Molecular Recognition at the Phosphatidylinositol 3,4,5-Trisphosphate-Binding Site. Studies Using the Permuted Isomers of Phosphatidylinositol Trisphosphate

Article abstract of DOI:10.1021/jo980356h

Permuted isomers of L-α-phosphatidyl-D-myo-inositol trisphosphate (PtdInsP₃), including PtdIns(3,4,5)P₃, PtdIns(3,4,6)P₃, PtdIns(3,5,6)P₃, and PtdIns(4,5,6)P₃, have been synthesized as part of our effort to understand the underlying principles governing ligand selection for PtdIns(3,4,5)P₃-specific binding proteins. These PtdInsP₃ isomers are examined by using two PtdIns(3,4,5)P₃-dependent functional assays: binding to the C-terminal SH2 domain of the p85 regulatory subunit of PI 3-kinase and platelet aggregation. Our data show that all these isomers bind to the SH2 domain with comparable affinity despite variation in the regioisomeric distribution of phosphate functions. Moreover, all these phospholipids are capable of triggering platelet aggregation with the relative potency of PtdIns(3,4,5)P₃ > PtdIns(3,5,6)P₃ > PtdIns(4,5,6)P₃ > PtdIns(3,4,6)P₃. Evidence suggests that these PtdInsP₃'s facilitate cell aggregation by activating Ca²⁺ influx across the plasma membrane. In contrast, other inositol lipids examined including PtdIns(3,4)P2, PtdIns(4,5)P₂, PtdIns(3)P, and PtdIns(4)P are ineffective in eliciting the aggregation even at much higher concentrations. Taken together, the present data suggest that the charge density on the phosphorylated inositol ring represents a key factor in determining the phosphoinositide binding specificity of target proteins. It is conceivable that the interaction with the PtdIns(3,4,5)P₃-binding motif requires the participation of all three phosphates on the headgroup of PtdIns(3,4,5)P₃. Consequently, other membrane phosphoinositides (e.g., the bis- and monophosphates) become thermodynamically unfavorable for the binding to these PtdIns(3,4,5)P₃ targets.

Full text of DOI:10.1021/jo980356h

5430
J . Org. Chem. 1998, 63, 5430-5437
Molecu la r Recogn ition a t th e P h osp h a tid ylin ositol
3,4,5-Tr isp h osp h a te-Bin d in g Site. Stu d ies Usin g th e P er m u ted
Isom er s of P h osp h a tid ylin ositol Tr isp h osp h a te
Da-Sheng Wang, Ao-Lin Hsu, Xueqin Song, Chi-Ming Chiou, and Ching-Shih Chen*
Division of Medicinal Chemistry and Pharmaceutics, College of Pharmacy, University of Kentucky,
Lexington, Kentucky 40536-0082
Received February 26, 1998
Permuted isomers of L-R-phosphatidyl-D-myo-inositol trisphosphate (PtdInsP3), including PtdIns-
(3,4,5)P3, PtdIns(3,4,6)P3, PtdIns(3,5,6)P3, and PtdIns(4,5,6)P3, have been synthesized as part of
our effort to understand the underlying principles governing ligand selection for PtdIns(3,4,5)P3-
specific binding proteins. These PtdInsP3 isomers are examined by using two PtdIns(3,4,5)P3-
dependent functional assays: binding to the C-terminal SH2 domain of the p85 regulatory subunit
of PI 3-kinase and platelet aggregation. Our data show that all these isomers bind to the SH2
domain with comparable affinity despite variation in the regioisomeric distribution of phosphate
functions. Moreover, all these phospholipids are capable of triggering platelet aggregation with
the relative potency of PtdIns(3,4,5)P3 > PtdIns(3,5,6)P3 > PtdIns(4,5,6)P3 > PtdIns(3,4,6)P3.
Evidence suggests that these PtdInsP3’s facilitate cell aggregation by activating Ca2+ influx across
the plasma membrane. In contrast, other inositol lipids examined including PtdIns(3,4)P2, PtdIns-
(4,5)P2, PtdIns(3)P, and PtdIns(4)P are ineffective in eliciting the aggregation even at much higher
concentrations. Taken together, the present data suggest that the charge density on the
phosphorylated inositol ring represents a key factor in determining the phosphoinositide binding
specificity of target proteins. It is conceivable that the interaction with the PtdIns(3,4,5)P3-binding
motif requires the participation of all three phosphates on the headgroup of PtdIns(3,4,5)P3.
Consequently, other membrane phosphoinositides (e.g., the bis- and monophosphates) become
thermodynamically unfavorable for the binding to these PtdIns(3,4,5)P3 targets.
In tr od u ction
tified
a
unique PtdIns(3,4,5)P3-binding peptide,4
WAAKIQASFRGHMARKK (single-letter amino acid code),
to which many PtdIns(3,4,5)P3-binding SH2 domains
contain homologous internal sequences (Chen, C.-S.,
submitted). This model peptide possesses discrete apolar
and polybasic segments, confirming that the interfacial
binding to PtdIns(3,4,5)P3 micelles entails both hydro-
phobic and electrostatic interactions.4 It is worthy to note
that all PtdIns(3,4,5)P3-binding proteins examined so far
exhibit a certain degree of tolerance with respect to the
aliphatic chain length. Generally speaking, the C8- and
C4-analogues of PtdIns(3,4,5)P3 retain the activity of their
C16-counterpart. However, the deacylated product cannot
interact with PtdIns(3,4,5)P3 targets in a productive
manner.5 This observation suggests that while hydro-
phobic interactions with the fatty-acyl chains are impor-
tant to the binding, electrostatic bonding plays a rela-
tively more prominent role in the ligand selection. Thus,
to gain insight into the mode of interaction between the
highly charged headgroup and its binding domain, we
have synthesized four permuted analogues of phosphati-
dylinositol trisphosphates (PtdInsP3) to scrutinize the
regioisomeric effect of phosphate functions on two PtdIns-
(3,4,5)P3-dependent processes: binding to the p85 C-
terminal SH2 domain6 and platelet activation through
the activation of the putative Ins(1,3,4,5)P4 receptor.7 The
present study indicates that the charge density on the
It is well documented that the lipid products of phos-
phoinositide 3-kinase (PI 3-kinase) exert their physi-
ological functions by recruiting and/or activating target
proteins at the plasma membrane.1 Among these targets
are several Src homology-2 (SH2) domain-containing
proteins (such as the p85 regulatory subunit of PI
3-kinase1a and PLC-γ12) that specifically bind phosphati-
dylinositol 3,4,5-trisphosphate [PtdIns(3,4,5)P3] and Akt/
PKB that specifically binds phosphatidylinositol 3,4-
bisphosphate [PtdIns(3,4)P2] via its Pleckstrin (PH)
domain.1c Considering the diversity of membrane phos-
pholipids, selective recognition of the D-3 phosphoinositi-
des represents a crucial issue in addressing the functional
role of these signaling enzymes. In the literature, several
approaches have been undertaken to explore the under-
lying mechanism of this discriminative binding. Prestwich
and co-workers successfully developed 4-benzoyldihydro-
cinnamoyl-containing photoaffinity probes to map the
phosphoinositide-binding domains.3 Previously, we iden-
* To whom correspondence should be addressed. Tel.: (606) 257-
2300 (ext. 261). Fax: (606) 257-2489. E-mail:cchen1@pop.uky.edu.
(1) Reviews: (a) Carpenter, C. L.; Cantley, L. C. Curr. Opin. Cell
Biol. 1996, 8, 153. (b) Toker, A.; Cantley, L. C. Nature 1997, 387, 673.
(c) Franke, T. F.; Kaplan, D. R.; Cantley, L. C. Cell 1997, 88, 435.
(2) Bae, Y. S.; Cantley, L. G. Cantley; Chen, C.-S.; Kim, S.-R.; Kwon,
K.-S.; Rhee, S. G. J . Biol. Chem. 1998, 273, 4465.
(3) (a) Prestwich, G. D. Acc. Chem. Res. 1996, 29, 503. (b) Prestwich,
G. D.; Dorman, G.; Elliot, J . T.; Marecak, D. M.; Chaudhary, A.
Photochem. Photobiol. 1997, 65, 222. (c) Tall, E.; Dorman, G.; Garcia,
P.; Runnels, L.; Shah, S.; Chen, J .; Profit, A.; Gu, Q. M.; Chaudhary,
A.; Prestwich, G. D.; Rebecchi, M. J . Biochemistry 1997, 36, 7239.
(4) Lu, P.-J .; Chen, C.-S. J . Biol. Chem. 1997, 272, 466.
(5) Hao, W.; Tan, Z.; Prasad, K.; Reddy, K. K.; Chen, J .; Prestwich,
G. D.; Falck, J . R.; Shears, S. B.; Lafer, E. M. J . Biol. Chem. 1997,
272, 6393.
(6) Rameh, L. E.; Chen, C.-S.; Cantley, L. C. Cell 1995, 83, 821.
S0022-3263(98)00356-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/03/1998

Phosphatidylinositol 3,4,5-Trisphosphate-Binding Site
J . Org. Chem., Vol. 63, No. 16, 1998 5431
a
Sch em e 1. Syn th esis of P td In s(3,4,6)P 3 a n d P td In s(3,5,6)P 3
a
Key: (a) Bu2SnO, AllBr, CsF; (b) NaH, BnBr; (c) AcCl/CH3OH-CH2Cl2; (d) Bu2SnO, PMBCl, CsF, NaI; (e) RhCl(PPh)3, DABCO, 1 N
HCl-acetone; (f) (BnO)2PN(i-Pr)2, 1H-tetrazole, m-CPBA; (g) TFA-CH2Cl2-CH3OH; (h) 11, 1H-tetrazole, m-CPBA; (i) palladium black/
H2 (55 psi); (j) Bu2SnO, BnBr, CsF; (k) NaH, AllBr; (l) AcCl/CH3OH-CH2Cl2.
phosphorylated inositol ring represents a key factor in
dictating the binding specificity.
of inositol phosphates and inositol lipids.8,9 The synthesis
of PtdIns(3,4,5)P3 from (-)-1 has been described else-
where9 and will not be elaborated here.
P td In s(3,4,6)P 3 a n d P td In s(3,5,6)P 3. The efficient
use of (-)-1 to prepare PtdIns(3,4,6)P3 and PtdIns(3,5,6)-
P3 was made of via key intermediates (-)-5 and (-)-17,
respectively.
For PtdIns(3,4,6)P3 synthesis, (-)-1 was first subjected
to di-O-allylation. Selective removal of the trans-cyclo-
hexylidene group to give (+)-2 was achieved by brief
exposure to CH3COCl/CH3OH at room temperature.8a
Since an extended exposure to the acid resulted in the
concomitant hydrolysis of the cis-1,2-ketal, the reaction
was carefully monitored by TLC and was stopped when
all the substrate disappeared. Stannylidene-activated
alkylation by reacting (+)-2 with n-Bu2SnO, followed by
allyl bromide and CsF, selectively introduced an allyl
function at the 6-OH over 5-OH,8a,9a,10 yielding (-)-3.
Subsequent 5-O-benzylation and removal of the cis-ketal
gave the diol (-)-5. Regioselective 1-O-p-methoxyben-
zylation of 5 via the aforementioned stannylidene activa-
tion followed by 2-O-benzylation afforded a fully protected
Resu lts a n d Discu ssion
Syn th esis of L-r-P h osph atidyl D-myo-In ositol Tr is-
p h osp h a tes. The four regioisomers were synthesized
using optically active 1 as precursors, i.e., PtdIns(3,4,5)-
P3, PtdIns(3,4,6)P3, and PtdIns(3,5,6)P3 from (-)-1 (Scheme
1) and PtdIns(4,5,6)P3 from (+)-1 (Scheme 2). Both
enantiomers were prepared by a facile enzymatic resolu-
tion and have versatile applications in the preparation
(8) (a) Gou, D.-M.; Liu, Y.-C.; Chen, C.-S. Carbohydr. Res. 1992, 234,
51. (b) Lu, P.-J .; Gou, D.-M.; Shieh, W.-R.; Chen, C.-S. Biochemistry
1994, 33, 11586.
(9) (a) Gou, D.-M.; Chen, C.-S. J . Chem. Soc., Chem. Commun. 1994,
2125. (b) Wang, D.-S.; Chen, C.-S. J . Org. Chem. 1996, 61, 5905.
(10) Desai, T.; Gigg, J .; Gigg, R.; Martin-Zamora, E. Carbohydr. Res.
1996, 296, 97.
(7) Lu, P.-J .; Hsu, A.-L.; Wang, D.-S.; Chen, C.-S. Biochemistry 1998,
in press.

5432 J . Org. Chem., Vol. 63, No. 16, 1998
Wang et al.
a
Sch em e 2. Syn th esis of P td In s(4,5,6)P 3
13. Subsequent 3-O-allylation gave (+)-14 that was
subjected to the selective removal of the trans-ketal and
di-O-allylation, in tandem, to furnish (-)-16. Hydrolysis
of the cis-1,2-di-O-cyclohexylidene group afforded the key
intermediate (-)-17. The synthetic route for PtdIns-
(3,5,6)P3 from 17 paralleled that converting (-)-5 to
PtdIns(3,4,6)P3. The overall yield of converting (-)-1 into
PtdIns(3,5,6)P3 was 22%.
P td In s(4,5,6)P 3. It has been demonstrated in the
literature that stannylidene-activated alkylation of 1 is
highly selective for the 6-OH over 1-OH.9b,11 Conse-
quently, (+)-24 was provided by reacting (+)-1 with
n-Bu2SnO, followed by allyl bromide and CsF. Subse-
quent 1-O-p-methoxybenzylation gave (-)-25. Selective
hydrolysis of the trans-ketal of 25 by a brief acid
treatment yielded (+)-26. The diol was subjected to 4,5-
di-O-allylation, followed by acid hydrolysis and bisben-
zylation, to afford the fully protected intermediate (+)-
29. Subsequent transformations leading to PtdIns(4,5,6)P3
were similar to that described above for converting (+)-7
to PtdIns(3,4,6)P3. The overall yield was 27% from (+)-
1.
Bin d in g to th e p 85 C-Ter m in a l SH2 Dom a in . SH2
domains are conserved sequences of approximately 100
amino acids that are able to bind phosphotyrosine in a
sequence-specific manner.12 Previously, Cantley and co-
workers reported that PtdIns(3,4,5)P3 bound directly and
selectively to the SH2 domains of the 85 kDa subunit of
PI 3-kinase.6 It is noteworthy that this binding provides
a negative feedback control of PI 3-kinase by disrupting
the binding of p85 to phosphotyrosine via the SH2
domains.6 In addition, such PtdIns(3,4,5)P3 recognition
is of implied significance to the membrane localization
or activation of SH2-containing enzymes. For example,
PtdIns(3,4,5)P3 has been demonstrated to activate PLC-γ
isozymes by interacting with their SH2 domains.2
The p85 subunit of PI 3-kinase contains two SH2
domains, designated as the N-terminal (NT) and C-
terminal (CT) SH2 domains.6 The CT-SH2 domain has
been reported to bind PtdIns(3,4,5)P3 with high affinity,
while its affinity with phosphatidylinositol bisphosphates
such as PtdIns(3,4)P2 and PtdIns(4,5)P2 was significantly
lower.6 Considering the stringent requirement of the
three phosphate groups for binding, we examined the
stereochemical effect on the protein-phospholipid inter-
action. The full-length CT-SH2 domain was prepared by
expression as a glutathione S-transferase (GST) fusion
protein in Escherichia coli according to a reported
procedure,6 followed by affinity purification and Factor
Xa hydrolysis to liberate the SH2 domain from the
immobilized GST. To assess the binding affinity with
individual PtdInsP3 isomers, equilibrium binding based
on the quenching of SH2 intrinsic tryptophan fluores-
cence was performed. These phospholipids alone did not
have appreciable emission (data not shown), but induced
a dose-dependent and saturable reduction in the fluo-
rescence intensity of the CT-SH2 domain in a similar
manner. Figure 1 A-D displays the spectral change
induced by PtdIns(3,4,5)P3, PtdIns(3,5,6)P3, PtdIns(3,4,6)-
P3, and PtdIns(4,5,6)P3, respectively. The dissociation
constants (Kd) were estimated from the linear relation-
ship between 1/[1 - (∆F/∆Fmax)] and [phosphoinositide-
a
Key: (a) Bu2SnO, AllBr, CsF; (b) NaH, PMBBr; (c) AcCl/
CH3OH-CH2Cl2; (d) NaH, AllBr; (e) AcCl/CH3OH-CH2Cl2; (f)
NaH, BnBr; (g) RhCl(PPh)3, DABCO, 1 N HCl-acetone; (h)
(BnO)2PN(i-Pr)2, 1H-tetrazole, m-CPBA; (i) TFA-CH2Cl2-CH3OH;
(j) 11, 1H-tetrazole, m-CPBA; (k) palladium black/H2 (55 psi).
intermediate (+)-7. Deallylation of 7 with RhCl(PPh3)3
and diazabicyclo[2.2.2]octane (DABCO) under reflux,
followed by 1 N HCl-acetone (1:9), gave the desired triol
(-)-8. Phosphorylation of 8 by reacting with dibenzyl
N,N-diisopropylphosphoramidite and 1H-tetrazole, fol-
lowed by m-CPBA, afforded the corresponding 3,4,6-tris-
(dibenzyl phosphate) (-)-9. The p-methoxybenzyl func-
tion was removed by trifluoroacetic acid (TFA)-CH2Cl2-
CH3OH (6:3:1; v/v) to furnish (-)-10 in order for the
subsequent formation of the phosphotriester linkage at
the C-1 position. Reaction of 10 with 1,2-di-O-palmitoyl-
sn-glycerol-3-(benzyl N,N-diisopropylphosphoramidite)
(11) in the presence of 1H-tetrazole, followed by m-CPBA,
gave the perbenzylated derivative (+)-12 that underwent
hydrogenolysis to afford PtdIns(3,4,6)P3 with an overall
yield of 30% from (-)-1.
Alternatively, treatment of (-)-1 with n-Bu2SnO fol-
lowed by benzyl bromide and CsF resulted in regioselec-
tive benzylation at the 4-OH over 3-OH,8b affording (+)-
(11) Yu, K.-L.; Fraser-Reid, B. Tetrahedron Lett. 1988, 29, 979.
(12) Reviews: (a) Schaffhausen, B. Biochim. Biophys. Acta 1995,
1242, 61. (b) Beattie, J . Cell Signal. 1996, 8, 75.

Phosphatidylinositol 3,4,5-Trisphosphate-Binding Site
J . Org. Chem., Vol. 63, No. 16, 1998 5433
F igu r e 1. Tryptophan fluorescence emission spectra of the
p85 CT-SH2 domain in the presence of various amounts of (A)
PtdIns(3,4,5)P3, (B) PtdIns(3,5,6)P3, (C) PtdIns(3,4,6)P3, and
(D) PtdIns(4,5,6)P3. Spectra were recorded with 12.5 µM p85
CT-SH2 domain according to the method described in the
Experimental Section. Molar ratios of the phospholipid to the
SH2 domain were (top to bottom) (A) for PtdIns(3,4,5)P3: 0:1,
1:1, 1.5:1, 2.5:1, 3.5:1, 4.5:1, 5.5:1, 6.5:1, and 7:1 (saturation);
(B) for PtdIns(3,5,6)P3: 0:1, 2:1, 2.5:1, 3.5:1, 4.5:1, 5.5:1, 6.5:
1, and 8:1 (saturation); (C) for PtdIns(3,4,6)P3: 0:1, 1:1, 2:1,
3:1, 4.5:1, 6.5:1, and 8:1 (saturation); (D) for PtdIns(4,5,6)P3:
0:1, 1:1, 1.5:1, 2.5:1, 3.5:1, 4.5:1, 6:1, 7:1, and 8:1 (saturation).
F igu r e 2. PtdInsP3-induced platelet aggregation. Washed
rabbit platelets were treated with (A) PtdIns(3,4,5)P3, (B)
PtdIns(3,5,6)P3, (C) PtdIns(3,4,6)P3, or (D) PtdIns(4,5,6)P3 at
80 µM (trace a), 40 µM (trace b), or 20 µM (trace c). The arrow
indicates the time point at which individual phosphoinositides
were added.
PtdIns(3,4)P2, PtdIns(4,5)P2, PtdIns(3)P, and PtdIns(4)P.
Evidence suggests that PtdIns(3,4,5)P3 triggered platelet
activation by facilitating transmembrane Ca2+ influx. It
is documented that platelets contain an Ins(1,3,4,5)P4-
mediated Ca2+ transport system on plasma membranes.13
Although the mechanism by which Ins(1,3,4,5)P4 operates
remains enigmatic, the putative Ins(1,3,4,5)P4 receptor
has been purified and characterized.13 PtdIns(3,4,5)P3
was able to compete with Ins(1,3,4,5)P4 for the receptor
binding with an identical potency.7 This cross-reactivity
may in part be attributable to the largely shared struc-
tural motifs between these two molecules. In fact, Ins-
(1,3,4,5)P4-binding sites on platelet plasma membranes
might represent PtdIns(3,4,5)P3 receptors since PtdIns-
(3,4,5)P3 contained Ins(1,3,4,5)P4 as its headgroup.13b
In this study, we examined the potency of permuted
PtdInsP3 analogues vis-a`-vis PtdIns(3,4,5)P3 in inducing
platelet activation. Washed rabbit platelets were exposed
to various exogenous phospholipids in the presence of 1
mM Ca2+. Figure 2 A-D displays the dose-dependent
effect of PtdIns(3,4,5)P3, PtdIns(3,5,6)P3, PtdIns(3,4,6)-
]total/(∆F/∆Fmax). The estimated Kd values were as fol-
lows: PtdIns(3,4,5)P3, 15.3 ( 1.7 µM; PtdIns(3,5,6)P3,
14.2 ( 2.5 µM; PtdIns(3,4,6)P3, 15.1 ( 4.4 µM; PtdIns-
(4,5,6)P3, 23.8 ( 3.0 µM. In contrast, the binding affinity
for PtdIns(3,4)P2 and PtdIns(4,5)P2 was at least 10-fold
lower (spectra not shown), indicating that the lipid
binding effect on SH2 fluorescence was specific.
These binding data indicate that the regioisomeric
arrangement of the phosphates did not have a significant
impact on the binding to the CT-SH2 domain. Appar-
ently, the charge density on the polyphosphorylated
inositol ring represented a key factor in the phosphoi-
nositide recognition. The effective binding requires the
presence of three adjacent phosphates, regardless of their
relative positions on the ring. This information is thus
of importance to the rational design of inhibitors to
intervene the binding of the p85 subunit to tyrosine-
phosphorylated proteins via the SH2 domain.
P la telet Activa tion . Previously, we reported that
exogenous PtdIns(3,4,5)P3-induced platelet aggregation
in a Ca2+ and dose-dependent manner.7 This platelet
activation was highly structurally specific and was not
noted with other inositol lipids examined including
(13) (a) Cullen, P. J .; Patel, Y.; Kakkar, V. V.; Irvine, R. F.; Authi,
K. S. Biochem. J . 1994, 298, 739. (b) Cullen, P. J .; Dawson, A. P.; Irvine,
R. F. Biochem. J . 1995, 305, 139. (c) Cullen, P. J .; Hsuan, J . J .; Truong,
O.; Letcher, A. J .; J ackson, T. R.; Dawson, A. P.; Irvine, R. F. Nature
1995, 376, 527. (d) O’Rourke, F.; Matthews, E.; Feinstein, M. B.
Biochem. J . 1995, 315, 1027-1034.

5434 J . Org. Chem., Vol. 63, No. 16, 1998
Wang et al.
DMF (4 mL) was treated with NaH (20.2 mg, 0.82 mmol) at 0
°C under argon for 30 min, followed by benzyl bromide (106
µL, 0.84 mmol). The reaction mixture was allowed to attain
40 °C and stirred for an additional 2 h. Excess NaH was
destroyed with CH3OH, and the solution was diluted with ethyl
acetate (25 mL), washed with water, dried, and concentrated.
Column chromatography (hexanes-ether, 25:1) of the residue
P3, and PtdIns(4,5,6)P3, respectively, on platelet aggrega-
tion (trace a, 80 µM; trace b, 40 µM, trace c, 20 µM).
Among these, the relative potency was in the order of
PtdIns(3,4,5)P3 > PtdIns(3,5,6)P3 > PtdIns(4,5,6)P3
>
PtdIns(3,4,6)P3. PtdIns(3,4,5)P3 was severalfold more
effective than other isomers in mediating the platelet
aggregation. It is worth mentioning that PtdIns(4,5)P2
and PtdIns(3,4)P2 at comparable concentrations failed to
elicit appreciable cell activation (data not shown). The
fact that all PtdInsP3’s tested were active in triggering
platelet activation while PtdIns(3,4)P2 and PtdIns(4,5)-
P2 were not active underscores the notion that the charge
density on the inositol ring was crucial to the receptor
binding. Moreover, in this case, the ligand specificity also
depended on the regioisomeric distribution of phosphate
functions on the inositol ring. It appears that the
phosphate function on the 5-OH played a major role in
the activity.
afforded (-)-4 (syrup, 194 mg, 98%): [R]23 ) -10.6° (c 3.1,
D
1
CHCl3); H NMR (CDCl3) δ 1.40-1.80 (m, 10 H), 3.31 (t, J )
9.6 Hz, 1 H), 3.57 (dd, J ) 3.6, 8.7 Hz, 1 H), 3.63 (dd, J ) 7.2,
9.6 Hz, 1 H), 3.74 (t, J ) 8.4 Hz, 1 H), 4.05 (dd, J ) 5.4, 6.9
Hz, 1 H), 4.23-4.40 (m, 7 H), 4.79 (s, 2 H), 5.15-5.35 (m, 6
H), 5.91-6.02 (m, 3 H), 7.26-7.39 (m, 5 H); MS (EI) m/z (rel
inten) 470 (M+, 0.1), 427 (6), 379 (2), 91 (100).
(-)-3,4,6-Tr i-O-a llyl-5-O-ben zyl-m yo-in ositol (5). A so-
lution of (-)-4 (157 mg, 0.33 mmol) in CH3OH-CH2Cl2 (1:3,
12 mL) was stirred with acetyl chloride (25 µL) at 23 °C for 1
h. Triethylamine (100 µL) was added, and the solution was
concentrated. Column chromatography (hexanes-ether, 10:1
f 1:3) of the residue yielded (-)-5 (amorphous, 114 mg, 87%):
1
[R]23 ) -6.9° (c 1.3, CHCl3); H NMR (CDCl3) δ 2.44 (br s, 2
D
Taken together, these observations suggest that the
ligand specificity of PtdIns(3,4,5)P3-binding proteins is
achieved through the participation of all three phosphate
functions of PtdIns(3,4,5)P3 in the binding. In principle,
a difference in the binding energy (∆∆Gb) of a few kcal/
mol will lead to a difference of several orders of magni-
tudes in the dissociation constants (Kd) [∆∆Gb ) -RT
ln(Kd/Kd′)]. Consequently, PtdIns(3,4)P2 and PtdIns(4,5)-
P2 become thermodynamically unfavorable for binding to
PtdIns(3,4,5)P3-specific targets. Further investigations
on the mode of interaction by NMR are currently in
progress in the laboratory.
H), 3.35 (dd, J ) 9, 18 Hz, 2 H), 3.47 (m, 1 H), 3.62-3.76 (m,
2 H), 4.19-4.45 (m, 7 H), 4.85 (dd, J ) 10.5, 29.1 Hz, 2 H),
5.20 (dd, J ) 9.6, 17.4 Hz, 3 H), 5.30 (t, J ) 17.4 Hz, 3 H),
5.91-6.01 (m, 3 H), 7.26-7.36 (m, 5 H); MS (EI) m/z (rel inten)
390 (M+, 1.5), 347 (12), 299 (6.5), 91 (100).
(-)-3,4,6-Tr i-O-a llyl-5-b en zyl-1-O-(p -m et h oxyb en zyl)-
m yo-in ositol (6). A solution of (-)-5 (110 mg, 0.28 mmol),
Bu2SnO (84 mg, 0.34 mmol), and toluene (25 mL) was stirred
under reflux, with azeotropic removal of water, for 2 h, and
then concentrated to dryness. To the residue were added DMF
(4 mL), CsF (107 mg, 0.71 mmol), NaI (5 mg, 0.03 mmol), and
4-methoxybenzyl chloride (61 µL, 0.42 mmol) at -15 °C. After
being stirred at -15 °C for 1 h, the reaction mixture was
allowed to warm to room temperature and stirred for an
additional 16 h. The solution was diluted with CH2Cl2 (20
mL), washed with water, dried, and concentrated. Column
chromatography (hexanes-ether, 25:1 f 10:1) of the residue
Exp er im en ta l Section
Ma ter ia l a n d Meth od s. Racemic 1,2:5,6-di-O-cyclohexy-
lidene-myo-inositol ((()-1) was prepared according to the
procedure by Garegg et al.14 Optically active 1 was prepared
with good yield by an enzymatic method in which the 6-butyryl
ester of 1 was subjected to enantioselective hydrolysis by
porcine pancreatic lipase (sigma, Type II) in a biphasic system
consisting of hexanes-ether/water.8 The optical purity of (+)-
and (-)-1 thus prepared was greater than 98% enantiomeric
excess after recrystallization. 1,2-Di-O-palmitoyl-sn-glycerol
3-(benzyl N,N-diisopropylphosphoramidite) (11) was prepared
using a method described by Dreef et al.15 (+)-3,4-Di-O-allyl-
1,2-O-cyclohexylidene-myo-inositol (+)-2 was synthesized from
(-)-1 as previously described.8a
provided (-)-6 (amorphous, 124 mg, 84%): [R]23 ) -1.3° (c
D
1
1.5, CHCl3); H NMR (CDCl3) δ 2.38 (br s, 1 H), 3.17 (d, J )
9.9 Hz, 1 H), 3.27 (t, J ) 9.6 Hz, 2 H), 3.68-3.73 (m, 2 H),
3.75 (s, 3 H), 4.10-4.12 (m, 4 H), 4.26-4.29 (m, 3 H), 4.59 (s,
2 H), 4.76 (d, J ) 0.9 Hz, 2 H), 5.06-5.26 (m, 6 H), 5.85-5.95
(m, 3 H), 6.84 (d, J ) 8.1 Hz, 2 H), 7.20-7.40 (m, 7 H); MS
(EI) m/z (rel inten) 509 (M+ - H, 2), 469 (0.5), 419 (0.7), 121
(100).
(+)-3,4,6-Tr i-O-a llyl-2,5-d i-O-b en zyl-1-O-(p -m et h oxy-
ben zyl)-m yo-in ositol (7). Conventional benzylation of (-)-6
(124.5 mg, 0.24 mmol), as described for (-)-4, gave (+)-7
(syrup, 140 mg, 96%): [R]23 ) +2.7° (c 2.2, CHCl3); 1H NMR
D
(-)-3,4,6-Tr i-O-a llyl-1,2-O-cycloh exylid en e-m yo-in osi-
tol (3). A mixture of (+)-2 (158 mg, 0.41 mmol), Bu2SnO (128
mg, 0.51 mmol), and toluene (25 mL) was stirred under reflux,
with azeotropic removal of water, for 2 h, and then concen-
trated to dryness. To the residue were added DMF (4 mL),
CsF (177 mg, 1.16 mmol), and allyl bromide (60.5 µL, 0.52
mmol) at -15 °C. After being stirred at -15 °C for 1 h, the
reaction mixture was allowed to warm to room temperature
and stirred for an additional 16 h. The solution was diluted
with CH2Cl2 (20 mL), washed with water, dried, and concen-
trated. Column chromatography (hexanes-ether, 20:1 f 10:
(CDCl3) δ 3.13 (dd, J ) 2.4, 9.9 Hz, 1 H), 3.25 (dd, J ) 2.4, 9.9
Hz, 1 H), 3.29 (t, J ) 9.0 Hz, 1 H), 3.73-3.84 (m, 5 H), 3.89 (t,
J ) 2.1 Hz, 1 H), 4.00-4.04 (m, 2 H), 4.22-4.34 (m, 4 H), 4.53
(q, J ) 11.4, 21.3 Hz, 2 H), 4.77 (s, 4 H), 5.06-5.13 (m, 3 H),
5.18-5.27 (m, 3 H), 5.78-5.98 (m, 3 H), 6.84 (d, J ) 8.7 Hz, 2
H), 7.20-7.35 (m, 12 H); MS (EI) m/z (rel inten) 559.2 (M+
41, 0.1), 509.2 (0.2), 91 (100).
-
(-)-2,5-Di-O-ben zyl-1-O-(p-m eth oxyben zyl)-m yo-in osi-
tol (8). A mixture of (+)-7 (80 mg, 0.13 mmol), RhCl(PPh3)3
(55.5 mg, 0.60 mmol), diazabicyclo[2.2.2]octane, and ethanol
(2 mL) was refluxed for 4 h and cooled to room temperature.
The solution was diluted with ether (20 mL), washed with
water, dried, and concentrated. The residue was mixed with
acetone-1 N HCl (9:1, 30 mL), refluxed for 30 min, and cooled
to room temperature, and NaHCO3 was added. The mixture
was concentrated, diluted with water, and extracted with ether
(3 × 10 mL). The organic phase was dried and concentrated.
Column chromatography (hexanes-ether, 10:1 f 0:1) of the
residue yielded (-)-8 (amorphous, 56 mg, 88%): [R]23D ) -2.9°
1) of the residue gave (-)-3 (syrup, 166 mg, 90%): [R]23
)
D
-2.3° (c 1.3, CHCl3); 1H NMR (CDCl3) δ 1.26-1.79 (m, 10 H),
2.62 (s, 1 H), 3.33 (t, J ) 9.6 Hz, 1 H), 3.48-3.60 (m, 3 H),
3.99 (t, J ) 6.6 Hz, 1 H), 4.13-4.21 (m, 4H), 4.26-4.36 (m, 3
H), 5.14 (d, J ) 9.9 Hz, 3 H), 5.26 (d, J ) 17.4 Hz, 3 H), 5.81-
5.97 (m, 3 H); MS (EI) m/z (rel inten) 380 (M+, 10), 337.0 (33),
41 (100).
(-)-3,4,6-Tr i-O-a llyl-5-O-ben zyl-1,2-O-cycloh exylid en e-
m yo-in ositol (4). A solution of (-)-3 (160 mg, 0.42 mmol) in
1
(c 2.0, CHCl3); H NMR (CDCl3) δ 2.30 (br s, 1 H), 2.54 (br s,
2 H), 3.21-3.31 (m, 2 H), 3.40-3.44 (m, 2 H), 3.82 (s, 3 H),
4.05-4.13 (m, 2 H), 4.52-4.82 (m, 4 H), 4.94-4.99 (m, 2 H),
6.88 (d, J ) 7.8 Hz, 2 H), 7.26-7.38 (m, 12 H); MS (EI) m/z
(rel inten) 479.9 (M+, 0.1), 388.9 (3), 268.9 (5), 91 (100).
(14) Garegg, P. J .; Iversen, T.; J ohansson, R.; Lindberg, B. Carbo-
hydr. Res. 1984, 130, 322.
(15) Dreef, C. E.; Elie, C. J . J .; Hoogerhout, P.; van der Marel, G.
A.; van Boom, J . H. Tetrahedron Lett. 1988, 29, 6513.

Phosphatidylinositol 3,4,5-Trisphosphate-Binding Site
J . Org. Chem., Vol. 63, No. 16, 1998 5435
(-)-2,5-Di-O-ben zyl-1-O-(p-m eth oxyben zyl)-m yo-in osi-
tol-3,4,6-tr is(d iben zyl p h osp h a te) (9). A solution of 1H-
tetrazole (159 mg, 0.23 mmol) and dibenzyl N,N-diisopropy-
lphosphoramidite (455 mg, 0.13 mmol) in dry dichloromethane
(4 mL) was stirred at 23 °C under argon for 30 min, and (-)-8
(50 mg, 0.10 mmol) was added in one portion. The mixture
was kept under the same conditions for an additional 4 h,
cooled to -40 °C, and then treated with m-CPBA (500 mg, 57%
purity, 0.94 mmol). The solution was stirred at -40 °C for 30
min and then allowed to attain room temperature. After 1 h,
the mixture was diluted with CH2Cl2 (20 mL), washed with,
in tandem, 10% aqueous Na2SO3, 10% aqueous NaHCO3, and
brine, dried, and concentrated. Column chromatography
ture and stirred for 16 h. The solution was then diluted with
CH2Cl2 (25 mL), washed with water, dried, and concentrated.
Column chromatography (hexanes-ether, 20:1 f 10:1) of the
residue afforded (+)-13 (syrup, 193 mg, 85%): [R]23 ) +0.8°
D
(c 2.0, CHCl3); 1H NMR (CDCl3) δ 1.41-1.75 (20 H), 2.65 (d, J
) 1.5 Hz, 1 H), 3.58 (dd, J ) 7.6, 10.5 Hz, 1 H), 3.93 (dd, J )
2, 7.8 Hz, 1 H), 4.06-4.07 (m, 1 H), 4.23 (dd, J ) 7.6, 10.5 Hz,
1 H), 4.38 (t, J ) 7.5 Hz, 1 H), 4.46 (dd, J ) 3.6, 7.5 Hz, 1 H),
4.75 (q, J ) 11.8, 35 Hz, 2 H), 7.28-7.43 (m, 5 H); MS (EI)
m/z (rel inten) 430 (M+, 6), 91 (100).
(+)-3-O-Allyl-4-O-ben zyl-1,2:5,6-d i-O-cycloh exylid en e-
m yo-in ositol (14). A solution of (+)-13 (180 mg, 0.42 mmol)
in DMF (3 mL) was treated with NaH (24 mg, 85% purity,
0.84 mmol) at 0 °C under argon for 30 min, followed by allyl
bromide (73 µL, 0.84 mmol), and then the reaction mixture
was allowed to attain 40 °C for 1 h. Excess NaH was destroyed
with CH3OH. The solution was diluted with ethyl acetate (25
mL), washed with water, dried, and concentrated. Column
chromatography (hexanes-ether, 50:1 f 20:1) of the residue
gave (+)-14 (syrup, 190 mg, 97%): [R]23D ) +6.3° (c 4, CHCl3);
1H NMR δ 1.34-1.73 (m, 20 H), 3.50 (dd, J ) 7.7, 10.7 Hz, 1
H), 3.69 (t, J ) 3.1 Hz, 1 H), 3.82 (dd, J ) 2.7, 7.7 Hz, 1 H),
4.01-4.12 (m, 3 H), 4.31 (t, J ) 7.0 Hz, 1 H), 4.39 (dd, J )
3.8, 6.9 Hz, 1 H), 4.75 (q, J ) 11.9, 43.3 Hz, 2 H), 5.13-5.31
(m, 2 H), 5.80-5.93 (m, 1 H), 7.25-7.39 (m, 5 H); MS (EI) m/z
(rel inten) 470.2 (M+, 0.5), 427.1 (6), 390.1 (10), 379.1 (4), 91
(100).
(hexanes-ether, 25:1 f 1:1) of the residue provided (-)-9
1
(syrup, 119 mg, 91%): [R]23 ) -0.1° (c 3, CHCl3); H NMR
D
(CDCl3) δ 3.32-3.40 (m, 1 H), 3.43-3.50 (m, 1 H), 3.80 (s, 3
H), 4.05-4.13 (m, 2 H), 4.53-4.99 (m, 20 H), 6.78 (d, J ) 7.8
Hz, 2 H), 6.94-7.36 (m, 42 H); 31P NMR (CDCl3, 85% H3PO4
as external standard) δ -1.30, -1.64, -1.75; HRMS (MALDI)
m/z (rel inten) 1261.405 (M + H, 100).
(-)-2,5-Di-O-ben zyl-3,4,6-tr is(diben zyl ph osph ate)-myo-
in ositol (10). Removal of the p-methoxybenzyl function was
achieved by exposing (-)-9 (110 mg, 0.083 mmol) to trifluo-
roacetic acid-CH2Cl2-CH3OH (6:3:1, 10 mL) at 0 °C for 1 h,
and 10% aqueous NaHCO3 was added. The organic layer was
washed with 10% aqueous NaHCO3 and water, dried, and
concentrated. Column chromatography (hexanes-ether, 5:1
f 1:2) of the residue afforded (-)-10 (syrup, 91 mg, 92%):
(-)-3-O-Allyl-4-O-b en zyl-1,2-O-cycloh exylid en e-m yo-
in ositol (15). A solution of (+)-14 (176 mg, 0.37 mmol) in
CH3OH-CH2Cl2 (1:3, 6 mL) was stirred with acetyl chloride
(8 µL) at 23 °C for 10 min. Triethylamine (30 µL) was then
added, and the solution was concentrated. Column chroma-
tography (hexanes-ether, 10:1 f 1:3) of the residue yielded
[R]23 ) -2.1° (c 2.0, CHCl3); 1H NMR (CDCl3) δ 3.32-3.40
D
(m, 1 H), 3.43-3.50 (m, 1 H), 4.05-4.13 (m, 2 H), 4.53-4.99
(m, 18 H), 6.94-7.36 (m, 40 H); 31P NMR (CDCl3, 85% H3PO4
as external standard) δ 1.08, -1.34, -1.62; HRMS (MALDI)
m/z (rel inten) 1141.339 (M + H, 100).
(+)-1-O-(1,2-Di-O-p a lm it oyl-sn -glycer ol-3-b en zyloxy-
ph osph or yl)-2,5-di-O-ben zyl-myo-in ositol-3,4,6-tr is(diben -
zyl p h osp h a te) (12). A solution of 1,2-di-O-palmitoyl-sn-
glycerol 3-(benzyl N,N-diisopropylphosphoramidite) (11) (56
mg, 0.069 mmol) and 1H-tetrazole (8.2 mg, 0.053 mmol) in dry
CH2Cl2 (2 mL) was stirred at 23 °C under argon for 30 min,
and (-)-10 (20 mg, 0.018 mmol) was added in one portion. The
mixture was kept under the same conditions for an additional
4 h, cooled to -40 °C, and treated with m-CPBA (30 mg, 57%
purity, 0.056 mmol). The mixture was stirred at -40 °C for
30 min and then allowed to attain room temperature for 1 h,
diluted with CH2Cl2 (20 mL), washed with, in tandem, 10%
aqueous Na2SO3, 10% aqueous NaHCO3, and brine, dried, and
concentrated. Column chromatography (hexanes-ether, 25:1
(-)-15 (amorphous, 117 mg, 80%): [R]23 ) -3.0° (c 2.0,
D
CHCl3); 1H NMR (CDCl3) δ 1.30-1.80 (m, 10 H), 2.74 (br s, 2
H), 3.32 (t, J ) 8.6 Hz, 1 H), 3.63-3.76 (m, 3 H), 3.98 (dd, J
) 2.2, 7.5 Hz, 1 H), 4.21-4.25 (m, 2 H), 4.44 (dd, J ) 3.8, 5.2
Hz, 1 H), 4.71 (d, J ) 11.1 Hz, 1 H), 5.01 (d, J ) 11.1 Hz, 1 H),
5.19-5.35 (m, 2 H), 5.90-6.03 (m, 1 H), 7.26-7.37 (m, 5 H);
MS (EI) m/z (rel inten) 389.1 (M+ - H, 2), 299.1 (12), 91 (100).
(-)-3,5,6-Tr i-O-a llyl-4-O-ben zyl-1,2-O-cycloh exylid en e-
m yo-in ositol (16). A solution of (-)-15 (100 mg, 0.26 mmol)
in DMF (3 mL) was treated with NaH (30 mg, 85% purity,
1.04 mmol) at 0 °C under argon for 30 min, followed by allyl
bromide (90 µL, 1.04 mmol). The reaction mixture was allowed
to warm to 40 °C and stirred for 1 h. Excess NaH was
destroyed with CH3OH. The mixture was diluted with ethyl
acetate (25 mL), washed with water, dried, and concentrated.
Column chromatography (hexanes-ether, 50:1 f 20:1) of the
f 1:1) of the residue provided (+)-12 (syrup, 28 mg, 86%):
1
[R]23 ) +0.2° (c 2.0, CHCl3); H NMR (CDCl3) δ 0.88 (t, J )
D
6.6 Hz, 6 H), 1.21-1.28 (m, 48 H), 1.40-1.58 (m, 4 H), 2.15-
2.24 (m, 4 H), 3.47-3.60 (m, 2 H), 3.96-4.44 (m, 6 H), 4.68-
5.08 (m, 21 H), 6.85-7.40 (m, 45 H); 31P NMR (CDCl3, 85%
H3PO4 as external standard) δ -0.351 (2P), -0.89 (br, 2 P);
HRMS (MALDI) m/z (rel inten) 1862.858 (M + 2 H, 100).
(+)-L-r-P h osph atidyl-D-myo-in ositol 3,4,6-Tr iph osph ate,
Dip a lm itoyl [P td In s(3,4,6)P 3]. A solution of (+)-12 (28 mg,
0.015 mmol) and palladium black (20 mg) in 80% aqueous
ethanol (2 mL) was shaken under H2 (55psi) for 16 h, filtered,
and concentrated. The filtrate was lyophilized to furnish
residue provided (-)-16 (syrup, 114 mg, 95%): [R]23 ) -4.2°
D
(c 2, CHCl3); 1H NMR (CDCl3) δ 1.37-1.78 (m, 10 H), 3.21 (dd,
J ) 8.8, 9.7 Hz, 1 H), 3.55-3.63 (m, 2 H), 3.81 (t, J ) 8.7 Hz,
1 H), 4.02 (dd, J ) 5.4, 7.0 Hz, 1 H), 4.21-4.28 (m, 5 H), 4.31-
4.38 (m, 2 H), 4.80 (q, J ) 10.7, 15.8 Hz, 2 H), 5.13-5.32 (m,
6 H), 5.89-6.01 (m, 3 H), 7.28-7.40 (m, 5 H); MS (EI) m/z (rel
inten) 470 (M+, 2), 380 (12), 91 (100).
(-)-3,5,6-Tr i-O-a llyl-4-O-ben zyl-m yo-in ositol (17). Re-
moval of the cis-cyclohexylidene group of (-)-16 (100 mg, 0.21
mmol) with acetyl chloride, as described for (-)-5, yielded (-)-
PtdIns(3,4,6)P3 (lyophilized powder, 15.5 mg, 98%): [R]23
)
17 (amorphous, 75 mg, 90%): [R]23 ) -6.0° (c 2, CHCl3); H
1
D
D
1
+3.6° (c 6, CHCl3); H NMR (CDCl3) δ 0.96 (t, J ) 7.2 Hz, 6
H), 1.31 (s, 48 H), 1.52-1.68 (m, 4 H), 2.30-2.37 (m, 4 H),
3.45 (t, J ) 9.3 Hz, 1 H), 3.86 (t, J ) 9.6 Hz, 1 H), 4.02-4.25
(m, 5 H), 4.41-4.57 (m, 3 H), 5.22-5.30 (m, 1 H); 31P NMR
(CDCl3, 85% H3PO4 as external standard) δ -0.715 (2P),
-1.063 (1P), -2.93 (1P); HRMS MALDI) m/z (rel inten)
1049.395 (M - H, 100).
NMR (CDCl3) δ 2.61 (d, J ) 6.7 Hz, 2 H), 3.25 (t, J ) 9.4 Hz,
1 H), 3.32 (dd, J ) 2.9, 9.5 Hz, 1 H), 3.38-3.46 (m, 1 H), 3.64
(t, J ) 9.5 Hz, 1 H), 3.81 (t, J ) 9.5 Hz, 1 H),4.17-4.46 (m, 7
H), 4.81 (q, J ) 10.6, 16.7 Hz, 2 H), 5.13-5.33 (m, 6 H), 5.86-
6.03 (m, 3 H), 7.28-7.39 (m, 5 H); MS (EI) m/z (rel inten) 389.1
(M+, 1), 299.1 (2), 91 (100).
(+)-3,5,6-Tr i-O-allyl-4-O-ben zyl-1-O-(p-m eth oxyben zyl)-
m yo-in ositol (18). Regioselective introduction of a p-meth-
oxybenzyl group to the C-1 of (-)-17 (69 mg, 0.018 mmol), as
(+)-4-O-Ben zyl-1,2:5,6-d i-O-cycloh exylid en e-m yo-in osi-
tol (13). A mixture of (-)-1 (180 mg, 0.53 mmol), Bu2SnO
(144.7 mg, 0.59 mmol), and toluene (25 mL) was stirred under
reflux, with azeotropic removal of water, for 2 h and then
concentrated to dryness. To the residue were added DMF (4
mL), CsF (201 mg, 1.34 mmol), and benzyl bromide (94 µL,
1.13 mmol) at -15 °C. After being stirred at -15 °C for 1 h,
the reaction mixture was allowed to warm to room tempera-
described for (-)-6, afforded (+)-18 (syrup, 83 mg, 92%): [R]23
D
1
) +0.5° (c 2, CHCl3); H NMR (CDCl3) δ 2.43 (s, 1 H), 3.14
(dd, J ) 2.5, 10 Hz, 1 H), 3.22-3.34 (m, 2 H), 3.79-3.82 (m, 4
H), 3.90-3.97 (m, 2 H), 4.04-4.06 (m, 2 H), 4.23-4.34 (m, 4
H), 4.50-4.57 (m, 2 H), 4.79-4.88 (m, 4 H), 5.14-5.19 (m, 2
H), 5.21-5.29 (m, 2 H), 5.83-6.03 (m, 3 H), 6.81-6.85 (m, 2

5436 J . Org. Chem., Vol. 63, No. 16, 1998
Wang et al.
H), 7.24-7.40 (m, 7 H); MS (EI) m/z (rel inten) 509.2 (M+
-
The reaction mixture was warmed to 40 °C and stirred for 2
h. The excess NaH was destroyed with CH3OH. The solution
was diluted with ethyl acetate (50 mL), washed with water,
dried, and concentrated. Column chromatography (hexanes-
ether, 50:1 f 20:1) of the residue gave (-)-25 (225 mg, 95%):
H, 0.2), 91 (100).
(+)-3,5,6-Tr i-O-a llyl-2,4-d i-O-b en zyl-1-O-(p -m et h oxy-
ben zyl)-m yo-in ositol (19). Conventional benzylation of (+)-
18 (72 mg, 0.014 mmol), as described for (-)-4, gave (+)-19
(syrup, 81 mg, 96%): [R]23 ) +4.4° (c 1, CHCl3); 1H NMR
[R]23 ) -5.7° (c 1.4, CHCl3); 1H NMR (CDCl3) δ 1.30-1.77
D
D
(CDCl3) δ 3.16 (dd, J ) 2.3, 9.6 Hz, 1 H), 3.23-3.31 (m, 2 H)
3.76-3.86 (m, 4 H), 3.90-3.97 (m, 2 H), 4.04-4.07 (m, 2 H),
4.25-4.34 (m, 4 H), 4.50-4.57 (m, 2 H), 4.77-4.17 (m, 4 H),
5.12-5.15 (m, 3 H), 5.23-5.29 (m, 3 H), 5.83-6.02 (m, 3 H),
6.81-6.86 (m, 2 H), 7.24-7.40 (m, 12 H); MS (EI) m/z (rel
inten) 600 (M+, 0.1), 559 (0.4), 91 (100).
(m, 20 H), 3.21-3.36 (m, 1 H), 3.55-3.80 (m, 5 H), 3.84-4.11
(m, 2 H), 4.20-4.35 (m, 2 H), 4.62-4.66 (m, 2 H), 5.12-5.20
(m, 1 H), 5.28 (d, J ) 3 Hz, 2 H), 5.76-6.02 (m, 1 H), 6.87 (d,
J ) 13.2 Hz, 2 H), 7.30 (d, J ) 13.2 Hz, 2 H); MS (EI) m/z (rel
inten) 500.2 (M+, 0.1), 457.2 (1), 379.2 (0.1), 121 (100).
(+)-6-O-Allyl-1-O-(p-m eth oxyben zyl)-2,3-O-cycloh exy-
lid en e-m yo-in ositol (26). Removal of the trans-cyclohexy-
lidene group of (-)-25 (195 mg, 0.39 mmol) with acetyl
chloride, as described for (-)-15, afforded (+)-26 (syrup, 126
(-)-2,4-Di-O-ben zyl-1-O-(p -m eth oxyben zyl)-m yo-in osi-
tol (20). Deallylation of (+)-19 (76 mg, 0.013 mmol) with
RhCl(PPh3)3, as described for (-)-8, yielded (-)-20 (amorphous,
44 mg, 72%): [R]23 ) -0.4° (c 2, CHCl3); H NMR (CDCl3) δ
mg, 77%): [R]23 ) +6.5° (c 3, CHCl3); 1H NMR (CDCl3) δ
1
D
D
1.85 (br s, 3 H), 3.46 (t, J ) 9.2 Hz, 1 H), 3.55 (m, 1 H), 3.73-
3.82 (m, 4 H), 3.88 (t, J ) 9.2 Hz, 3 H), 4.06 (t, J ) 2.7 Hz, 1
H), 4.65-4.87 (m, 5 H), 6.81-6.85 (m, 2 H), 7.25-7.40 (m, 12
H); MS (EI) m/z (rel inten) 479.0 (M+ - H, 0.5), 389.0 (3), 359.0
(2), 91 (100).
1.30-1.73 (m, 10 H), 2.01 (br s, 2 H), 3.21-3.30 (m, 1 H), 3.40-
3.90 (m, 7 H), 4.10-4.30 (m, 2 H), 4.41-4.50 (m, 1 H), 4.67 (s,
2 H), 5.18-5.34 (m, 2 H), 5.85-6.10 (m, 1 H), 6.90 (d, J )
12.1 Hz, 2 H), 7.29 (d, J ) 12.2 Hz, 2 H); MS (EI) m/z (rel
inten) 420.1 (M+, 0.2), 377.1 (1), 299.1 (0.5), 121 (100).
(-)-4,5,6-Tr i-O-a llyl-1-O-(p-m eth oxyben zyl)-2,3-cyclo-
h exylid en e-m yo-in ositol (27). Allylation of (+)-26 (126 mg,
0.30 mmol) with NaH and allyl bromide, as described for (+)-
14, yielded (-)-27 (syrup, 142.5 mg, 95%): [R]23D ) -9.0° (c 1,
(+)-2,4-Di-O-ben zyl-1-O-(p-m eth oxyben zyl)-m yo-in osi-
tol 3,5,6-Tr is(d iben zyl p h osp h a te) (21). Phosphorylation
of (-)-20 (40 mg, 0.008 mmol) via the phosphoramidite method,
as described for (-)-9, yielded (+)-21 (syrup, 90 mg, 86%):
1
[R]23 ) +0.5° (c 2, CHCl3); 1H NMR (CDCl3) δ 3.43-3.50 (m,
CHCl3); H NMR (CDCl3) δ 1.30-1.78 (m, 10 H), 3.12 (t, J )
D
1 H), 3.75 (s, 3 H), 4.06-4.13 (m, 1 H), 4.24-5.03 (m, 22 H),
6.74-6.78 (m, 2 H), 6.90-7.40 (m, 42 H); 31P NMR (CDCl3,
85% H3PO4 as external standard) δ -1.45, -1.61, -1.86;
HRMS (MALDI) m/z (rel intel) 1261.392 (M + H, 100).
(+)-2,4-Di-O-b en zyl-m yo-in osit ol 3,5,6-Tr is(d ib en zyl
p h osp h a te) (22). Removal of the p-methoxybenzyl function
of (+)-21 (80 mg, 0.006 mmol) with TFA, as described for (-)-
10.2 Hz, 1 H), 3.49-3.57 (m, 2 H), 3.67 (t, J ) 10.2 Hz, 1 H),
3.79 (s, 3 H), 3.90 (dd, J ) 8.1, 10.5 Hz, 1 H), 4.12-4.66 (m, 7
H), 4.70 (q, J ) 19.8, 33 Hz, 2 H), 5.10-5.32 (m, 6 H), 5.85-
6.09 (m, 3 H), 6.83-6.88 (m, 2 H), 7.25-7.32 (m, 2 H); MS
(EI) m/z (rel inten) 500.2 (M+, 0.2), 457.2 (0.2), 444.2 (0.4),
417.2 (0.2), 121 (100).
(+)-4,5,6-Tr i-O-a llyl-1-O-(p-m eth oxyben zyl)-m yo-in osi-
tol (28). Removal of the cis-cyclohexylidene group of (-)-27
(172 mg, 0.34 mmol) with acetyl chloride, as described for (-)-
10, afforded (+)-22 (syrup, 63 mg, 87%): [R]23 ) +3.6° (c 2,
D
1
CHCl3); H NMR (CDCl3) δ 3.67-3.70 (m, 1 H), 4.01-4.07 (t,
J ) 9.9 Hz, 1 H), 4.25-4.33 (m, 2 H), 4.52 (q, J ) 9.6, 18.9
Hz, 1 H), 4.59-5.08 (m, 17 H), 6.94-7.33 (m, 40 H); 31P NMR
(CDCl3, 85% H3PO4 as external standard) δ 1.16, -1.05, -1.63;
HRMS (MALDI) m/z (rel inten) 1141.337 (M + H, 100).
(+)-1-O-(1,2-Di-O-p a lm it oyl-sn -glycer ol-3-b en zyloxy-
ph osph or yl)-2,4-di-O-ben zyl-myo-in ositol 3,5,6-Tr is(diben -
zyl p h osp h a te) (23). Coupling of (+)-22 (55 mg, 0.005 mmol)
with the 1,2-di-O-palmitoyl-sn-glycerophosphoryl moiety, as
5, gave (+)-28 (amorphous, 130 mg, 90%): [R]23 ) +10.3° (c
D
3.5, CHCl3); 1H NMR (CDCl3) δ 2.43 (s, 2 H), 3.19 (t, J ) 14.2
Hz, 1 H), 3.28-3.40 (m, 2 H), 3.55-3.73 (m, 2 H), 3.81 (s, 3
H), 4.13-4.39 (m, 7 H), 4.63 (q, J ) 5.5, 19 Hz, 2 H), 5.15-
5.32 (m, 6 H), 5.89-6.03 (m, 3 H), 6.90 (d, J ) 12.7 Hz, 2 H),
7,26 (d, J ) 12.2 Hz, 2 H); MS (EI) m/z (rel inten) 420.1 (M+,
0.1), 379.1 (0.5), 299.1 (0.2), 121 (100).
(+)-4,5,6-Tr i-O-a llyl-2,3-d i-O-b en zyl-1-O-(p -m et h oxy-
ben zyl)-m yo-in ositol (29). Benzylation of (+)-28 (100 mg,
0.24 mmol) with NaH and benzyl bromide, as described for
(-)-4, gave (+)-29 (syrup, 140 mg, 95%): [R]23D ) +5.4° (c 3.7,
CHCl3); 1H NMR (CDCl3) δ 3.16-3.21 (m, 3 H), 3.74-3.69 (m,
5 H), 4.17-4.42 (m, 7 H), 4.43-4.70 (m, 4 H), 4.78 (s, 2 H),
5.08-5.28 (m, 6 H), 5,75-6.09 (m, 3 H), 6.81-6.85 (m, 2 H),
7.19-7.42 (m, 12 H); MS (EI) m/z (rel inten) 600.1 (M+, 0.),
559.1 (0.1), 509.1 (1), 479.1 (0.1), 91 (100).
(-)-2,3-Di-O-ben zyl-1-O-(p-m eth oxyben zyl)-m yo-in osi-
tol (30). Deallylation of (+)-29 (134 mg, 0.22 mmol) with
RhCl(PPh3)3, as described for (-)-8, afforded (-)-30 (amor-
phous, 87,1 mg, 81%): [R]23D ) -21.8° (c 2.2, CHCl3); 1H NMR
(CDCl3) δ 2.57 (br s, 3 H), 3.13-3.24 (m, 2 H), 3.32-3.47 (m,
2 H), 3.78 (s, 3 H), 3.93-4.03 (m, 2 H), 4.39-4.75 (m, 6 H),
6.88 (d, J ) 12.7, 2 H), 7.19-7.45 (m, 12 H); MS (EI) m/z (rel
inten) 479.9 (M+, 0.1), 388.9 (3), 268.9 (5), 91 (100).
described for (+)-12, provided (+)-23 (syrup, 75 mg, 84%):
1
[R]23 ) +0.9° (c 2.6, CHCl3); H NMR (CDCl3) δ 0.88 (t, J )
D
6.9 Hz, 6 H), 1.25 (s, 48 H), 1.43-1.62 (m, 4 H), 2.26-2.40 (m,
4 H), 3.97-4.26 (m, 6 H), 4.28-4.58 (m, 3 H), 4.60-5.07 (m,
20 H), 6.94-7.40 (m, 45 H); 31P NMR (CDCl3, 85% H3PO4 as
external standard) δ -0.20, -0.50, -0.73 (d), -1.05 (d); HRMS
(MALDI) m/z (rel inten) 1862.830 (M + 2 H, 100).
L-r-P h osp h a tid yl-D-m yo-in ositol 3,5,6-Tr isp h osp h a te,
Dip a lm itoyl [P td In s(3,5,6)P 3]. The perbenzylated deriva-
tive (+)-23 (65 mg, 0.003 mmol) was subjected to hydrogenoly-
sis, as described for PtdIns(3,4,6)P3, to provide PtdIns(3,5,6)P3
(lyophilized powder, 36 mg, 98%): [R]23D ) -10.9° (c 5, CHCl3);
1H NMR (CDCl3) δ 0.89 (t, J ) 7.2 Hz, 6 H), 1.28 (s, 48 H),
1.52-1.66 (m, 4 H), 2.24-2.35 (m, 4 H), 3.96-4.28 (m, 7 H),
4.40-4.45 (m, 2 H), 4.63-4.75 (m, 1 H), 5.21-5.30 (m, 1 H);
31P NMR (CDCl3, 85% H3PO4 as external standard) δ -0.92
(2P), -1.31 (1P), -2.74 (1P); HRMS (MALDI) m/z (rel inten)
1049.399 (M - H, 100).
(+)-2,3-Di-O-ben zyl-1-O-(p-m eth oxyben zyl)-m yo-in osi-
tol 4,5,6-Tr is(d iben zyl p h osp h a te) (31). Phosphorylation
of (-)-30 (82.1 mg, 0.17 mmol) via the phosphoramidite method
as described for (-)-9 yielded (+)-31 (syrup, 215 mg, 99%):
(+)-6-O-Allyl-2,3:4,5-d i-O-cycloh exylid en e-m yo-in osi-
tol (24). Regioselective allylation of (+)-1 (200 mg, 0.59
mmol), as described for (-)-3, afforded (+)-24 (196.3 mg,
[R]23 ) +2.0° (c 3.5, CHCl3); 1H NMR (CDCl3) δ 3.38-3.45
D
87%): [R]23 ) +4.3° (c 2.8, CHCl3); 1H NMR (CDCl3) δ 1.31-
(m, 3 H), 3.78 (s, 3 H), 4.34-4.71 (m, 5 H), 4.86-5.07 (m, 16
H), 6.80 (d, J ) 8.7 Hz, 2 H), 7.11-7.35 (m, 42 H); 31P NMR
(CDCl3, 85% H3PO4 as external standard) δ 1.49, -0.86, -1.54;
HRMS (MALDI) m/z (rel inten) 1261.401 (M + H, 100).
(+)-2,3-Di-O-b en zyl-m yo-in osit ol 4,5,6-Tr is(d ib en zyl
p h osp h a te) (32). Removal of the p-methoxybenzyl function
of (+)-31 (204 mg, 0.16 mmol) with TFA, as described for (-)-
D
1.77 (m, 20 H), 2.58 (d, J ) 2.4 Hz, 1 H), 3.48 (dd, J ) 12, 15.9
Hz, 1 H), 3.80 (dd, J ) 2.7, 12 Hz, 1 H), 3.90-4.02 (m, 1 H),
4.12-4.23 (m, 3 H), 4.30-4.45 (m, 2 H), 5.16-5.36 (m, 2 H),
5.69-6.09 (m, 1 H); MS (EI) m/z (rel inten) 380.1 (M+, 30),
337.1 (60), 55 (100).
(-)-6-O-Allyl-1-O-(p-m eth oxyben zyl)-2,3:4,5-d i-O-cyclo-
h exylid en e-m yo-in ositol (25). A solution of (+)-24 (180 mg,
0.47 mmol) in DMF (4 mL) was treated with NaH (21.2 mg,
85% purity, 0.71 mmol) at 0 °C under argon for 30 min,
followed by p-methoxybenzyl chloride (71.8 µL, 0.71 mmol).
10, gave (+)-32 (148 mg, 80%): [R]23 ) +3.2° (c 2, CHCl3);
D
1H NMR (CDCl3) δ 2.05 (br s, 1 H), 3.44-3.52 (m, 1 H), 3.67-
3.74 (m, 1 H), 4.02 (br s, 1 H), 4.14-4.28 (m, 1 H), 4.40-5.08
(m, 18 H), 7.06-7.41 (m, 40 H); 31P NMR (CDCl3, 85% H3PO4

Phosphatidylinositol 3,4,5-Trisphosphate-Binding Site
J . Org. Chem., Vol. 63, No. 16, 1998 5437
as external standard) δ 1.39, -0.80, -1.34; HRMS (MALDI)
m/z (rel inten) 1141.335 (M + H, 100).
(+)-1-O-(1,2-Di-O-p a lm it oyl-sn -glycer ol-3-b en zyloxy-
ph osph or yl)-2,3-di-O-ben zyl-myo-in ositol 4,5,6-Tr is(diben -
zyl p h osp h a te) (33). Coupling of (+)-32 (40 mg, 0.035 mmol)
with the 1,2-di-O-palmitoyl-sn-glycerophosphoryl moiety, as
100 mM KCl, and 1 mM DTT. Varying amounts of the
phospholipid, in 15 µL of the same buffer, were gradually
introduced into 785 µL of the buffer containing the CT-SH2
domain (12.5 µM).
Changes in the fluorescence intensity were used as a
measure of the binding affinity of the protein-lipid complex
according to the following equation17
described for (+)-12, yielded (+)-33 (54.5 mg, 84%): [R]23
)
D
+6.4° (c 1.2, CHCl3); 1H NMR (CDCl3) δ 0.88 (t, J ) 6.0 Hz, 6
H), 1.14-1.36 (m, 48 H), 1.46-1.62 (m, 4 H), 2.16-2.28 (m, 4
H), 3.50-3.59 (m, 1 H), 3.91-4.24 (m, 4 H), 4.42-4.74 (m, 5
H), 4.77-5.08 (m, 19 H), 7.06-7.38 (m, 45 H); 31P NMR (CDCl3,
85% H3PO4 as external standard) δ -0.63 (d), -0.73 (d), -1.48
(d), -1.51 (d); HRMS (MALDI) m/z (rel inten) 1862.847 (M +
2 H, 100).
1/[[1 - (∆F/∆Fmax)]Ka] ) [[phosphoinositide]total/
(∆F/∆Fmax)] - [CT-SH2]total
where Ka denotes the association constant.
P r ep a r a tion of Wa sh ed Ra bbit P la telets. Platelets were
obtained from adult New Zealand White rabbits and washed
using a modified method of Baenziger and Majerus.18 In brief,
whole blood was anticoagulated by treating 9 volumes of blood
with 1 volume of anticoagulant that consisted of 3.8% (w/v)
trisodium citrate and 140 mM dextrose. Platelet-rich plasma
was obtained by centrifugation of whole blood at 200g for 15
min. The platelets were pelleted by centrifugation at 1500g
for 10 min. The platelet pellets were resuspended in a platelet-
washing buffer, consisting of 4.3 mM Na2HPO4, 24.3 mM
NaH2PO4, 4.3 mM K2HPO4, pH 6.5, 113 mM NaCl, 5.5 mM
glucose, and 0.5% bovine serum albumin, washed twice, and
resuspended in buffer A consisting of 20 mM Tris/HCl, pH 7.0,
containing 150 mM NaCl and 5 mM glucose. Platelet concen-
trations were determined by using a Coulter cell counter
(Coulter Electronics, Inc., Hialeah, FL) and were adjusted to
2-3 × 108 cells/mL for all experiments. These washed
platelets stood at room temperature for 1-2 h before use.
P la telet Aggr ega tion . Washed platelets were added to
an aggregation cuvette, and the reaction was initiated by
adding indicated amounts of the phosphoinositide in the
presence of 1 mM Ca2+ with stirring at 1000 rpm. Platelet
aggregation was monitored on a two-channel Chrono-Log
aggregometer (Chrono-Log Corp., Havertown, PA) at 37 °C.
L-r-P h osp h a tid yl-D-m yo-in ositol 4,5,6-Tr isp h osp h a te,
Dip a lm itoyl [P td In s(4,5,6)P 3]. The perbenzylated deriva-
tive (+)-33 (34 mg, 0.018 mmol) was subjected to hydrogenoly-
sis, as described for PtdIns(3,4,6)P3, to provide PtdIns(4,5,6)P3
(lyophilized powder, 18.8 mg, 98%): [R]23D ) +4° (c 2, CHCl3);
1H NMR (CDCl3) δ 0.89 (t, J ) 7.2 Hz, 6 H), 1.28 (s, 48 H),
1.52-1.66 (m, 4 H), 2.24-2.35 (m, 4 H), 3.53-3.56 (m, 1 H),
3.64-3.68 (m, 2 H), 4.03-4.43 (m, 5 H), 4.46-4.60 (m, 1 H),
4.62-4.78 (m, 1 H), 5.22-5.30 (m, 1 H); 31P NMR (CDCl3, 85%
H3PO4 as external standard) δ -0.22 (2P), -0.95 (1P), -1.51
(1 P); HRMS (MALDI) m/z (rel inten) 1049.382 (M - H, 100).
P r ep a r a tion of th e C-Ter m in a l SH2 Dom a in s of th e
p 85 Su bu n it of P I 3-Kin a se. A pGEX vector containing the
cDNA sequence encoding the p85 CT-SH2 domain was ex-
pressed in E. coli as a GST fusion protein by IPTG induction.6
The GST fusion protein and free SH2 domain were prepared
as follows. The bacterial lysates were incubated with glu-
tathione-coupled Sepharose 4B beads (Sigma) for 2 h at 37 °C
and washed three times with 30 mM Hepes, pH 7.0, containing
100 mM NaCl, 1 mM EDTA, and 0.5% NP-40, followed by the
same buffer without NP-40 twice. Subsequently, the im-
mobilized GST fusion protein was subjected to Factor Xa
hydrolysis to isolate the free SH2 domain. After incubation
with the protease at 4 °C overnight, the mixture was subjected
to centrifugation at 200g for 5 min to separate the free SH2
domain from the GST bead. The SH2 domain was concen-
trated by ultrafiltration. The homogeneity of the isolated SH2
domain was indicated by a single band on 15% SDS-PAGE
with molecular masses of 20 kDa for the CT-SH2 domain.
Ack n ow led gm en t . This investigation was sup-
ported in part by National Institutes of Health Grant
No. R01 GM 53448. The authors thank Professor Lewis
C. Cantley for providing E. coli cells expressing the GST-
SH2 fusion protein.
Su p p or tin g In for m a tion Ava ila ble: NMR spectra for
obtained compounds (33 pages). This material is contained in
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
F lu or escen ce Sp ectr oscop y. Fluorescence spectra were
recorded at 30 °C with a fluorescence spectrophotometer
according to a method previously described.16 Interactions
between the SH2 domain and individual PtdInsP3 isomers
were assessed by monitoring the tryptophan fluorescence with
excitation wavelength at 292 nm. The buffer used for the
fluorescence experiments consisted of 25 mM Tris/HCl, pH 7.5,
J O980356H
(17) Ward, L. D. Methods Enzymol. 1985, 117, 400.
(18) Baenziger, N. L.; Majerus, P. W. Methods Enzymol. 1974, 31,
149.
(16) Lu. P.-J .; Shieh, W.-R.; Rhee, S.-G.; Yin, H. L.; Chen, C.-S.
Biochemistry 1996, 35, 14027.