Teod,wJnm btus,
oo4o-4039193 $5.00 + .oo Pagamon Press Lad
Vol. 34. No. 2, pp. 219-222 1993
PAed in f&eat Britain
SYNTHESIS
AND
BINDING
STUDIES
OF AN OPTICALLY
1,4,5-TRIS(METHYLENESULFONlC
ACID)
PURE
ANALOGUE
HEXADEOXY-
OF
IPg
A. P. Kozikowski*a, V. I. Ognyanov? C. Chen? P. Kurianh, and F. T. Crew& aNeurochemistry Research, Mayo Foundation for Education and Research, 4500 San Pablo Road, Jacksonville, Florida 32224, and huniversity of Florida, Box J-267, JHM Heatth Center, Gainesville,
Abstract:
Florida 3261 o-0267.
The synthesis of the (-)-sodium salt of the hexadeoxy-1,4,5_trfs(methylene-
sulfonic acid) analogue of IPs by use of an asymmetric Diels-Alder reaction is described together with the effects of this compound in binding to the IPs receptor. The seminal investigation
idea that opened
the area of myo-inositol
signalling
for
intensive
came in 1975 with Michell’s suggestion that the turnover of phosphoinositides
coupled cell surface receptors to cellular Ca*+ mobilization. 1 Subsequently, Berridge was led to conclude that the water soluble product of Pl(4,5)P2 hydrolysis, likely
intracellular
candidate
that released
namely IP(1,4.5)3, was the
Ca *+ from intracellular
stores and increased
cytoplasmic free Ca*+ concentrations. 2 By a series of events that are still not fully understood, but probably involve a network of interacting transcription
protein kinases and phosphorylation
factors,3 the increase in intracellular
leads to transcriptional
of nuclear
Ca*+ together with the activation
activation of genes responsible for a biological response.4ls
of PKC
In an effort
to identify IPQ-like molecules that might find therapeutic application in the treatment of certain neurodegenerative
disorders such as Alzheimer’s disease,6 a disease which may at least in part
owe its etiology to improper calcium signalling, we have become interested in the creation of certain metabolically stable isosteres of IPQ. Herein we report the synthesis of a molecule which represents
a simplified
methylenesulfonic
version
of IPs, namely the hexadeoxy
acid groups in place of the phosphate
analogue
of IP3 bearing
groups of the parent structure.
Although the 6-hydroxy group of IPs does play a more prominent role in Ca*+ release than either of the hydroxy groups at positions 2 or 3,7,* we were interested in constructing this minimally functionalized
compound first in order to properly set the stage for further mapping the SAR
requirements of the IPa recognition site. Synthesis of the required sulfur analogue began by employing the di-(+)-menthyl ester of fumaric acid9 1 (Scheme 1) as the Diels-Alder dienophile in the cycloaddition benzyloxymethyl-1 ,Sbutadiene oxidationlo
(2). This diene was prepared from isoprene by selenium dioxide
followed by 0-benzylation.
using diisobutylaluminum
reaction with 2-
The cycloaddition
reaction was carried out at -40 “C
chloride as the homogeneous catalyst, the preferred catalyst in such 219
220
(Scheme 1.
Diastereofacially Selective DiebAlder Route to an Analoaue of IPI.
m&A’c’
1,
CH,OBn
3 (95 o/o;de=98%)
4 (41%) LiAIH4 1
~~ic$=BB.B) 6 (79%)
9 (68%)
8 (88%)
OH HO 3-03Po
2-oapo
-&
OPOi HO
natural IP3
1
221
asymmetric Die&Alder cyclohexene
reactions as elegantly described by Yamamoto and coworkers.*1
The
3 was formed in 95% yield which by HPLC determination was found to possess a
diastereomeric purity of 98%. the chromatographically
Next the cyclohexene
3 was hydrogenated over Pd/C to deliver
separable compounds 4 and 5, with the former isomer predominating.
Reduction of these diesters individually
with lithium aluminum hydride (LAH) then provided 6
and 7. The assignment of stereochemistry to these cyclohexanes tsC NMR chemical shift of the hydroxymethylene of steric compression,13
the axial hydroxymethylene
strength than the equatorial chemical shifts).
C-l
was made by examining the
group at the C-l postion.
hydroxymethylene
group of 6 (see structures 6 and 7 for
Thus, the major product of the hydrogenation
synthesis of the target compound.
As a consequence
group of 7 appears at a higher field
After some experimentation,
reaction was taken on to the we found that sulfur was best
introduced into this molecule by the Mitsunobu reaction invoiving use of thiolacetic acid.14 The desired tris-thiolacetate
6 was obtained in 68% yield.
Interestingly, when this compound was
deprotected using catalytic amounts of sodium methoxide in methanol, the disulfide 9 could be isolated in 88% yield.
Therefore,
in order to avoid potential
problems in the subsequent
oxidation of this disulfide 9 to sulfonic acid, 8 was reduced directly to the his-thiol
10 by LAH
treatment, and 10 was then oxidized with concentrated nitric acid.15 The desired IPs mimic 11 was isolated as its tris-sodium salt by treatment with sodium hydroxide. The sodium salt was obtained as a white powder which exhibited an optical rotation of -24.5 (c 1.O, H20). To examine the ability of 11 to function as a mimic of IPs, we investigated its ability to displace [sH]IPs from rat cerebellar P2 membrane fraction. 1s As is apparent from the graph presented in Figure 1, 11 is a rather poor mimic of natural IP3, and exhibits a Ki of > 17 pM compared to a Kr of 10 nM for natural IPs.
Several factors are likely to contribute to the poor
binding affinity of this compound, which may include the following: (a) the lack of the hydroxy groups at positions 2, 3, and 8; (b) substitution of one of the oxygen atoms present in each of the three phosphate groups of IP3 by methylene; and (c) replacement of phosphate by sulfonic acid.17
While the synthesis
contributions
of additional
IP3 analogues
which each of the specific structural
will be required to sort out the
features delineated
above contribute
to
binding, we note that deletion of the 8-hydroxy group from IPs results in about a lOO-fold loss both in binding affinity and ability to release Ca s+ from rat parotid microsomal fraction.7 Deletion of the 3-hydroxy group, however, has been shown to provide an IPs analogue of comparable Can+ releasing activity. s Furthermore, a recent article has revealed that myo-inositol
1,4,5-tris-
sulphate is biologically inactive.16 In summation,
while the IP3 mimic disclosed
herein is not useful from a biological
standpoint, the chemical route developed may find application in the construction of improved compounds of this genre. Acknowledgements.
We are indebted to the National
Institute on Aging for their
support of this work. We also wish to thank Dr. Roberto Pellicciari for valuable discussions.
222
Figure 1. ConpetMor~ binding of [3H]IP3 with I I. Each assay tube contained12,000 dpm of PH]IP3, 250 pg of rat Cerebellaf9 membranefractionand 50 pL of I4 (0.06 to 2000 pmol) or 11 (0.6 to 106 pmol) in 100 mM Tris-HCI and 4 mM EDTA (pH = 7.8) In a final volume of 200 pL. Tubes were incubatedfor 30 min at 4 “C. Separationof boundand free liiand was achievedby rapidfiltration. The Ki was computedusingcomputerassisted curve fitting(McPherson,G. A., BIOSOFT, Cambridge,U.K., 1965).
6. 9. 10. 11. 12. 13.
14. 15. 16. 17. 16.
References and Notes Mkhell, R. H. Wiochem. Wiophys. Acta 1975, 415, 61-147. Berrkfge,M. J. Wiochem.J. 1963,212, 649-656. Bohmann,D. Cancer Cells 1990,2, 337-344. Feam, J. C.; King, A. C. Ce111965, 40, 991-1000. Kelly, K.; Co&ran. B. H.; Stiles, C. D.; Leder .P. Cell 1963, 35, 603-610. Khachaturian,2. Neumbbbgy of Aging 1992, 13, 197-196. B_Deoxv:Bonis, D.; Sezan, A.; Mat&it. P.; Dubreuil,D.; Cleophax,J.; Gem, S. D.; Rossignol,B. Wiochem. Wiophys. Research Commun. 1991, 775, 694-900; Safrany, S. T.; Wojcikiewicz,R. J. H.; Strupish,J.; Dubreuil,D.; Cleophax,J.; Gero, S. D.; Nahorski,S. Ft.; Potter, B. V. L. FfWS Lett. 1991, 278, 252-256; P_Deoxv:Hirata, M.; Watanabe,Y.; Ishimatsu,T.; Ikebe, T.; Kimura,Y.; Yamaguchi.K.; Ozaki, S.; Koga,T. J. Wiol.Chem. 1969,284, 20303-20306. 8_Deoxv: Kozlkowski,A. P.; Fauq, A. H. J. Chem. Sot., Chem. Commun.1990, 1636. Heathcock,C. H.; Davis, B. Ft.;Hadley, C. R. J. Med. Chem. 1969.32, 197-202. Tanaka,J.; Suzuki.T.; Takabe, K.; Katagirl,T. NipponKagaku Kaishi1973, 2, 292-295. Furuta.K.; lwanaga, K.; Yamamoto,H. TetrahedronLett. 1966, 27, 4507-4510. Confirmationof the optical purity of the diil prepared by LAH reductionof 3 was obtained by HPLC analysisof the MPTA estersas describedin reference11. Dalling, D. K.; Grant, D. M. J. Am. Chem. Sot. 1972, 94, 5316-5324. In addition,alcohols4 and 5 were transformedto their correspondingaldehydesby Swem oxidationwhich proceededwithoutepimerfzationin each case. The lH NMR spectrumof the aklehyde derivedfrom 5 revealeda signalfor the protonat the akiehydebearingcarbonat 6 = 2.52. This signalappeared as a quintetwith J= 4.7 Hz, thus reflectingits equatorialnature. No comparablesignal was observedin the spectrumof the aldehyde derived from 4, for this axially Situatedprotonwould be expectedto resonateat higherfield strengthsand thus to overlap with other signals. The trans-diequatorialnature of the alkoxycarbonyl roups of 5 was also readily apparent from an examinationof the appropriatering protonscentered at B = 2.73 (ABq, JAB = 9.7 Hz, with both parts split into dd with J= 9.2, 4.1 and 9.2, 3.6 Hz, respectively). The aldehydesderivedfrom 4 and 5 were further shown to be equilibratedby DBU/CD$N treatmentto afford a 3:l mixture in which the aldehydederivedoriginallyfrom 4 predominated. Volante, R. P. Tetrahedmn Len. 1961,22, 3119-3122. Bauer, L.; Oardella, L. A. J. 0~. Chem. 1961, 28, 62-65. Maki, T.; Kowatch,M. A.; Baum, B. J.; Ambudkar.I. S.; Roth, 0. S. Wiochim. Wiophys. Acta 1969, 7074.7377. Ab initii calculationsreveal a stereochemicaldifferentiationof hydrogenbondingto sulfate and phosphate groups. See, inter alia, Thatcher, G. R.; Cameron, D. R.; Nagelkerke, R.: Schmitke, J. J. Chem. Sot., Chem. Commun. 1992, 366-366. Willems. H. A. M.: Veeneman. G. H.: Westerduin.P. Tetrahedron Lett. 1992, 33, 2075-2076; Westerduin, Reitz, A. B., Ed., ACS Symposium P.; van .Boeckel,6. A. A. in inosifoi Phosphates and Derivatives, Series, 1991, 483, 111-121.
(Received in USA 27 July 1992; accepted8 October 1992)