 2.08 (s, 3 H, CH3CO), 2.04 (s, 3 H,
CH3CO) and 1.99 (s, 3 H, CH3CO); MS: (FAB) m/z 854.2 [(M
+ H)+, 48%], m/z calcd for C45H50N5O12 [M + H] 854.3567,
found m/z 854.3619; Anal. Calcd for C45H50N5O12 C, 63.30; H,
5.90; N 8.27%. Found: C, 62.80; H, 5.30; N 7.97%.
C
44H49N5O12 [M + H] 840.3455 found m/z 840.3469.

Purine-Modified Mimics of Adenophostin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 4869
2′′,5′,6′′-Tr i-O-ben zyl-3′O-r-D-glu cop yr a n osyl-6-(d im e-
th yla m in o)-9-â-D-r ibofu r a n osylp u r in e (26). Deacetylation
of 25 (290 mg, 0.34 mmol) was performed as described for the
synthesis of 23. Crystallization from MeOH, gave pure 26 (140
18. Purification by flash chromatography on silica using
EtOAc-hexane (2:3 then 1:1 then 4:1) gave the title compound
(55 mg, 68%) as a colorless oil; 1H NMR (400 MHz; CDCl3) δH
8.28 (s, 1 H, H-8), 7.82 (s, 1 H, H-2), 7.39-6.94 (m, 45 H,
ArCH), 6.32 (d, 1 H, J 1′,2 6.6 Hz, H-1′), 5.69 (br s, 1 H, NH),
5.64-5.59 (m, 1 H, H-2′), 5.31 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 5.08-
4.30 (m, 21 H, H-3′, H-4′, H-3′′, H-4′′, 8.5 × OCH2Ar,), 4.30
(AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2Ar), 3.89-3.81 (m, 1 H,
H-5′′), 3.66-3.23 (m, 5 H, H-5′a, H-5′b, H-2′′, H-6′′a, H-6′′b)
and 2.18-1.18 (m, 11 H, cyclohexyl); 31P NMR (162 MHz;
CDCl3; 1H decoupled) δP -0.23, -0.89, -1.05; MS: (FAB) m/z
1563.6 [(M + H)+, 64%], m/z calcd for C85H90N5O18P3 [M + H]+
1563.5605 found m/z 1563.5602.
mg, 57%); mp 153-155 °C; [R]20 +2.2 ( 1 (c 0.45, CHCl3); 1H
D
NMR (400 MHz; CDCl3) δH 8.26 (s, 1 H, H-8), 8.02 (s, 1 H,
H-2), 7.35-7.24 (m, 15 H, ArCH), 6.20 (d, 1 H, J 1′,2 6.6 Hz,
H-1′), 4.85 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.76-4.38 (m, 8 H, H-2′,
H-3′, 3 × OCH2Ar), 4.25-4.23 (m, 1 H, H-4′), 4.08 (t, 1 H, J 3′′,2
) J 3′′,4′′ 9.8 Hz, H-3′′), 3.91-3.88 (m, 1 H, H-5′′), 3.69-3.49 (m,
11 H, H-5′a, H-5′b, H-4′′, H-6′′a, H-6′′b, N(CH3)2) and 3.41 (dd,
1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.8 Hz, H-2′′); MS: (FAB) m/z calcd for
C
39H45N5O9 [M + H]+ 728.3250 found m/z 728.3280.
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4-tr is-O-[d i(ben zyloxy)p h os-
3′O-r-D-Glu cop yr a n osyl-6-cycloh exyla m in o-1-â-D-r ibo-
fu r a n osylp u r in e 2′,3′′,4′′-Tr isp h osp h a te (Na + sa lt) (11).
Deprotection of 29 (25 mg, 0.016 mmol) and purification were
performed as described for the synthesis of 7 to give the target
compound 11 (15.27 µmol, 46%) as a highly hygroscopic solid;
1H NMR (400 MHz; D2O) δH 8.07 and 8.03 (2 s, 2 H, H-2, H-8),
6.09 (d, 1 H, J 1′,2 6.3 Hz, H-1′), 5.19 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′),
5.14-5.08 (m, 1 H, H-2′), 4.59-4.49 (m, 1 H, H-3′), 4.35-4.26
(m, 3 H, H-4′, H-3′′, H-4′′) and 3.89-3.58 (m, 6 H, H-5′a, H-5′b,
H-2′′, H-5′′, H-6′′a, H-6′′b), 1.89-1.82 (m, 2 H, cyclohexyl),
1.73-1.59 (m, 2 H, cyclohexyl), 1.49-1.45 (m, 1 H, cyclohexyl)
and 1.35-1.08 (m, 6 H, cyclohexyl); 31P NMR (162 MHz; D2O;
1H decoupled) δP 3.31, 2.66 and 1.49; MS: (FAB) m/z 752.1
[(M + H)+, 60%], m/z calcd for C22H36N5O18P3 [M + H]+
752.1346 found m/z 752.1361.
p h or yl]-3′O-r-D-glu cop yr a n osyl-6-(d im et h yla m in o)-9-â-
D-r ibofu r a n osylp u r in e (27). Phosphorylation of 26 (40 mg,
0.05 mmol) was performed as described for the synthesis of
18. Purification was achieved by flash chromatography on
silica using EtOAc-hexane (1:1 then 4:1) to give the title
compound (45 mg, 54%) as a colorless oil; 1H NMR (400 MHz;
CDCl3) δH 8.26 (s, 1 H, H-8), 7.89 (s, 1 H, H-2), 7.39-6.95 (m,
45 H, ArCH), 6.37 (d, 1 H, J 1′,2 6.3 Hz, H-1′), 5.69-5.64 (m, 1
H, H-2′), 5.32 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 5.09 (AB, 1 H, J AB
11.7 Hz, 0.5 × OCH2Ar), 5.06-4.85 (m, 10 H, H-3′′, 4.5 ×
OCH2Ar), 4.47-4.36 (m, 10 H, H-3′, H-4′, H-4′′, 3.5 × OCH2-
Ar), 4.29 (AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2Ar), 3.84-3.82
(m, 1 H, H-5′′) and 3.65-3.43 (m, 11 H, H-5′a, H-5′b, H-2′′,
H-6′′a, H-6′′b, N(CH3)2); 31P NMR (162 MHz; CDCl3; 1H
decoupled) δP -0.05, -0.66, -0.82; MS: (FAB) m/z 1508.4 [(M
+ H)+, 79%], m/z calcd for C81H84N5O18P3 [M + H]+ 1508.5102
found m/z 1508.5068.
2′,3′′,4′′-Tr i-O-a cet yl-2′′,5′,6′′-t r i-O-b en zyl-3′O-r-D-glu -
cop yr a n osyl-2-ch lor o-6-m et h yla m in o-9-â-D-r ib ofu r a n o-
sylp u r in e (30). The reaction was carried out with methy-
lamine hydrochloride (179 mg, 2.60 mmol) and 16 (390 mg,
0.44 mmol), for 4 h as described for 22, and purification by
flash chromatography on silica using CH2Cl2-acetone (40:1)
as eluent gave the title compound (260 mg, 67%) as a colorless
3′O-r-D-Glu cop yr a n osyl-6-(d im eth yla m in o)-9-â-D-r ibo-
fu r a n osylp u r in e 2′,3′′,4′′-Tr isp h osp h a te (10). Deprotection
of 27 (20 mg, 0.013 mmol) was performed as described for the
synthesis of 7 to give the target compound 10, which was
purified by HPLC (see Methods) to give 10 (0.004 mmol, 31%)
as the free acid; 1H NMR (400 MHz; D2O) δH 8.29 and 8.17 (2
s, 2 H, H-2, H-8), 6.19 (d, 1 H, J 1′,2 6.6 Hz, H-1′), 5.12 (d, 1 H,
J 1′′,2 3.9 Hz, H-1′′), 5.10-5.05 (m, 1 H, H-2′), 4.46-4.44 (m, 1
H, H-3′), 4.34-4.27 (m, 2 H, H-4′, H-3′′), 3.93-3.86 (m, 1 H,
H-4′′) and 3.71-3.56 (m, 12 H, H-5′a, H-5′b, H-2′′, H-5′′, H-6′′a,
1
oil; H NMR (400 MHz; CDCl3) δH 7.99 (s, 1 H, H-8), 7.36-
7.24 (m, 15 H, ArCH), 6.27 (d, 1 H, J 1′,2 5.5 Hz, H-1′), 6.16
(broad s, 1 H, NH), 5.61 (t, 1 H, J 2′,1′ ) J 2′,3′ 5.5 Hz, H-2′), 5.43
(dd, 1 H, J 3′′,2′′ 10.1 Hz, J 3′′,4′′ 9.4 Hz, H-3′′), 5.05 (t, 1 H, J 4′′,3
)
J 4′′,5′′ 9.4 Hz, H-4′′), 4.97 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.71 (t, 1
H, J 3′,2 ) J 3′,4 5.5 Hz, H-3′), 4.62 (AB, 1 H, J AB 12.1 Hz, 0.5 ×
OCH2Ar), 4.55-4.47 (m, 5 H, H-4′, 2 × OCH2Ar), 4.34 (AB, 1
H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.00-3.96 (m, 1 H, H-5′′), 3.63
(dd, 1 H, J 5′a,4 2.7 Hz, J 5′a,5′b 10.9 Hz, H-5′a), 3.55 (dd, 1 H,
J 5′b,4 3.1 Hz, J 5′b,5′ 10.5 Hz, H-5′b), 3.40 (dd, 1 H, J 2′′,1 3.5 Hz,
J 2′′,3′′ 10.1 Hz, H-2′′), 3.36-3.26 (m, 2 H, H-6′′a, H-6′′b), 3.16
(broad s, 3 H, NCH3), 1.98 (s, 3 H, CH3CO), 1.94 (s, 3 H, CH3-
CO) and 1.89 (s, 3 H, CH3CO); MS: (FAB) m/z 874.1 [(M +
H)+, 90%], m/z calcd for C44H47N5O12Cl [M + H]+ 37Cl 876.2992
found m/z 876.3042, [M + H]+ 35Cl 874.3021 found m/z
874.3043.
1
H-6′′b, N(CH3)2); 31P NMR (162 MHz; D2O; H decoupled) δP
4.61 (2 × P) and 4.52 (1 P); MS: (FAB) m/z 698.0 [(M + H)+,
40%], m/z calcd for C18H30N5O18P3 [M + H]+ 698.0877 found
m/z 698.0894.
2′,5′,6′′-Tr i-O-b en zyl-3′O-r-D-glu cop yr a n osyl-6-cyclo-
h exyla m in o-9-â-D-r ibofu r a n osylp u r in e (28). The reaction
of 15 with cyclohexylamine (0.08 mL, 1.1 mmol) was carried
out overnight as described for the synthesis of 22. The oil
obtained was dissolved in MeOH and NaOMe (30 mg, 0.56
mmol) was added, and the mixture was stirred for 30 min. It
was then concentrated under reduced pressure to leave a white
solid, which was washed with water (10 mL) and extracted
with CHCl3 (3 × 20 mL). The combined extracts were dried
(MgSO4), filtered, and concentrated under reduced pressure
to give an oil which was crystallized from MeOH (70 mg, 50%);
2′′,5′,6′′-Tr i-O-ben zyl-3′O-r-D-glu cop yr a n osyl-2-ch lor o-
6-m eth yla m in o-9-â-D-r ibofu r a n osylp u r in e (31). Deacety-
lation of 30 (70 mg, 0.08 mmol) was performed as described
for the synthesis of 23 to yield the title compound (50 mg, 84%)
as an oil; 1H NMR (400 MHz; CDCl3) δH 7.97 (s, 1 H, H-8),
7.37-7.21 (m, 15 H, ArCH), 6.28 (br s, 1 H, NH), 6.22 (d, 1 H,
J 1′,2 6.3 Hz, H-1′), 4.73 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.65 (AB, 1
H, J AB 11.7 Hz, 0.5 × OCH2Ar) 4.50-4.38 (m, 7 H, H-2′, H-4′,
2.5 × OCH2Ar), 4.16-4.11 (m, 3 H, H-3′, H-3′′, H-5′′), 3.92-
3.87 (m, 1 H, H-6′′a), 3.74 (t, 1 H, J 3′′,2 ) J 3′′,4′′ 10.0 Hz, H-4′′),
3.66 (dd, 1 H, J 6′′b,5 3.1 Hz, J 6′′b,6′′ 10.5 Hz, H-6′′b), 3.53-3.48
(m, 2 H, H-5′a, H-5′b), 3.41 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 10.0
Hz, H-2′′) and 3.14 (s, 3 H, NCH3); MS: (FAB) m/z 748.2 [(M
+ H)+, 36%], m/z calcd for C38H42N5O9Cl [M + H]+ 37Cl
750.2719 found m/z 750.2754, [M + H]+ 35Cl 748.2749 found
m/z 748.2759.
2′′,5′,6′′-Tr i-O-ben zyl-3′O-r-D-glu copyr an osyl-2-m eth oxy-
6-m eth yla m in o-9-â-D-r ibofu r a n osylp u r in e (32). NaOMe
(61 mg, 1.13 mmol) was added to a solution of 31 (100 mg,
0.11 mmol) in MeOH (10 mL), and the mixture was heated at
reflux overnight. The mixture was allowed to cool and then
concentrated under reduced pressure to leave a white solid,
which was washed with water (10 mL) and extracted with
mp 103-106 °C; [R]20 +0.01 ( 1 (c 0.38, CHCl3); 1H NMR
D
(400 MHz; CDCl3) δH 8.26 (s, 1 H, H-8), 8.01 (s, 1 H, H-2),
7.37-7.22 (m, 15 H, ArCH), 6.25 (d, 1 H, J 1′,2 6.3 Hz, H-1′),
5.85 (broad s, 2 H, NH, OH), 4.78 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′),
4.65 (AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2Ar), 4.59-4.38 (m, 7
H, H-2′, H-3′, 2.5 × OCH2Ar), 4.17-4.09 (m, 3 H, H-4′, H-3′′,
OH), 3.97-3.94 (m, 1 H, H-5′′), 3.76-3.64 (m, 3 H, H-4′′, H-6′′a,
H-6′′b), 3.58-3.51 (m, 2 H, H-5′a, H-5′b), 3.47 (d, 1 H, J 5.1
Hz, cyclohexyl), 3.41 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.9 Hz, H-2′′),
2.11-1.91 (m, 3 H, cyclohexyl), 1.91-1.77 (m, 2 H, cyclohexyl),
1.68-1.66 (m, 1 H, cyclohexyl), 1.52-1.42 (m, 2 H, cyclohexyl)
and 1.34-1.22 (m, 2 H, cyclohexyl); MS: (FAB) m/z 782.3 [(M
+ H)+, 60%], m/z calcd for C43H51N5O9 [M + H]+, 782.3765
found m/z 782.3768. Anal. Calcd for C43H51N5O9 C, H, N.
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4′′tr is-O-[d i(ben zyloxy)p h os-
p h or yl]-3′O-r-D-glu cop yr a n osyl-6-cycloh exyla m in o-9-â-
D-r ibofu r a n osylp u r in e (29). Phosphorylation of 28 (40 mg,
0.05 mmol) was performed as described for the synthesis of

4870 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Rosenberg et al.
CHCl3 (3 × 20 mL). The combined extracts were dried
(MgSO4), filtered, and concentrated under reduced pressure
to give an oil which was purified by flash chromatography on
silica using CH2Cl2-acetone (40:1) as eluent to give the title
compound, which was crystallized from MeOH (60 mg, 71%);
982.3819 found m/z 982.3854, [M + H]+ 35Cl 980.3848 found
m/z 980.3848.
2′′,5′,6′′-Tr i-O-ben zyl-3′O-r-D-glu cop yr a n osyl-2-ch lor o-
6-(3-n or adam an tylam in o)-9-â-D-r ibofu r an osylpu r in e (35).
Deacetylation of 34 (90 mg, 0.092 mmol) was performed as
described for the synthesis of 23. Purification was achieved
by flash chromatography on silica using CH2Cl2-MeOH (20:
mp 192-193 °C; [R]20 0 ( 1 (c 0.40, CHCl3); 1H NMR (400
D
MHz; CDCl3) δH 7.79 (s, 1 H, H-8), 7.37-7.22 (m, 15 H, ArCH),
6.10 (d, 1 H, J 1′,2 6.3 Hz, H-1′), 5.93 (br s, 1 H, NH), 5.06 (br
s, 1 H, OH), 4.80 (d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.70 (AB, 1 H,
J AB 11.7 Hz, 0.5 × OCH2Ar), 4.60-4.42 (m, 6 H, H-2′, H-3′, 2
× OCH2Ar), 4.70 (AB, 1 H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.31-
4.29 (m, 1 H, H-4′), 3.74 (t, 1 H, J 3′′,2 ) J 3′′,4′′ 9.8 Hz, H-3′′),
3.97-3.94 (m, 2 H, H-5′′, OH), 3.79 (dd, 1 H, J 5′a,4 2.7 Hz, J 5′b,5′
10.5 Hz, H-5′), 3.68-3.61 (m, 4 H, H-5′b, H-4′′, H-6′′a, H-6′′b),
3.86 (s, 3 H, OCH3), 3.42 (dd, 1 H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.8 Hz,
H-2′′) and 3.14 (s, 3 H, NCH3); MS: (FAB) m/z 744.2 [(M +
H)+, 61%], m/z calcd for C39H45N5O18 [M + H]+ 37Cl 744.3244
found m/z 744.3246; Anal. Calcd for C4H4N5O12 C, 62.98; H,
6.10; N 9.42%. Found: C, 62.50; H, 6.10; N 9.25%.
1) as eluent to give the title compound (65 mg, 84%) as a
1
colorless oil; [R]20 +6.8 ( 1.0 (c 0.59, CHCl3); H NMR (400
D
MHz; CDCl3) δH 7.98 (s, 1 H, H-8), 7.37-7.21 (m, 15 H, ArCH),
6.24 (d, 1 H, J 1′,2 5.9 Hz, H-1′), 5.84 (br s, 1 H, NH), 4.69 (d, 1
H, J 1′′,2 3.5 Hz, H-1′′), 4.63 (AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2-
Ar), 4.59-4.36 (m, 7 H, H-2′, H-3′, 2.5 × OCH2Ar), 4.15 (t, 1
H, J 3′′,2 ) J 3′′,4′′ 9.8 Hz, H-3′′), 4.06-4.01 (m, 2 H, H-4′, H-5′′),
3.95 (dd, 1 H, J 6′′a,5 2.1 Hz, J 6′′a,6′′b 10.5 Hz, H-6′′a), 3.76 (t, 1
H, J 4′′,3 ) J 4′′,5 9.4 Hz, H-4′′,) 3.66 (dd, 1 H, J 6′′b,5 7.4 Hz, J 6′′b,6′′a
10.5 Hz H-6′′b), 3.51-3.44 (m, 2 H, H-5′a, H-5′b), 3.41 (dd, 1
H, J 2′′,1 3.5 Hz, J 2′′,3′′ 9.8 Hz, H-2′′), 2.64-2.61 (m, 1 H,
noradamantane), 2.35-2.18 (m, 6 H, noradamantane), 2.02-
1.94 (m, 2 H, noradamantane) and 1.69-1.56 (m, 4 H,
noradamantane); MS: (FAB) m/z 854.2 [(M + H)+, 52%], m/z
calcd for C46H52N5O9Cl [M + H]+ 37Cl 857.3535 found m/z
857.3583, [M + H]+ 35Cl 854.3531 found m/z 854.3502.
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4′′tr is-O-[d i(ben zyloxy)p h os-
p h or yl]-3′O-r-D-glu cop yr a n osyl-2-m eth oxy-6-m eth yla m i-
n o-9-â-D-r ibofu r a n osylp u r in e (33). Phosphorylation of 32
(30 mg, 0.040 mmol) was performed as described for the
synthesis of 18. Purification was achieved by flash chroma-
tography on silica using EtOAc-hexane (2:3 then 1:1 then 4:1)
2′′,5′,6′′-Tr i-O-ben zyl-2′,3′′,4-tr is-O-[d i(ben zyloxy)p h os-
ph or yl]-3′O-r-D-glu copyr an osyl-2-ch lor o-6-(3-n or adam an -
tyla m in o)-9-â-d -r ibofu r a n osylp u r in e (36). Phosphorylation
of 35 (40 mg, 0.04 mmol) was performed as described for the
synthesis of 18. Purification was achieved by flash chroma-
tography on silica using EtOAc-hexane (2:3 then 1:1 then 4:1)
1
to give the title compound (35 mg, 57%) as a colorless oil; H
NMR (400 MHz; CDCl3) δH 7.63 (s, 1 H, H-8), 7.37-6.96 (m,
45 H, ArCH), 6.37 (d, 1 H, J 1′,2 6.3 Hz, H-1′), 5.82 (broad s, 1
H, NH), 5.71-5.61 (m, 1 H, H-2′), 5.33 (d, 1 H, J 1′′,2 3.5 Hz,
H-1′′), 5.02-4.87 (m, 8 H, H-3′′, 3.5 × OCH2Ar), 4.77-4.27 (m,
14 H, H-3′, H-4′, H-4′′, 5.5 × OCH2Ar), 3.86-3.80 (m, 1 H,
H-5′′), 3.70 (s, 3 H, OCH3), 3.69-3.53 (m, 5 H, H-5′a, H-5′b,
H-2′′, H-6′′a, H-6′′b) and 3.11 (s, 3 H, NCH3); 31P NMR (162
MHz; CDCl3; 1H decoupled) δP 0.04, -0.58, -0.79; MS: (FAB)
m/z 1524.1 [(M), 89%], m/z calcd for C81H84N5O18P3 [M + H]+
1525.5085 found m/z 1525.5159.
1
to give the title compound (35 mg, 46%) as a colorless oil; H
NMR (400 MHz; CDCl3) δH 7.72 (s, 1 H, H-8), 7.51-6.99 (m,
45 H, ArCH), 6.24 (d, 1 H, J 1′,2 6.3 Hz, H-1′), 6.06 (br s, 1 H,
NH), 5.51-5.46 (m, 1 H, H-2′), 5.30 (d, 1 H, J 1′′,2 4.3 Hz, H-1′′),
5.04-4.33 (m, 21 H, H-3′, H-4′, H-3′′, H-4′′, 8.5 × OCH2Ar),
4.30 (AB, 1 H, J AB 11.7 Hz, 0.5 × OCH2Ar), 3.82-3.78 (m, 1
H, H-5′′), 3.72-3.52 (m, 5 H, H-5′a, H-5′b, H-2′′, H-6′′a, H-6′′b),
2.63-2.59 (m, 1 H, noradamantane), 2.34-2.17 (m, 6 H,
noradamantane), 2.09-1.98 (m, 2 H, noradamantane) and
1.65-1.56 (m, 4 H, noradamantane);31P NMR (162 MHz;
CDCl3; 1H decoupled) δP -0.23, -0.91, -0.11; MS: (FAB) m/z
calcd for C88H91N5O18P3Cl [M + H]+ 1636.5309 found m/z
1636.5334.
3′O-r-D-Glu cop yr a n osyl-6-(3-n or a d a m a n tyla m in o)-9-â-
D-r ibofu r a n osylp u r in e 2′,3′′,4′′-Tr isp h osp h a te (Na + sa lt)
(12). Deprotection of 36 (10 mg, 0.0061 mmol)) and purification
were performed as described for the synthesis of 7 to give the
target compound 12 (3.6 µmol, 59%) as a highly hygroscopic
solid; 1H NMR (400 MHz; D2O) δH 8.04 and 8.03 (2 s, 2 H,
H-2, H-8), 6.05 (d, 1 H, J 1′,2 5.5 Hz, H-1′), 5.19 (d, 1 H, J 1′′,2 3.5
Hz, H-1′′), 5.06-5.02 (m, 1 H, H-2′), 4.57-4.22 (m, 3 H, H-3′,
H-4′, H-3′′), 3.83-3.55 (m, 7 H, H-5′a, H-5′b, H-2′′, H-4′′ H-5′′,
H-6′′a, H-6′′b), 2.37-2.35 (m, 1 H, noradamantane), 2.18-2.01
(m, 6 H, noradamantane) 1.88-1.86 (m, 2 H, noradamantane)
and 1.45-1.42 (m, 4 H, noradamantane); 31P NMR (162 MHz;
D2O; 1H decoupled) δP 3.44, 2.49 and 2.49; MS: (FAB) m/z
788.0 [(M - H)-, 100%], m/z calcd for C25H37N5O18P3 [M - H]-
788.1301 found m/z 788.1336.
3′O-r-D-Glu cop yr a n osyl-2-m eth oxy-6-m eth yla m in o-9-
â-D-r ibofu r an osylpu r in e 2′,3′′,4′′-Tr isph osph ate (13). Depro-
tection of 33 (20 mg, 0.013 mmol) and purification were
performed as described for the synthesis of 7 to give the target
compound 13 (0.012 mmol, 92%) which was further purified
by HPLC to give the title trisphosphate (6.1 µmol, 47%) as
1
the free acid; H NMR (400 MHz; D2O) δH 7.82 (s, 1 H, H-8),
5.99 (d, 1 H, J 1′,2 5.1 Hz, H-1′), 5.28-5.23 (m, 1 H, H-2′), 5.11
(d, 1 H, J 1′′,2 3.5 Hz, H-1′′), 4.30-4.18 (m, 3 H, H-3′, H-4′, H-3′′),
3.83-3.80 (m, 1 H, H-4′′), 3.77 (s, 3 H, OCH3), 3.74-3.31 (m,
6 H, H-5′a, H-5′b, H-2′′, H-5′′, H-6′′a, H-6′′b) and 3.12 (s, 3 H,
1
NCH3); 31P NMR (162 MHz; D2O; H decoupled) δP 3.15, 2.28
and 1.33; MS: (FAB) m/z 714.0 [(M + H)+, 69%], m/z calcd for
C
18H30N5O18P3 [M + H]+ 714.0826 found m/z 714.0860.
2′,3′′,4′′-Tr i-O-a cet yl-2′′,5′,6′′-t r i-O-b en zyl-3′O-a -d -glu -
cop yr a n osyl-2-ch lor o-6-(3-n or a d a m a n tyla m in o)-9-â-D-r i-
bofu r a n osylp u r in e (34). The reaction was carried out with
3-aminonoradamantane hydrochloride (212 mg, 1.2 mmol) and
16 (179 mg, 0.2 mmol), for 48 h as described for 22, and the
product was purified by flash chromatography on silica using
EtOAc-hexane (4:1) as eluent to give the title compound (160
mg, 80%) as a colorless oil; 1H NMR (400 MHz; CDCl3) δH 7.97
(s, 1 H, H-8), 7.36-7.24 (m, 15 H, ArCH), 6.26 (d, 1 H, J 1′,2 5.5
Hz, H-1′), 6.16 (br s, 1 H, NH), 5.61 (dd, 1 H, J 2′,1′ 5.5 Hz, J 2′,3′
5.0 Hz, H-2′), 5.43 (t, 1 H, J 3′′,2 ) J 3′′,4′′ 9.9 Hz, H-3′′), 5.04 (t,
1 H, J 4′′,3 ) J 4′′,5′′ 9.9 Hz, H-4′′), 4.97 (d, 1 H, J 1′′,2 3.3 Hz, H-1′′),
4.72 (t, 1 H, J 3′,2 ) J 3′,4 5.0 Hz, H-3′), 4.62 (AB, 1 H, J AB 12.1
Hz, 0.5 × OCH2Ar), 4.55-4.46 (m, 5 H, H-4′, 2 × OCH2Ar),
4.36 (AB, 1 H, J AB 12.1 Hz, 0.5 × OCH2Ar), 4.10-3.97 (m, 1
H, H-5′′), 3.71 (dd, 1 H, J 5′a,4 2.9 Hz, J 5′a,5′b 10.6 Hz, H-5′a),
3.63 (dd, 1 H, J 5′b,4 3.3 Hz, J 5′b,5′ 10.6 Hz, H-5′b), 3.54 (dd, 1 H,
J 2′′,1 3.3 Hz, J 2′′,3′′ 9.9 Hz, H-2′′), 3.43-3.35 (m, 2 H, H-6′′a,
H-6′′b), 2.65-2.61 (m, 1 H, noradamantane), 2.35-2.19 (m, 7
H, noradamantane), 2.05-1.94 (m, 1 H, noradamantane), 1.98
(s, 3 H, CH3CO), 1.94 (s, 3 H, CH3CO), 1.89 (s, 3 H, CH3CO)
and 1.69-1.57 (m, 4 H, noradamantane); MS: (FAB) m/z 980.3
[(M + H)+, 80%], m/z calcd for C52H58N5O12Cl [M + H]+ 37Cl
Ack n ow led gm en t. We thank the Wellcome Trust
for a Prize Studentship (H.J .R.) and Program Grant
Support (0060554 to B.V.L.P. and 039662 to C.W.T.).
We also thank Mr. J ames R. Robinson, University of
Bath, UK, and Dr. Yousif Shamoo, Rice University,
Houston, TX, for useful discussions on molecular model-
ing and protein-nucleotide interactions, respectively.
Su p p or tin g In for m a tion Ava ila ble: 13C NMR data for
1
selected compounds and H NMR spectra for compounds 7-13.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Refer en ces
(1) Berridge, M. J .; Lipp, P.; Bootman, M. D. The versatility and
universality of calcium signaling. Nat. Rev. Mol. Cell Biol. 2000,
1, 11-21.

Purine-Modified Mimics of Adenophostin A
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 4871
(2) Taylor, C. W.; Laude, A. J . IP3 receptors and their regulation
by calmodulin and cytosolic Ca2+. Cell Calcium 2002, 32, 321-
334.
(3) Potter, B. V. L.; Lampe, D. Chemistry of inositol lipid mediated
cellular signaling. Angew. Chem., Int. Ed. Engl. 1995, 34, 1933-
1972.
(4) Wilcox, R. A.; Primrose, W. U.; Nahorski, S. R.; Challiss, R A.
J . New developments in the molecular pharmacology of the myo-
inositol 1,4,5-trisphosphate receptor. Trends Pharmacol. Sci.
1998, 19, 467-475.
(5) Riley, A. M.; Morris, S. A.; Nerou, E. P.; Correa, V.; Potter, B.
V. L.; Taylor, C. W. Interactions of inositol 1,4,5-trisphosphate
(IP3) receptors with synthetic poly(ethylene glycol)-linked dimers
of IP3 suggest close spacing of the IP3-binding sites. J . Biol.
Chem. 2002, 277, 40290-40295.
(6) Takahashi, M.; Kagasaki T.; Hosoya, T.; Takahashi, S. Adeno-
phostins A and B: potent agonists of inositol 1,4,5-trisphosphate
receptor produced by Penicillium brevicompactum. J . Antibiot.
1993, 46, 1643-1647.
(7) Takahashi, M.; Tanzawa, K.; Takahashi, S. Adenophostins,
newly discovered metabolites of Penicillium brevicompactum, act
as potent agonists of the inositol 1,4,5-trisphosphate Receptor.
J . Biol. Chem. 1994, 269, 369-372.
(8) Hirota, J .; Michikawa, T.; Miyawaki, A.; Takahashi, M.; Tan-
zawa, K.; Okura, I.; Furuichi, T.; Mikoshiba, K. Adenophostin-
mediated quantal Ca2+ release in the purified and reconstituted
inositol 1,4,5-trisphosphate receptor type 1. FEBS Lett. 1995,
368, 248-252.
(9) Murphy, C. T.; Riley, A. M.; Lindley, C. J .; J enkins, D. J .;
Westwick, J .; Potter, B. V. L. Structural analogues of D-myo-
inositol 1,4,5-trisphosphate and adenophostin A: Recognition by
cerebellar and platelet inositol 1,4,5-trisphosphate receptors.
Mol. Pharmacol. 1997, 52, 741-748.
(10) Marchant, J . S.; Beecroft, M. D.; Riley, A. M.; J enkins, D. J .;
Marwood, R. D.; Taylor, C. W.; Potter, B. V. L. Disaccharide
polyphosphates based upon adenophostin A activate hepatic
D-myo-inositol 1,4,5-trisphosphate receptors. Biochemistry 1997,
36, 12780-12790.
(11) Correa, V.; Riley, A. M.; Shuto, S.; Horne, G.; Nerou, E. P.;
Marwood, R. D.; Potter, B. V. L.; Taylor, C. W. Structural
determinants of adenophostin A activity at inositol trisphosphate
receptors. Mol. Pharmacol. 2001, 59, 1206-1215.
(12) Hotoda, H.; Takahashi M.; Tanzawa, K.; Takahashi, S.; Kaneko,
M. IP3 receptor-ligand. 1: Synthesis of adenophostin A. Tetra-
hedron Lett. 1995, 36, 5037-5040.
(13) van Straten N. C. R.; van der Marel, G. A.; van Boom, J . H. An
expeditious route to the synthesis of adenophostin A. Tetrahe-
dron 1997; 53, 6509-6522.
(14) Marwood, R. D.; Correa, V.; Taylor, C. W.; Potter, B. V. L.
Synthesis of adenophostin A. Tetrahedron: Asymmetry 2000, 11,
397-403.
(15) Marwood, R. D.; Riley, A. M.; J enkins, D. J .; Potter, B. V. L.
Synthesis of adenophostin A and congeners modified at glucose.
J . Chem. Soc., Perkin Trans. 1 2000, 1935-1947.
(16) Wilcox, R. A.; Erneux, C.; Primrose, W. U.; Gigg, R.; Nahorski,
S. R. 2-hydroxyethyl R-D-glucopyranoside 2, 3′, 4′-trisphosphate,
a novel, metabolically resistant, adenophostin A and myo-inositol
1,4,5-trisphosphate analogue potently interacts with the myo-
inositol 1,4,5-trisphosphate receptor. Mol. Pharmacol. 1995, 47,
1201-1211.
(17) Moitessier, N.; Chre´tien, F.; Chapleur, Y.; Humeau, C. Synthesis
and biological activities of inositol 1,4,5-trisphosphate mimics
related to xylopyranosides. Tetrahedron Lett. 1995, 36, 8023-
8026.
(24) de Kort, M.; Regenbogen, A. D.; Overkleeft, H. S.; Challiss, R.
A. J .; Iwata, Y.; Miyamoto, S.; van der Marel, G. A.; van Boom,
J . H. Synthesis and biological evaluation of cyclophostin: a 5′,6′′-
tethered analog of adenophostin A. Tetrahedron 2000, 56, 5915-
5928.
(25) de Kort, M.; Luijendijk, J .; van der Marel, G. A.; van Boom, J .
H. Synthesis of photoaffinity derivatives of adenophostin A. Eur.
J . Org. Chem. 2000, 3085-3092.
(26) Rosenberg, H. J .; Riley, A. M.; Marwood, R. D.; Correa, V.;
Taylor, C. W.; Potter, B. V. L. Xylopyranoside-based agonists of
D-myo-inositol 1,4,5-trisphosphate receptors: synthesis and ef-
fect of stereochemistry on biological activity. Carbohydr. Res.
2001, 332, 53-66.
(27) Roussel, F.; Moitessier, N.; Hilly, M.; Chre´tien, F.; Mauger, J .-
P.; Chapleur, Y. d-myo-inositol 1,4,5-trisphosphate and adeno-
phostin mimics: importance of the spatial orientation of a
phosphate group on the biological activity. Bioorg. Med. Chem.
2002, 10, 759-768.
(28) Riley, A. M.; J enkins, D. J .; Marwood, R. D.; Potter, B. V. L.
Synthesis of glucopyranoside-based ligands for D-myo-inositol
1,4,5-triphosphate receptors. Carbohydr. Res. 2002, 337, 1067-
1082.
(29) Marwood, R. D.; Shuto, S.; J enkins, D. J .; Potter, B. V. L.
Convergent synthesis of adenophostin A analogues via a base
replacement strategy. Chem. Commun. 2000, 219-220.
(30) Marwood, R. D.; J enkins, D. J .; Correa, V.; Taylor, C. W.; Potter,
B. V. L. Contribution of the adenine base to the activity of
adenophostin A investigated using a base replacement strategy.
J . Med. Chem. 2000, 43, 4278-4287.
(31) Shuto, S.; Horne, G.; Marwood, R. D.; Potter, B. V. L. Total
synthesis of nucleobase-modified adenophostin A mimics. Chem.
Eur. J . 2001, 7, 4937-4946.
(32) de Kort, M.; Correa, V.; Valentijn, A. R. P. M.; van der Marel,
G. A.; Potter, B. V. L.; Taylor, C. W.; van Boom, J . H. Synthesis
of potent agonists of the D-myo-inositol 1,4,5-trisphosphate
receptor based on clustered disaccharide polyphosphate ana-
logues of adenophostin A. J . Med. Chem. 2000, 43, 3295-3303.
(33) Schaeffer, H. J .; Thomas, H. J . Synthesis of potential anticancer
agents. XIV. Ribosides of 2,6-disubstituted purines. J . Am.
Chem. Soc. 1958, 80, 3738-3742.
(34) Bosanac, I.; Alattia, J .-R.; Mal, T. K.; Chan, J .; Talarico, S.; Tong,
F. K.; Tong, K. I.; Yoshikawa, F.; Furuichi, T.; Iwai, M.;
Michikawa, T.; Mikoshiba, K.; Ikura, M. Structure of the inositol
1,4,5-trisphosphate receptor binding core in complex with its
ligand. Nature 2002, 420, 696-700.
(35) Morris, S. A.; Nerou, E. P.; Riley, A. M.; Potter, B. V. L.; Taylor,
C. W. Determinants of adenophostin A binding to inositol
trisphosphate receptors. Biochem. J . 2002, 367, 113-120.
(36) VanLingen, S.; Sipma, H.; De Smet, P.; Callewaert, G.; Missiaen,
L.; De Smedt, H.; Parys, J . B. Ca2+ and calmodulin differentially
modulate myo-inositol 1,4,5-trisphosphate (IP3)-binding to the
recombinant ligand-binding domains of the various IP3 receptor
isoforms. Biochem. J . 2000, 346, 275-280.
(37) Glouchankova, L.; Krishna, U. M.; Potter, B. V. L.; Falck, J . R.;
Bezprozvanny, I. Association of the inositol (1, 4, 5)-trisphos-
phate receptor ligand binding site with phosphatidylinositol
(4,5)-bisphosphate and adenophostin A. Mol. Cell. Biol. Res.
Commun. 2000, 3, 153-158.
(38) J ones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible
docking. J . Mol. Biol. 1997, 267, 727-748.
(39) Teague, S. J . Implications of protein flexibility for drug discovery.
Nat. Rev. Drug Discovery 2003, 2, 527-541.
(40) Hotoda, H.; Murayama, K.; Miyamoto, S.; Iwata, Y.; Takahashi,
M.; Kawase, Y.; Tanzawa, K.; Kaneko, M. Molecular recognition
of adenophostin A, a very potent Ca2+ inducer, at the D-myo-
inositol 1,4,5-trisphosphate receptor. Biochemistry 1999, 38,
9234-9241.
(41) Flocco, M. M.; Mowbray, S. L. Planar stacking interactions of
arginine and aromatic side-chains in proteins. J . Mol. Biol. 1994,
235, 709-717.
(42) Meyer, E. A.; Castellano, R. K.; Diederich, F. Interactions with
aromatic rings in chemical and biological recognition. Angew.
Chem., Int. Ed. 2003, 42, 1210-1250.
(43) Allers, J .; Shamoo, Y. Structure-based analysis of protein-RNA
interactions using the program ENTANGLE. J . Mol. Biol. 2001,
311, 75-86.
(44) Ahvazi, B.; Coulombe, R.; Delarge, M.; Vedadi, M.; Zhang, L.;
Meighen, E.; Vrielink, A. Crystal structure of the NADP+-
dependent aldehyde dehydrogenase from Vibrio harveyi: struc-
tural implications for cofactor specificity and affinity. Biochem.
J . 2000, 349, 853-861.
(45) Sawyer, D. A.; Potter, B. V. L. Total synthesis of fluorinated
analogues of inositol and inositol 1,4,5-trisphosphate. J . Chem.
Soc., Perkin. Trans. 1 1992, 923-932.
(18) J enkins, D. J .; Marwood, R. D.; Potter, B. V. L. A disaccharide
polyphosphate mimic of 1d-myo-inositol 1,4,5-trisphosphate.
Chem. Commun. 1997, 449-450; Corrigendum 805.
(19) Shuto, S.; Tatani, K.; Ueno, Y.; Matsuda, A. Synthesis of
adenophostin analogues lacking the adenine moiety as novel
potent IP3 receptor ligands: some structural requirements for
the significant activity of adenophostin A. J . Org. Chem. 1998,
63, 8815-8824.
(20) Marwood, R. D.; Riley, A. M.; Correa, V., Taylor, C. W.; Potter,
B. V. L. Simplification of adenophostin A defines a minimal
structure for potent glucopyranoside-based mimics of d-myo-
inositol 1,4,5-trisphosphate. Bioorg. Med. Chem. Lett. 1999, 9,
453-458.
(21) Rosenberg, H. J .; Riley, A. M.; Correa, V.; Taylor, C. W.; Potter,
B. V. L. C-glycoside based mimics of D-myo-inositol 1,4,5-
trisphosphate. Carbohydr. Res. 2000, 329, 7-16.
(22) de Kort, M.; Regenbogen, A. D.; Valentijn, A. R. P. M.; Challiss,
R. A. J .; Iwata, Y.; Miyamoto, S.; van der Marel, G. A.; van Boom,
J . H. Spirophostins: conformationally restricted analogues of
adenophostin A. Chem. Eur. J . 2000, 6, 2696-2704.
(23) Riley, A. M.; Correa, V.; Mahon, M. F.; Taylor, C. W.; Potter, B.
V. L. Bicyclic analogues of D-myo-inositol 1,4,5-trisphosphate
related to adenophostin A: synthesis and biological activity. J .
Med. Chem. 2001, 44, 2108-2117.
J M030883F