Poly(ethylene glycol)-linked dimers of D-myo-inositol 1,4,5- trisphosphate

Article abstract of DOI:10.1039/b002312o

The first poly(ethylene glycol)-linked dimers of D-myo-inositol 1,4,5- trisphosphate have been synthesised as probes for multi-subunit binding proteins of this ubiquitous second messenger.

Full text of DOI:10.1039/b002312o

Poly(ethylene glycol)-linked dimers of -myo-inositol 1,4,5-trisphosphate
D
Andrew M. Riley and Barry V. L. Potter*
Wolfson 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
Received (in Cambridge, UK) 23rd March 2000, Accepted 27th April 2000
The first poly(ethylene glycol)-linked dimers of
D
-myo-
linker derived from hexa(ethylene glycol) (estimated4 r.m.s.
length 1.5 nm), and then three larger dimers were synthesised
from PEGs with average molecular weights 1450, 3350 and
8000 (estimated rms lengths 3, 5 and 8 nm, respectively).
The synthesis begins with diol 16 (Scheme 1). Stannylene-
mediated regioselective alkylation of the equatorial 1-OH group
with p-methoxybenzyl chloride in the presence of CsF gave the
alcohol 2 in 80% yield. Alkylation of 2 with bromoacetonitrile
and sodium hydride in acetonitrile at reduced temperatures7
then gave the 2-O-cyanomethyl derivative 3 {mp 120–121 °C,
inositol 1,4,5-trisphosphate have been synthesised as probes
for multi-subunit binding proteins of this ubiquitous second
messenger.
D-myo-inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] acts as an
intracellular second messenger in almost all mammalian cells
by evoking the release of Ca2+ ions from intracellular stores
through Ins(1,4,5)P3-gated ion channels1 [Ins(1,4,5)P3 re-
ceptors, IP3Rs] located in the endoplasmic reticulum. IP3Rs are
tetramers, composed of four subunits, each with a single
Ins(1,4,5)P3 binding site, surrounding the central ion channel.
This suggests the possibility of designing multivalent ligands
for IP3Rs, in which two or more copies of the natural ligand
connected by a linker may be able to access multiple binding
sites simultaneously, potentially giving enhanced potency,
selectivity or other novel effects. To this end, a synthesis of
bivalent and tetravalent analogues has been reported in which
two or four molecules of a synthetic carbohydrate-based
Ins(1,4,5)P3 mimic were directly attached to a small hydro-
phobic hub.2 As the IP3R is large (12 nm width, estimated from
electron microscopy3) and the locations of the Ins(1,4,5)P3
binding sites are unknown, a more promising and versatile
approach may be to use bivalent ligands with hydrophilic
polymeric linkers in a realistic range of lengths. This strategy
has been successfully employed, using very simple chemistry,
to identify super-potent dimeric ligands for tetrameric cyclic
nucleotide gated (CNG) channels of photoreceptor and olfac-
tory neurones by using a series of dimers of the 8-thio derivative
of the natural ligand (cGMP) linked by PEG chains.4 These
PEG-linked dimers of cGMP also showed partial agonist
properties at photoreceptor CNG channels, and some were
membrane-permeant.
18
[a]D +56 (c 1, CHCl3)} in 83% yield. The nitrile was
smoothly reduced with LAH in THF to the primary amine,
which was not isolated but temporarily protected as the
trifluoroacetamide by reaction with ethyl trifluoroacetate in
THF at room temperature.8 Finally, the acid-labile butanediace-
tal (BDA) and PMB protecting groups were cleaved using TFA,
exposing the hydroxy groups at positions 1, 4 and 5. The
trifluoroacetyl protection was not affected under these condi-
tions, and the triol 4 {mp 130–131 °C, [a]D18 +3 (c 1, CHCl3)}
was obtained in 77% overall yield from 3. Phosphitylation using
bis(benzyloxy)(N,N-diisopropylamino)phosphine and 1H-tetra-
zole followed by in situ oxidation with MCPBA gave crystalline
20
5 {mp 85–87 °C, [a]D 26 (c 1, CHCl3)}.
Application of this approach to Ins(1,4,5)P3 is, however,
synthetically much more difficult. Nevertheless, studies have
shown that bulky groups may be attached to the axial 2-oxygen
atom of Ins(1,4,5)P3 with minimal reduction in affinity for the
IP3R,5 suggesting that at the Ins(1,4,5)P3 binding sites of the
IP3R this area may be open to solvent. We therefore chose to
synthesise Ins(1,4,5)P3 dimers linked via the 2-position (Fig. 1).
The synthetic strategy was first explored using a short PEG
Scheme 1 Reagents and conditions: i, (a) Bu2SnO, MeOH, 4 Å sieves,
Soxhlet, reflux, 16 h; (b) PMBCl, CsF, DMF, 50 °C, 5 h, 84%; ii, NaH,
BrCH2CN, CH3CN, 220 to 240 °C 5 h then room temp., 16 h, 83%; iii, (a)
LAH, THF, room temp, 1 h, (b) ethyl trifluoroacetate, THF, room temp., 1 h,
(c) TFA–CH2Cl2–H2O (19+20+1), room temp., 30 min, 77% from 3; iv,
(BnO)2PNPr12, 1 H-tetrazole, CH2Cl2, room temp., 1 h, then MCPBA,
278 °C to room temp., 30 min (96%); v, LiOH·H2O (10 equiv.) in THF–
MeOH–H2O (2+2+1), room temp., 1 h, 91%. Bn = benzyl, PMB = p-
methoxybenzyl.
It was now necessary to expose the primary amine by
selective removal of the trifluoroacetyl group. Treatment of 5
with methanolic ammonia or K2CO3 in aqueous methanol was
only partially successful; long reaction times were required,
leading to the formation of various polar products, presumably
from partial cleavage of benzylphosphate esters. However, it
was found that the trifluoroacetyl group could cleanly be
removed by treatment with LiOH in THF–MeOH–H2O9 for 1 h.
Fig. 1 Ins(1,4,5)P3 dimers could target multiple receptor binding sites.
DOI: 10.1039/b002312o
Chem. Commun., 2000, 983–984
This journal is © The Royal Society of Chemistry 2000
983

potential pharmacological tools to probe the IP3R. Other
proteins are known to have multiple binding sites11 for
Ins(1,4,5)P3 or phosphatidylinositol 4,5-bisphosphate and the
Ins(1,4,5)P3 dimers could also be used to investigate these. A
particularly attractive application might be to use Ins(1,4,5)P3
metabolising enzymes to convert Ins(1,4,5)P3 dimers into
dimers of other inositol phosphates, which may have their own
intracellular targets proteins.
We thank the Wellcome Trust for Programme Grant support
(045491) and Drs S. W. Garrett and I. S. Blagborough for
valuable advice.
Notes and references
† 7a: this was synthesised by reaction of hexa(ethylene glycol) with 6
equivalents of bis(p-nitrophenyl) carbonate in DMF in the presence of
diisopropylethylamine, and was purified by flash chromatography on silica
gel before use. 7b and 7d were synthesised in a similar way from PEGs with
average molecular weights of 1450 and 8000, respectively. The bis(p-
nitrophenyl carbonate)-PEG 7c, derived from a PEG of average molecular
weight 3350, is commercially available (Sigma).
‡ Selected data for 8a–8d: 8a dH(CDCl3, 400 MHz) 3.24–3.30 (4H, m,
OCH2CH2N), 3.38 (2H, dd, J 10.0, 2.1 Hz, 3-H), 3.50 (4H, s, OCH2CH2N),
3.55–3.58 (16H, m, 8 3 PEG CH2), 3.66–3.70 (4H, m, 2 3 PEG CH2), 4.00
(2H, dd, J 9.7, 9.4 Hz, 6-H), 4.04 (2H, br s, 2-H), 4.10–4.16 (4H, m, 2 3
PEG CH2), 4.18 (2H, ddd, J 9.7, 7.3, 2.3 Hz, 1-H), 4.48 (2H, ddd, J 9.4, 9.4,
9.1 Hz, 5-H), 4.47, 4.57 (4H, ABq, JAB 11.7 Hz, OCH2Ph), 4.62–4.68 (2H,
0.5 of ABq with 3JHP coupling, JAB 11.7, JHP 8.5 Hz, POCH2Ph), 4.74–5.07
(28H, m, 4-H and 6.5 AB systems of OCH2Ph), 5.60 (2H, br t, J 5.3 Hz,
NH), 6.96–6.98 (4H, m, Ph), 7.08–7.36 (76H, Ph); protected dimers 8b–8d
had similar 1H NMR spectra to 8a, except; 8b: d 3.55–3.70 (approx. 130H,
m, CH2 of PEG), 8c: d 3.55–3.70 (approx. 300H, m, CH2 of PEG); 8d: d
3.55–3.70 (approx. 700H, m, CH2 of PEG).
Scheme 2 Reagents and conditions: i, 6 (3–4 equiv.), DMF, room temp., 24
h, 31–58%; ii, Pd–C, H2, 50 psi, room temp., 24 h, 52–65%. Bn = benzyl,
PNP = p-nitrophenyl. *a, n = 4; b, n ≈ 30; c, n ≈ 75; d, n ≈ 180.
Within this time there was little effect on the benzylphosphate
groups. The amine 6 was found to be unstable, and was
therefore freshly prepared for each cross-linking reaction and
used immediately.
§ Selected data for 9a–d: 9a: dH(CD3OD, 400 MHz) d ca. 3.3 [4H, m
(buried), OCH2CH2N], 3.61–3.70 (22H, m, 3-H and 10 3 PEG CH2),
3.77–3.85 (2H, m, OCHHCH2N), 3.92–4.00 (6H, m, 5-H, 6-H and
OCHHCH2N), 4.01–4.06 (4H, m, 1-H and 2-H), 4.14–4.20 (4H, m, 2 3
CH2 of PEG), 4.32 (2H, ddd, J 9.4, 8.9, 8.6 Hz, 4-H); dP(CD3OD, 162 MHz)
2.06 (2 P), 3.19 (2 P) and 3.66 (2 P); MS m/z (2ve ion FAB, relative
intensity); 1281 (90%), 1259 [M2, 80%], 97 [H2PO42, 100%]; Accurate
mass FAB2: calc. for C30H61N2O39P62, 1259.127; found 1259.122. Dimers
Cross-linking of two molecules of 6 was first attempted using
the bis(p-nitrophenylcarbonate) derivative† 7a of hexa(ethylene
glycol) (Scheme 2). Reaction of 7a with 3 equivalents of 6 in
DMF gave the protected PEG-linked dimer 8a, which was
isolated in moderate yield (58% based on 7a) yield after
purification by flash chromatography on silica gel. The 1H
1
9b–d had similar H NMR spectra to that of 9a, except for the increasing
integral of the signal at d ca. 3.60–3.70 corresponding to CH2 of PEG. Their
31P NMR spectra were also similar to that of 9a.
NMR spectrum of 8a confirmed that it was a dimer‡ and the 31
P
NMR spectrum showed three signals, each corresponding to
two equivalent phosphorus atoms in 8a. Removal of all sixteen
benzyl groups from 8a was easily achieved by hydrogenation
over Pd–C. Purification by ion-exchange chromatography on Q-
Sepharose Fast Flow resin, eluting with a gradient of triethyl-
ammonium hydrogencarbonate buffer gave 9a as the triethyl-
ammonium salt, which eluted between 0.7 and 0.9 mol dm23
buffer. The structure of 9a was confirmed by 1H and 31P NMR
spectroscopy, and by negative ion FAB mass spectrometry§
before accurate quantification by total phosphate assay.10 The
larger protected dimers 8b, 8c and 8d were then synthesised by
reaction of 6 with bis(p-nitrophenylcarbonate)-PEGs 7b, 7c and
7d under the conditions established for 7a. In each case, the
product was purified by flash chromatography and its dimeric
structure was confirmed by 1H NMR spectroscopy before
deprotection as for 8a. Finally, purification of each dimer by ion
exchange chromatography as for 9a gave 9b, 9c and 9d as their
triethylammonium salts, which were all freely soluble in
water.
1 M. J. Berridge, Nature (London), 1993, 361, 315. For recent reviews,
see: R. A. Wilcox, W. U. Primrose, S. R. Nahorski and R. A. J. Challiss,
Trends Pharmacol. Sci., 1998, 19, 467; C. W. Taylor, Biochim. Biophys.
Acta, 1998, 1436, 19.
2 M. de Kort, A. R. P. M. Valentijn, G. A. van der Marel and J. H. van
Boom, Tetrahedron Lett., 1997, 38, 7629.
3 E. Katayama, H. Funahashi, T. Michikawa, T. Shiraishi, T. Ikemoto, M.
Iino, K. Hirosawa and K. Mikoshiba, EMBO J., 1996, 15, 4844.
4 R. H. Kramer and J. W. Karpen, Nature (London), 1998, 395, 710.
5 M. Hirata, F. Yanaga, T. Koga, T. Ogasawara, Y. Watanabe and S.
Ozaki, J. Biol. Chem., 1990, 265, 8404; B. V. L. Potter and D. Lampe,
Angew. Chem., Int. Ed. Engl., 1995, 34, 1933.
6 A. M. Riley and B. V. L. Potter, Tetrahedron Lett., 1999, 40, 2213.
7 C. Malet and O. Hindsgaul, J. Org. Chem., 1996, 61, 4649.
8 D. Xu, K. Prasad, O. Repic and T. J. Blacklock, Tetrahedron Lett., 1995,
36, 7357.
9 R. Baker and J. L. Castro, J. Chem. Soc., Perkin Trans. 1, 1990, 47.
10 B. N. Ames and D. T. Dubin, J. Biol. Chem., 1960, 235, 769.
11 See for example: C. B. Baron, D. R. Tolan, K. H. Choi and R. F. Coburn,
Biochem. J., 1999, 341, 805; M. J. Bottomley, P. Lo Surdo and P. C.
Driscoll, Curr. Biol., 1999, 9, R301.
Thus we have demonstrated, for the first time, a viable
synthetic route to high molecular weight bivalent ligands as
984
Chem. Commun., 2000, 983–984