Synthesis of inositol phosphodiesters by phospholipase C-catalyzed transesterification

Article abstract of DOI:10.1021/ja9616084

Transesterification of primary alcohols with inositol 1,2-cyclic phosphate (IcP) in the presence of phosphatidylinositol-specific phospholipase C (PI-PLC) resulted in the formation of O-alkyl inositol 1-phosphates. The starting IcP was obtained in a single step by PI-PLC catalyzed cleavage of phosphatidylinositol from the soybean phospholipid. The transesterification reaction was performed with a series of 20 structurally diverse hydroxyl compounds, ranging in the structural complexity from methanol to the serine-containing Ser-Tyr-Ser-Met tetrapeptide, to give the corresponding phosphodiesters with 20-80% yields, depending mainly on the solubility of alcohols in water. The rates of transesterifications, and of the competing hydrolysis of IcP to inositol 1-phosphate (IP), were relatively insensitive to the alcohol structure. With polyhydroxyl compounds such as glycerol and hexitols, the enzyme displayed complete preference toward formation of the inositol phosphate derivatives of the primary hydroxyl groups. On the other hand, PI-PLC did not discriminate between primary hydroxyl groups in different environments and showed low stereoselectivity with racemic alcohols featuring a chiral center at the β-position. The O-alkyl inositol phosphates formed were readily separable from the hydrolytic product, IP, by the anion-exchange chromatography, and were fully characterized by means of ¹H and ³¹P NMR and electrospray MS. Our results provide a new, simple, and efficient two-step synthetic route to substituted O-alkyl inositol phosphates from inexpensive starting materials. The reported reaction was successfully applied to synthesis of complex inositol phosphate derivatives, as illustrated by inositol phosphoesters of mono- and oligosaccharides, nucleosides and peptides. The synthetic usefulness of this reaction, however, is not limited to the examples shown. Because transesterification activity of phospholipase C has not been reported before, its mechanism is discussed in a broad context of mechanisms of phosphoesterases.

Full text of DOI:10.1021/ja9616084

J. Am. Chem. Soc. 1996, 118, 7679-7688
7679
Synthesis of Inositol Phosphodiesters by Phospholipase
C-Catalyzed Transesterification†
Karol S. Bruzik,*,‡ Zhiwen Guan,‡ Suzette Riddle,§ and Ming-Daw Tsai§
Contribution from the Department of Medicinal Chemistry and Pharmacognosy, UniVersity
of Illinois at Chicago, Chicago, Illinois 60612-7231, and Departments of Chemistry and
Biochemistry, The Ohio State UniVersity, Columbus, Ohio 43210
ReceiVed May 14, 1996X
Abstract: Transesterification of primary alcohols with inositol 1,2-cyclic phosphate (IcP) in the presence of
phosphatidylinositol-specific phospholipase C (PI-PLC) resulted in the formation of O-alkyl inositol 1-phosphates.
The starting IcP was obtained in a single step by PI-PLC catalyzed cleavage of phosphatidylinositol from the soybean
phospholipid. The transesterification reaction was performed with a series of 20 structurally diverse hydroxyl
compounds, ranging in the structural complexity from methanol to the serine-containing Ser-Tyr-Ser-Met tetrapeptide,
to give the corresponding phosphodiesters with 20-80% yields, depending mainly on the solubility of alcohols in
water. The rates of transesterifications, and of the competing hydrolysis of IcP to inositol 1-phosphate (IP), were
relatively insensitive to the alcohol structure. With polyhydroxyl compounds such as glycerol and hexitols, the
enzyme displayed complete preference toward formation of the inositol phosphate derivatives of the primary hydroxyl
groups. On the other hand, PI-PLC did not discriminate between primary hydroxyl groups in different environments
and showed low stereoselectivity with racemic alcohols featuring a chiral center at the â-position. The O-alkyl
inositol phosphates formed were readily separable from the hydrolytic product, IP, by the anion-exchange
1
chromatography, and were fully characterized by means of H and 31P NMR and electrospray MS. Our results
provide a new, simple, and efficient two-step synthetic route to substituted O-alkyl inositol phosphates from inexpensive
starting materials. The reported reaction was successfully applied to synthesis of complex inositol phosphate
derivatives, as illustrated by inositol phosphoesters of mono- and oligosaccharides, nucleosides and peptides. The
synthetic usefulness of this reaction, however, is not limited to the examples shown. Because transesterification
activity of phospholipase C has not been reported before, its mechanism is discussed in a broad context of mechanisms
of phosphoesterases.
Introduction
methods still take more than 10 steps from inositol or any other
precursor. The biggest effort in these syntheses is expended
on elaborate protection-deprotection schemes, necessary to
achieve the desired regio- and stereoselective phosphorylation
pattern of inositol framework. Creating the molecular diversity
needed for studies of structure-activity relationship of inositol-
related enzymes and receptors,5,7 while feasible, is extremely
labor intensive. In the past we have embarked on devising
chemical syntheses of phosphoinositides applicable toward a
goal of a broader application.8-10 In this paper we report on
the first enzymatic synthesis of O-alkyl inositol phosphates
starting from the corresponding alcohols, and the inexpensive
and readily available soybean phospholipid, using phosphati-
dylinositol-specific phospholipase C (PI-PLC†).
The action of PI-PLC on phosphatidylinositol (PI, 1), or its
phosphorylated derivatives, produces the mixture of the corre-
sponding 1,2-cyclic phosphate (IcP, 2) and inositol 1-phosphate
(IP, 3) (Scheme 1) or their 4- and 5-phosphorylated con-
geners.8,11-13 The bacterial species of PI-PLC cleave the non-
phosphorylated phosphatidylinositols (PI and glycosyl-PI) to
afford initially the cyclic phosphate11,14 as a sole product. This
enzyme also catalyzes the subsequent slow hydrolysis of IcP
into inositol 1-phosphate.11,14 In our earlier report8 we have
Development of new synthetic methods leading to inositol
phosphates is of considerable interest in view of their role as
second messengers in the transduction of cellular signals.1,2 The
current state of art allows synthesis of all known naturally
occurring inositol phosphates and phospholipids, and many of
their analogs with almost any desired pattern of the phosphodi-
ester and phosphomonoester substitution of inositol.3-10 Despite
significant advances, however, most of the described synthetic
† Abbreviations: ESMS, electrospray mass spectrometry; glycero-3-
phospho-(1-myo-inositol), Gro-IP; HPAEC, high performance anion ex-
change chromatography; IcP, inositol 1,2-cyclic phosphate; IP, inositol
1-phosphate; MOPS, morpholinepropanesulfonic acid; PAD, pulse ampero-
metric detection; PI, phosphatidylinositol; PI-PLC, phosphatidylinositol-
specific phospholipase C; SDC, sodium deoxycholate; Tris, tris(hydroxy-
methylene)aminomethane.
‡ University of Illinois at Chicago.
§ The Ohio State University.
X Abstract published in AdVance ACS Abstracts, August 1, 1996.
(1) Berridge, M. J.; Irvine, R. F. Nature 1989, 341, 197-205.
(2) Berridge, M. J. Nature 1993, 361, 315-325.
(3) Beaucage, S.; Iyer, R. P. Tetrahedron 1993, 49, 10441-10488.
(4) Potter, B. L. V. Nat. Prod. Rep. 1990, 1-24.
(5) (a) Potter, B. V. L.; Lampe, D. Angew. Chem., Int. Ed. Engl. 1995,
34, 1933-1972. (b) Safrany, S. T.; Mills, S. J.; Liu, C.; Lampe. D.; Noble,
N. J.; Nahorski, S. R.; Potter, B. V. L. Biochemistry 1994, 33, 10763-
10769. Lu, P.-J.; Gou, D.-M.; Shieh, W.-R.; Chen, C.-S. Biochemistry 1994,
33, 11586-11597.
(10) (a) Bruzik, K. S.; Kubiak, R. J. Tetrahedron Lett. 1995, 36, 2415-
2418. (b) Bruzik, K. S.; Myers, J.; Tsai, M.-D. Tetrahedron Lett. 1992, 33,
1009-1012.
(6) Billington, D. C. Chem. Soc. ReV. 1989, 18, 83-122.
(7) Billington, D. C. The Inositol Phosphates: Chemical Synthesis and
Biological Significance; Weinheim: New York, 1993.
(8) Bruzik, K. S.; Morocho, A. M.; Jhon, D. Y.; Rhee, S. G.; Tsai, M.-
D. Biochemistry 1992, 31, 5183-93.
(11) Volwerk, J. J.; Shashidhar, M. S.; Kuppe, A.; Griffith, O. H.
Biochemistry 1990, 29, 8056-8062.
(12) Kim, J. W.; Ryu, S. H.; Rhee, S. G. Biochem. Biophys. Res.
Commun. 1989, 163, 177-182.
(9) Bruzik, K. S.; Tsai, M.-D. J. Am. Chem. Soc. 1992, 114, 6361-
6374.
(13) Wilson, D. B.; Bross, T. E.; Sherman, W. R.; Berger, R. A.; Majerus,
P. W. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 4013-4017.
S0002-7863(96)01608-3 CCC: $12.00 © 1996 American Chemical Society

7680 J. Am. Chem. Soc., Vol. 118, No. 33, 1996
Bruzik et al.
Scheme 1
provided evidence that in spite of the apparently different
behavior of mammalian and bacterial enzymes, reflected by
simultaneous vs sequential formation of the cyclic and acyclic
products, respectively, both enzymes utilize analogous trans-
esterification mechanisms, involving attack of the inositol
2-hydroxyl group on the phosphorus atom as a key step of the
phospholipid cleavage.
Reverse reactions of esterases, proteases, and glucosidases
are widely used in the synthesis of esters, peptides, and
oligosaccharides, respectively. In contrast, examples of syn-
thetic applications of phosphoesterases are limited to the
transphosphatidylation activity of phospholipase D15-18 and
transesterification activity of pancreatic ribonuclease,19,20 snake
venom phosphodiesterase,21,22 and alkaline phosphatase.23 This
report is intended to fill this gap by exploring the synthetic
potential of PI-PLC toward synthesis of the diverse series of
inositol phosphodiesters.
Figure 1. 31P NMR spectra of reaction mixtures of IcP and PI-PLC
under the following conditions: (A) IcP (10 mM) in Tris-HCl (70 mM)
and EDTA (10 mM) buffer; (B) sample of DPPI (10 mM) dispersed
with sodium deoxycholate (0.1 M) in Tris-HCl (0.2 M) 17 h after
addition of PI-PLC from B. cereus (3.5 µg); (C) sample of IcP (10
mM) in 70 mM MOPS-Na, 10 mM EDTA treated with dialyzed PI-
PLC; (D) same as trace C, but in the presence of 0.5 M glycerol; (E)
same as the trace C, but in the presence of 0.5 M Tris-HCl; and (F)
same as C, but after 16 days. All experiments were performed at pH
7.0.
Results
Discovery of the Reverse Transesterification Catalyzed by
PI-PLC. Hydrolysis of 1,2-dipalmitoyl-sn-glycero-3-phospho-
(1-myo-inositol) (DPPI, 1a)8 dispersed as mixed micelles with
sodium deoxycholate (SDC) in 70 mM Tris-HCl buffer (pH
7.0) catalyzed by PI-PLC from both Bacillus thuringiensis or
Bacillus cereus provided IcP (2, δ31P 16.1 ppm, Figure 1A).
The monitoring of the further hydrolytic cleavage of IcP
catalyzed by PI-PLC from either source revealed formation of
inositol 1-phosphate (IP, 3, δ 3.4 ppm) as expected, and two
unidentified products giving rise to signals at -0.3 and -0.5
ppm (Figure 1B). Since both enzyme preparations contained
significant amounts of glycerol, the control reaction was
performed using the dialyzed enzyme from B. thuringiensis in
70 mM MOPS-Na buffer at pH 7.0, which produced IP as an
exclusive product (Figure 1C). In another experiment, the
treatment of IcP with PI-PLC from B. thuringiensis in MOPS
buffer, and in the presence of 0.5 M glycerol, afforded IP and
a product giving rise to the signal at -0.3 ppm (Figure 1D).
These products were resolved by the anion-exchange chroma-
tography on Sephadex QAE, and their identity was determined
by NMR, MS, and high performance anion-exchange chroma-
tography (HPAEC). The comparison of spectral and chromato-
graphic data of this compound with those of the product of
deacylation of phosphatidylinositol, unequivocally established
its structure as glycero-3-phospho-(1-myo-inositol) (Gro-IP, 4).
The treatment of IcP with the dialyzed B. thuringiensis PI-PLC
in the presence of 0.5 M Tris-HCl buffer at pH 7.0 resulted in
the formation of a product giving rise to the signal at -0.5 ppm
(Figure 1E), in addition to IP. Purification of this product and
(14) (a) Griffith, O. H.; Volwerk, J. J.; Kuppe, A. Methods Enzymol.
1991, 197, 493-502. (b) Leigh, J. A.; Volwerk, J. J.; Griffith, O. H.; Keana,
J. F. Biochemistry 1992, 31, 8978-8993.
1
its analysis by H NMR and MS determined its structure as
1-O-(2-amino-2-hydroxymethylene-3-hydroxypropyl) 1-myo-
inositol phosphate (5). Hence, both products 4 and 5 are formed
as a result of the phospholipase C catalyzed transesterification
of glycerol or Tris, respectively, through the ring opening of
the five-membered cyclic phosphate. In the blank experiments,
incubation of IcP with each of the alcohols in the absence of
enzyme over several days produced no reaction.
(15) Yang, S. F.; Freer, S.; Benson, A. A. J. Biol. Chem. 1967, 242,
477-484.
(16) Holbrook, P. G.; Wurtman, R. J. Biochim. Biophys. Acta 1990, 1046,
185-188.
(17) Siddiqui, R. A.; Exton, J. H. J. Biol. Chem. 1992, 267, 5755-5761.
(18) Wang, P.; Schuster, M.; Wang, Y.-F.; Wong, C.-H. J. Am. Chem.
Soc. 1993, 115, 10478-10491.
(19) Saenger, W.; Suck, D.; Eckstein, F. Eur. J. Biochem. 1974, 46, 559-
Driving Force for Transesterification. The hydrolysis of
the cyclic five-membered phosphodiesters is energetically more
567.
(20) Usher, D. A.; Erenrich, E. S.; Eckstein, F. Proc. Natl. Acad. Sci.
U.S.A. 1972, 69, 115-118.
(21) Garcia-Diaz, M.; Avalos, M.; Cameselle, J. C. Eur. J. Biochem.
1991, 196, 451-457.
(24) 31P NMR chemical shifts are somewhat sensitive to the pH and
buffer conditions. The chemical shifts of signals in the spectra shown in
Figure 1 recorded for buffered samples differ, therefore, from those reported
in the Experimental Section, obtained for samples of pure phosphodiesters
dissolved in D2O only by 0.5-1 ppm.
(22) Garcia-Diaz M.; Avalos M.; Cameselle J. C. Eur. J. Biochem. 1993,
213, 1139-1148.
(23) Han, R.; Coleman, J. E. Biochemistry 1995, 34, 4238-4245.

Synthesis of Phosphodiesters with Phospholipase C
J. Am. Chem. Soc., Vol. 118, No. 33, 1996 7681
Table 1. 1H-1H and 31P-1H Coupling Constants for IcP, IP, and
Inositola
compd H1-H2 H2-H3 H3-H4 H4-H5 H5-H6 H6-H1 P-H1 P-H2
IcP
IP
inositol
4.8
2.7
2.7
3.8
2.7
2.7
10.1
9.7
9.8
10.1
8.9
8.8
9.4
8.9
8.8
8.3
9.6
9.8
20.1 ca. 0
8.6
a All coupling constants are for aqueous solutions at neutral pH; 0.2
Hz digital resolution.
Table 2. Maximum Yields of O-Alkyl Inositol Phosphates and
Their Physicochemical Data
Figure 2. MM2 optimized conformations of IP and IcP. The numerical
values designate C3-C2-C1-C6 and C2-C1-C6-C5 torsional
angles.
retention
maximum δ31P [ppm] ES MS timea
alcohol [M]
yield [%] (multiplicity) [m/z]
[min]
in anhydrous DMSO in the presence of boron trifluoride resulted
in the formation of the mixture of the corresponding O-butyl
inositol 1- and 2-phosphates. The low yield (15%) of the acyclic
diesters and the low regioselectivity of this reaction indicated,
however, that the utility of IcP as a nonenzymatic phosphory-
lating agent is doubtful (see also ref 26).
methanol [6.0]
70
63
15
33
74
46
55
81
30
24
40
50
nd
18
nd
nd
25
15
10
1.1 (qtr)
0.29
0.86
0.63
0.14 (q)
0.07 (q)
273.6 3.7
n-propanolc
301 nd
315.5 4.0
299.4 4.0
303.4 4.5
317.7 3.6
359.8 3.5
333.8 5.8
423.0 12.11
423.0 10.1
317.0 2.5
nd
butanol [0.6]
allyl [1.4]
ethane-1,2-diol [4.5]
(DL)-propane-1,2-diol [6.0]
hexane-1,6-diol [2.0]
glycerol [4.0]
D-mannitol [1.5]
L-iditol [0.15]
3-aminopropanol [6.0]
Tris buffer [0.5]
choline [0.5]
(DL)-serine [0.67]
pentaethylene glycold
biotinylpentaethylene glycole
glucose [3.0]
Enzymatic Synthesis of IcP. For transesterification reaction
of alcohols with IcP to be of any synthetic utility, the latter
starting material has to be readily available. The earlier reported
syntheses of the racemic IcP,27 and its phosphorothioate
analog,28 from inositol have been ruled out as a source of this
material, due to a greater difficulty in synthesis of the enan-
tiomerically pure IcP. Alternatively, we have considered
naturally occurring phosphoinositides as a starting material. The
crude phosphoinositide, containing PI as a major component,
can be readily obtained in the multigram quantity from the
inexpensive soybean phospholipid by the precipitation with
methanol from its chloroform solution.29 The treatment of the
sodium deoxycholate (SDC) dispersion of this phosphoinositide
fraction with phospholipase C followed by the detergent removal
via chloroform-methanol extraction at the pH 5.0, and further
anion-exchange chromatography afforded IcP contaminated with
only small amounts of the glycosylated IcP. In a typical
preparation, treatment of 20 g of phosphoinositide fraction with
only 20 µg of PLC afforded 1.7 g of IcP. This procedure can
be further simplified by omitting the use of the detergent. The
treatment of the heterogeneous ether-water dispersion of the
soybean phospholipid with PI-PLC resulted in the exhaustive
cleavage of phosphatidylinositol from the mixture. The removal
of the unhydrolyzed phospholipid by extraction with chloroform-
methanol, and the anion-exchange chromatography of the water-
soluble product using ammonium carbonate step-gradient af-
forded pure IcP.
Time Course of Transesterification of Alcohols with IcP.
A representative time course of the enzymatic transesterification
of an alcohol with IcP catalyzed by PI-PLC, exemplified by
the reaction of IcP with 1,6-hexanediol, is shown in Figure 3.
Formation of the O-1-(6-hydroxyhexyl) 1-myo-inositol phos-
phate (6) was followed by HPAEC on Carbopack PA1 column
using pulse-amperometric detector (PAD) for monitoring peak
elution. The disappearance of the IcP signal (tR 10.5 min) was
accompanied by the formation of the phosphodiester 6 at 3.5
min and IP (tR 13.5 min). At longer reaction times, the
concentration of IP increased mainly at a cost of the acyclic
diester product, while the concentration of IcP declined only
very slowly. At the very long reaction times (several days) IP
0.14
0.44 (dtr)
0.63
0.68
-1.56
-1.06 (q)
329.9 nd
-0.74, -0.81 346 19.0
0.49
0.44
0.06
-0.47
nd
479
706
1.9
b
421 15.0
485 2.7
727 15.8, 16.2
uridine [0.4]
Ser-Tyr-Ser-Met [0.04]
a Carbopack PA 1, 4 × 250 mm, isocratic 0.1 NaOH in NaOAc
gradient 40-200 mM in 30 min. b Product not visible by HPAEC.
c Propanol concentration 50% v/v. d Concentration 22% w/v. e Con-
centration 26% w/v; nd, not determined.
favorable in comparison with hydrolysis of the acyclic phos-
phodiesters by 4.6 kcal/mol.25a The origin of this free enthalpy
difference, once thought to arise from dioxaphospholane ring
strain,25a has now been ascribed to the differential solvation
energies of the substrate and the product.25b In the case of a
five-membered cyclic phosphate fused to a six-membered ring,
as in the case of IcP, the energy difference should be enhanced
due to an additional torsional strain of the cyclohexyl ring
produced by the bridging of the axial and equatorial hydroxyl
groups of inositol by the phosphate function. This strain would
be relieved upon ring opening giving rise to a greater driving
force for the ring opening. The examination of the vicinal 1H-
1H coupling constants in IcP and IP and inositol (Table 1)
revealed that there are significant differences between these
parameters in IcP and IP. The increase in H1-H2, H2-H3
couplings and the decrease in H1-H6 coupling constants in
IcP indicate flattening of the cyclohexyl ring in its C3-C2-
C1-C6 fragment. For example, the C3-C2-C1-C6 torsional
angle is changed from -57° in IP to -47° in IcP (Figure 2),
and the C2-C1-C6-C5 angle is similarly decreased from 57°
to 51°. In contrast, IP displays very similar couplings to those
of the unphosphorylated inositol, indicating that the distortion
of the six-membered ring is due to the presence of the five-
membered ring, and not phosphorylation.
Consistently, the half-life of IcP at pH 1 is less than 1 min,13
and IcP hydrolysis is already detectable at pH 4.0, suggesting
that ring opening of IcP could be a useful phosphorylation
reaction, if an alcohol could be substituted for water. In fact,
the treatment of tetra-n-butylammonium salt of IcP with butanol
(26) Schultz, C.; Metschies, T.; Gerlach, B.; Stadler, C.; Jastorff, B.
Synlett. 1990, 3, 163-5.
(27) Watanabe, Y.; Ogasawara, T.; Nakahira, H.; Matsuki, T.; Ozaki, S.
Tetrahedron Lett. 1988, 29, 5259-62.
(28) Schultz, C.; Metschies, T.; Jastorff, B. Tetrahedron Lett. 1988, 29,
3919-3920.
(25) (a) Haake, P. C.; Westheimer, F. H. J. Am. Chem. Soc. 1961, 83,
1102-1109. (b) Dejaegere A.; Karplus M. J. Am. Chem. Soc. 1993, 115,
5316-5317.
(29) Hawthorne, J. N.; White, D. A. Vitam. Horm. (New York) 1975,
33, 529-573.

7682 J. Am. Chem. Soc., Vol. 118, No. 33, 1996
Bruzik et al.
Figure 3. Time course of transesterification of hexane-1,6-diol with
IcP in the presence of PI-PLC at room temperature followed by
HPAEC-PAD. Reaction conditions: [diol], 2 M; [IcP], 0.4 M; PI-PLC,
15 µg. Chromatographic conditions: Carbopack PA1 column 4.6 ×
250 mm with 4.6 × 50 mm guard column was eluted with isocratic
NaOH (100 mM) and sodium acetate gradient (40-200 mM in 30 min).
Scheme 2
Figure 4. (A) Rates of formation of O-alkyl inositol phosphates from
IcP with various alkanols in the presence of PI-PLC. Reaction
conditions: initial concentration of alkanols 0.5 M, [IcP] 10 mM, PI-
PLC 3.5 µg. All reactions were followed by 31P NMR using ca. 30
min accumulation time for each spectrum. The experimental points
denoted refer to the middle of the accumulation period of the spectrum.
(B) Rate of formation of IP in the presence of 0.5 M alkanols.
yields were obtained in cases of simple alcohols, where their
high molar concentrations were possible to achieve. The yields
in some cases were limited by the poor solubility of the alcohols
in water. The transesterification reaction was found to be of a
general scope, with most alcohols participating at a comparable
efficiency (Figure 4A). For example, the calculated alcohol
efficiency parameter, defined as the ratio of the mole fraction
of the O-alkyl inositol phosphate in the product mixture to the
mole fraction of alcohol in the starting alcohol/water solution,
was as follows: methanol (6), mercaptoethanol (12), propanol
(17), serine (18), propane-1,2-diol (21), ethylene glycol (25),
glycerol (28), mannitol (29), and Tris (57). The rates of the
formation of the acyclic diesters with such alcohols as ethylene
glycol, glycerol, mannitol, and serine were very close indicating
that the reaction is quite insensitive to the nature of substituents
at the â-position such as hydroxyl, amino, tetraalkylammonium,
carboxyl, and sulfhydryl group. The only exception among
alcohols studied was Tris, which produced ca. 2-fold higher
yield of the product. As shown in Figure 4B the presence of
alcohols and their structural variation had little effect on the
initial rates of the competing IcP hydrolysis, indicating that none
of the alcohols studied was bound in the IcP site.
Although the range of alcohol structures accepted by the
enzyme was quite broad, we did note a few examples where
we were unable to obtain products of transesterification.
Reactions with secondary alcohols, isopropyl alcohol, 2-butanol,
and cyclohexanol did not afford the corresponding phosphodi-
esters, nor did the presence of such alcohols affect the rate of
the hydrolysis of IcP (results not shown). Likewise, reactions
with long chain primary alcohols such as dodecanol and
hexadecanol also failed to afford the product. The attempt to
was an exclusive product (not shown, but see Figure 1F for
analogous reaction). Due to the breakdown of the acyclic
phosphodiesters at the longer reaction times, their optimal yields
in reactions of various alcohols were assured by monitoring the
reactions by HPAEC-PAD, reverse phase HPLC, or 31P NMR
and stopping the reaction when the concentration of an acyclic
diester reached a highest point.
Alcohol Specificity of PI-Specific Phospholipase C. In
order to determine structural specificity of the enzyme we have
examined a series of reactions of IcP with primary and secondary
alcohols (Scheme 2). The results are summarized in Table 2
and in Figure 4A,B. The reactions of alcohols with IcP in the
presence of PLC resulted in the formation of the corresponding
acyclic phosphodiesters 4-23 with the yields generally exceed-
ing 20%, but in some cases greater than 70%. The highest

Synthesis of Phosphodiesters with Phospholipase C
J. Am. Chem. Soc., Vol. 118, No. 33, 1996 7683
Figure 5. Dependence of the yield of diester 6 on the concentration
of hexanediol. The reactions were monitored by HPAEC as described
in Figure 2. The yields of 6 are expressed as a molar fraction of all
inositol components after 2 h. The reaction conditions are the same as
in Figure 3, except for alcohol concentration.
increase the alcohol solubility in water by adding the hydrophilic
function, as in ω-hydroxyhexanoic acid, produced a negative
result, as well. On the other hand, the addition of 0.5 M inositol
did not provide the acyclic diester product, but it almost
completely inhibited the hydrolysis of IcP, indicating inositol
binding in the IcP site as previously reported.30 Other unsuc-
cessful attempts included alcohols with added electrophilic
reactivity, such as in bromoethanol and glycidol.
In order to optimize transesterification conditions we have
investigated the dependence of the yield of the phosphodiester
upon the alcohol concentration (Figure 5). The yield of the
diester 6 linearly increased with the concentration of hexane-
1,6-diol and started leveling off only at concentrations above 2
M, indicating weak binding affinity of the alcohol to PI-PLC.
Regio- and Stereospecificity of PLC Catalyzed Transes-
terification. With compounds featuring multiple hydroxyl
groups such as mannitol and glycerol there is a possibility of
two types of product isomerism due to phosphorylation of the
primary vs secondary hydroxyl group in mannitol, and the pro-S
vs the pro-R hydroxyl group in glycerol. In all cases of polyols
examined we have seen only phosphorylation of the primary
hydroxyl group, as evidenced by observation of either a quartet
Figure 6. 31P NMR spectra of Gro-PI (4) (A and B). Spectrum A
shows the product obtained by transesterification of glycerol with IcP.
Spectrum B was acquired after adding the diasteromerically pure Gro-
PI, obtained by deacylation of PI, to the sample shown in trace A.
symmetry. Further experiments with racemic alcohols such as
dl-propane-1,2-diol (12) and dl-serine (17) also provided almost
equimolar mixtures of diastereomers (not shown). In summary,
while PI-PLC displays a strong preference for the primary
hydroxyl groups, it exhibits very low stereoselectivity with
regard to configuration at the C-2 of an alcohol. This conclusion
is consistent with our earlier results showing that PI cleavage
by PI-PLC is nonstereospecific with regard to configuration at
the C-2 of the diacylglycerol moiety.8
1
or a doublet of triplets (not shown) in H-coupled 31P NMR
Application of PI-PLC Catalyzed Transesterification to
Synthesis of Complex Inositol Phosphodiesters. The potential
of the PI-PLC-catalyzed transesterification reaction as applied
to derivatization of structurally complex compounds is illustrated
by the syntheses discussed below: (A) 6-glucosylphospho-
inositol (20), (B) 5′-uridyl phosphoinositol (22), and (C) inositol
1-phosphate derivative of Ser-Tyr-Ser-Met tetrapeptide (23).
(A) Monitoring of the transesterification of glucose (the
mixture of R- and â-anomers) with IcP by HPAEC-PAD
indicated formation of a new product giving rise to the peak at
15.4 min, in addition to those of IcP and IP (Figure 7A).
Purification of this product by the preparative anion-exchange
chromatography and analysis by 1H, 31P, and ESMS confirmed
the structure of the product as 6-glucosyl phosphoinositol. The
product 20 constituted a mixture of both R- and â-anomers as
indicated by the presence of two doublets at 5.20 ppm (3.8 Hz)
spectra. Phosphorylation of the primary hydroxyl group is also
evident from the analysis of the 1H NMR spectra. The reaction
of IcP with glycerol afforded two isomers giving rise to very
closely spaced 31P NMR signals with 45:55 intensity ratio31
(Figure 6A). The addition to this product of a single isomer of
Gro-PI (4), obtained by the deacylation of PI, showed the
1
increase of the down-field signal (Figure 6B). Both H NMR
1
and H-1H COSY spectra were consistent with the presence
of two stereoisomers having either the pro-S or pro-R hydroxyl
group phosphorylated.32 With mannitol and iditol only a single
product was formed with the phosphorylated primary hydroxyl
group. This is due to the fact that both primary OH groups
present in these molecules are homotopic because of the C2
(30) Heinz, D. W.; Ryan, M.; Bullock, T. L.; Griffith, O. H. EMBO J.
1995, 15, 3855-3863.
1
(31) The two 31P NMR signals of Gro-IP obtained by transesterification
of glycerol are observed only when high resolution spectra of the Gro-IP
sample are obtained, typically requiring sample treatment with Chelex resin
or EDTA. This signal doubling was not observable in the spectra shown in
Figure 1.
and at 4.62 ppm (8.0 Hz), respectively, in H NMR spectrum.
The two anomeric products are not resolved by HPAEC and
31P NMR. An analogous reaction of lactose with IcP produced
a complex mixture of two R- and two â-isomers (21), due to
the random phosphorylation of the 6-hydroxyl groups in the
galactose and glucose rings. The obtained mixture of lactose-
phosphoinositols gave rise to a broad 31P NMR signal at 0.5
ppm, and two HPAEC peaks at 13.6 min and 14.0 min. This
1
(32) The H NMR spectrum of this diastereomeric mixture (giving rise
to a 31P NMR spectrum shown in Figure 6A) displays additional splitting
of glycerol H-1′A and H-1′B protons (by ca. 2.7 Hz (A) and 1 Hz (B),
respectively). Other resonances remain identical to those of the pure sn-3-
Gro-1-IP.

7684 J. Am. Chem. Soc., Vol. 118, No. 33, 1996
Bruzik et al.
Figure 7. (A) Formation of glucose inositol phosphate as observed
by HPAEC-PAD on Carbopack PA1. Identities of peaks are tR 2.8 min,
glucose, tR 11.2 min, IcP; tR 15 min, IP; tR 15.5 min, compound 20.
(B) HPAEC of the purified 20. Chromatographic conditions are the
same as in Figure 3.
Figure 9. (A) A representative HPLC of the reaction mixture of Ser-
Tyr-Ser-Met tetrapeptide with IcP and PI-PLC after 48 h showing
phosphorylated products 23a at tR 15.4 and 23b at tR 15.8 min. Reaction
conditions: tetrapeptide 0.04 M, PI-PLC 4 µg. Chromatographic
conditions are as described in Experimental Section. (B) An overlay
of the UV spectra of the tetrapeptide, the product 23a and the product
23b.
identical UV spectra to that of the starting material (Figure 9B),
suggesting that the phosphorylation took place at one of the
serines rather than at the tyrosine, since phosphorylation of the
latter would most likely affect the UV profile of the tetrapeptide.
The two isomers were preparatively separated by RP-HPLC and
analyzed by ES-MS. The presence of a single inositol phosphate
residue in the product was verified by the observation of the
molecular ion at m/z 727 for each isomer. However, the small
amounts of material produced precluded assignment of chro-
matographic peaks to particular positional isomers by using
NMR.
Demonstration of Reversibility of PI-PLC-Catalyzed Re-
action. The observed transesterifcation reaction is a manifesta-
tion of the reversibility of the formation of IcP from PI. This
reversibility is further demonstrated by the conversion of the
acyclic diesters obtained from IcP back into IcP in the presence
of PI-PLC. For example, the reaction of IcP with allyl alcohol
in the presence of PI-PLC afforded O-allyl inositol phosphate
(10). The treatment of the purified diester 10 with PI-PLC has
led to the mixture of IcP and IP, and eventually to IP after long
reaction times. In contrast, we have been unable to detect
formation of the phosphodiester from IP, or an equilibrium
between IcP and IP, using a sensitive HPAEC-PAD detection
method.
Figure 8. (A) A representative RP-HPLC of the reaction mixture of
transesterification of uridine with IcP to form uridine inositol phosphate
22. (B) HPLC of the purified 22. See Experimental Section for
chromatographic conditions. (C) A superposition of the UV spectra of
the eluates at 3.9 and 2.7 min.
experiment demonstrated a rather low regioselectivity of PI-
PLC toward the primary hydroxyl groups of oligosaccharides.
(B) The reaction of IcP with uridine was followed by the
reverse phase HPLC, monitoring the appearance of the product
by the photodiode array UV detector (Figure 8A). The product
giving rise to the peak at 2.5 min had a UV spectrum identical
to that of uridine itself (tR 3.9 min) (Figure 8C), indicating the
presence of the intact uracyl residue. This product was purified
by the anion-exchange chromatography (Figure 8B). The
analysis of the product by 1H, 31P NMR, and ES MS confirmed
its structure as 5′-uridyl phosphoinositol (22).
(C) The Ser-Tyr-Ser-Met tetrapeptide was chosen to test
positional specificity of PI-PLC in the phosphorylation of
peptide hydroxyl groups in different environments, since it
features two nonequivalent serine residues and a phenolic
hydroxyl group. The reaction progress was followed by reverse
phase HPLC. After 48 h formation of two products giving rise
to two peaks of equal intensity at 15.5 (23a) and 15.8 min (23b)
was observed, in addition to that of the starting tetrapeptide at
18.7 min (Figure 9A). Further reaction resulted in the diminu-
tion of one of the peaks. Both products 23a and 23b had
Discussion
The results reported above are significant in two aspects: (i)
they show a vast potential of PI-PLC for the synthesis of O-alkyl
inositol phosphodiesters, limited mainly by the hydrophobicity
of the primary alkanol; (ii) our results indicate far reaching
analogies between various groups of phosphoesterases, which
allow certain generalizations to be made.
Synthetic Utility of IcP-Alcohol Transesterification Cata-
lyzed by IcP. We have demonstrated that the transesterification
reaction catalyzed by PI-PLC can be used to synthesize a wide

Synthesis of Phosphodiesters with Phospholipase C
J. Am. Chem. Soc., Vol. 118, No. 33, 1996 7685
variety of substituted O-alkyl phosphoinositols. The enzymatic
synthesis is advantageous over chemical synthesis in that (i) it
uses a readily available soybean phospholipid as a starting
material, (ii) no protection/deprotection steps of alcohols are
required in order to accomplish synthesis of the phosphodiester
linkage with many polyfunctional alcohols, and (iii) the reaction
can be performed with a wide range of primary alcohols and
polyols with a high degree of selectivity between primary and
secondary hydroxyl groups. It is thus possible with minimal
effort to produce a series of polyfunctional alkyl phosphoinosi-
tols for studies of specificity of inositol-dependent enzymes or
enzyme inhibition. The procedure also eliminates the need for
synthesis of appropriately protected inositol derivatives to be
used in the phosphoester group assembly. It should, further-
more, be applicable for one-step incorporation of fluorescent
and radioisotope labels into inositol phosphodiesters to be used
as molecular probes in biological studies. Many applications
of the reaction may be possible as a need for a particular
structure arises.
more efficient substrate than water. We cannot address the
problem of the competition between the diacyl glycerol and an
alcohol since we have been using the purified IcP, rather than
the PI, as a substrate. The relatively weak binding affinity of
an alcohol substrate seems to be a common property of not only
PI-PLC but also other phosphoesterases as discussed in the next
section. This low binding affinity results probably from the
optimization of the catalysis for the forward reaction, in which
the tight alcohol binding would cause product inhibition.
(ii) PI-PLC catalyzes the cleavage of PI to IcP with kcat
approaching 1200 s-1 14
but the subsequent hydrolysis of IcP
,
to IP is at least 500 times slower.11 The cleavage of the non-
hydrophobic inositol phosphodiesters synthesized in this work
is still several-fold slower than the hydrolysis of IcP. The large
reduction in the cleavage rate of such non-hydrophobic RO-PI
is probably due to an absence of the interfacial activation of
PI-PLC with non-hydrophobic substrates (low kcat).33 Unlike
with other phospholipases, the observed interfacial activation
of PI-PLC, however, is relatively weak.33 For example, the
cleavage of dihexanoyl-PI with the bacterial and mammalian
δ-type PI-PLC is activated only 2-33b and 5-fold,33c respectively,
at the critical micelle concentration. In contrast, the nonaggre-
gating dibutyroyl-PI is completely resistant to PI-PLC-δ,33c
glycerophosphoinositol 4-phosphate is resistant to PI-PLC-R,34
and Gro-IP is almost fully resistant to bacterial PI-PLC35 (see
also this work). It, thus, appears that both bacterial and
mammalian enzymes display three regions of activity depending
on the substrate hydrophobicity: (i) very low activity with
hydrophilic substrates bearing an equivalent of eight methylene
residues or less in their alcohol leaving group; (ii) ca. 20-50%
activity with more hydrophobic substrates below cmc; and (iii)
full activity with hydrophobic substrates above cmc. The large
gap in reactivity between non-hydrophobic substrates such as
Gro-IP, or dibutyroyl-PI, and dihexanoyl-PI below cmc can be
explained by the enzyme-induced aggregation of dihexanoyl-
PI at submicellar concentration, but not Gro-IP, nor dibutyroyl-
PI. Since the binding of an acyclic alkyl inositol phosphate is
likely to be comparable to that of PI,34 it is most probable that
the reduction in the kcat is responsible for the resistance of non-
hydrophobic alkyl inositol phosphates to cleavage by PI-PLC.
(iii) The thermodynamic equlibrium between the cyclic
phosphodiester and monoester is clearly shifted in favor of the
monoester36 due to favorable solvation factors, and the earlier
discussed ring strain (see Results). Similarly, the equilibrium
between IcP and RO-PI should be shifted toward RO-PI due to
the ring strain present in IcP.
To prove the applicability of this reaction to syntheses of
more complex compounds we have used PLC to attach inositol
phosphate residue to carbohydrates, nucleosides, and peptides.
With hexoses and ribonucleosides only the primary 6-hydroxyl
and 5′-hydroxyl groups, respectively, were phosphorylated. In
contrast, the reaction was very regio-nonspecific with regard
to primary hydroxyl groups in different environments. This was
demonstrated by phosphorylation of the lactose and Ser-Tyr-
Ser-Met tetrapeptide. The results indicate that the enzyme can
phosphorylate indiscriminately hydroxymethylene groups of
serine and hexoses, but not the tyrosine hydroxyl. It is
conceivable that this reaction could be extended to attach inositol
phosphate moiety to solvent-exposed serines of larger peptides
or even proteins, and various oligosaccharides. Although we
have failed to phosphorylate the long chain fatty alcohols, the
phosphorylation of the non-hydrophobic long-chain alcohols
such as pentaethylene glycol (PEG) can be achieved. Such
alcohols can be used as flexible linkers to attach inositol
phosphate to various functionalities. In the past various
radioactive and fluorescently labeled inositol phosphate deriva-
tives have been synthesized and used as molecular
probes.3,5a,6,7,57-59 In the current work, we have synthesized
the biotinylated PEG inositol phosphate (19) to be used as a
ligand for an affinity chromatography of proteins with affinity
for the inositol phosphate residue. An extension of this
experiment for synthesis of a fluorescently labeled inositol
phosphodiesters can be easily envisioned.
Transterification as a Common Property of Phospho-
esterases which Form Ester Intermediates. Phosphatidyl-
inositol-specific phospholipase C is a member of a large group
of phosphoesterases which employ a transesterification mech-
anism for the phosphoester bond cleavage.37 These enzymes
can be broadly divided into two groups: (I) those which use
the intrinsic serine/threonine/tyrosine hydroxyl group of the
active site as a nucleophile, exemplified by phospholipase
D,15-18,38,39 snake venom phosphodiesterase,21,22 and alkaline
Factors Affecting Efficiency of Transesterification. The
above described transesterification reaction is a reverse reaction
of PI-PLC, in which different alcohols substitute for the
diacylglycerol. The competing hydrolysis of IcP can be also
considered as a reverse reaction, with water molecule substitut-
ing for the diacylglycerol. The feasibility of the transesterifi-
cation reaction as a source of O-alkyl inositol phosphates
depends on three factors: (i) kinetic competition between diacyl
glycerol, externally added alcohol and water molecules; (ii) the
ability of a product of the reverse reaction to serve as a substrate
for the forward reaction; and (iii) the equilibrium constant
between O-alkyl inositol phosphate (RO-PI), and ROH and IcP.
The importance of these factors is discussed below.
(33) (a) Volwerk, J. J.; Filhuth, E.; Griffith, O. H.; Jain, M. K.
Biochemistry 1994, 33, 3364-3474. (b) Lewis, K. A.; Garigapati, V. R.;
Zhou C.; Roberts, M. F. Biochemistry 1993, 32, 8836-8841. (c) Rebecchi,
M. J.; Eberhardt, R.; Delaney, T.; Ali, S.; Bittman, R. J. Biol. Chem. 1993,
268, 1735-1741.
(34) Cruz-Rivera, M.; Bennett, C. F.; Crooke, S. T. Biochim. Biophys.
Acta 1990, 1042, 113-118.
(i) The fact that the rate of IcP hydrolysis is close to the rate
of formation of RO-PI with alcohol concentration in the 0.5 M
range (Figure 4A) suggests that alcohol can efficiently compete
with water for the diacylglycerol binding site. This is reflected
by the magnitude of the efficiency parameter in the range of
6-57 for several studied alcohols, showing that alcohol is a
(35) Shashidhar, M. S.; Volwerk, J. J.; Griffith, O. H.; Keana, J. F. W.
Chem. Phys. Lipids 1991, 60, 101-110.
(36) Cheung, W. I.; Patrick, S. M.; Pennington, S. N. Int. J. Biochem.
1974, 5, 311-319.
(37) For a broad discussion of mechanisms of phosphotransferases, see:
Knowles, J. R. Annu. ReV. Biochem. 1980, 49, 877-919.

7686 J. Am. Chem. Soc., Vol. 118, No. 33, 1996
Bruzik et al.
Scheme 3
diester. The observed transesterification of alcohols by IcP in
the presence of PI-PLC falls, thus, into a general property of
group I and II phosphoesterases. It is interesting to note the
similarities between the transesterification reaction catalyzed by
PI-PLC to those of other phosphoesterases.
(A) Snake venom phosphodiesterase catalyzes the cleavage
of 3′,5′-dinucleotides and nucleoside triphosphates by a trans-
esterification mechanisms involving formation of the covalent
phosphoryl-enzyme intermediate.46,47 This enzyme also cata-
lyzes transesterification of various primary alcohols with
nucleoside triphosphates through alcohol attack at the reactive
phosphoryl-enzyme intermediate.21,22 The rate of transesteri-
fication is slow as compared to ATP cleavage and relatively
independent of the alcohol structure. In general, the alcohol
efficiency parameters for this enzyme22 are several-fold lower
than those for PI-PLC, observed in this work. Another group
I enzyme, phospholipase D, has been known to catalyze
transesterification (transphosphatidylation) with a broad variety
of primary alcohols,15,48 through formation of the phosphoryl-
enzyme intermediate.38 The recently reported ester group
exchange in phosphatidylcholine by PLD in the presence of
several hexitols and glucose is proposed to play a role in the
diabetic metabolism,49 while transphosphatidylation of ethanol
with phosphatidylcholine could be important in fetal alcohol
syndrome.50
(B) Ribonuclease A catalyzes transesterification of simple
primary alkanols with 2′,3′-cyclic nucleoside phosphates, at a
much reduced rate as compared to RNA cleavage.19 This
enzyme is related to the bacterial PI-PLC by the analogous
catalytic mechanism,30,40,51 and by the fact that the cyclic
phosphate is the physiological end-product of the RNA cleav-
age.40,52 The low rate of the cleavage of 2′,3′-cyclic phosphate,
mostly due to reduction in the kcat, results from the absence of
a part of the substrate structure needed to bind to the p2 site, to
achieve full activity.53,54 In the case of PI-PLC the analogous
rate reduction is brought about by the removal of the hydro-
phobic chains from the substrate.
phosphatase23 (Scheme 3I), and (II) those which use an internal
hydroxyl group of a substrate as an attacking nucleophile,
represented by ribonuclease A40-42 and PI-PLC8,30 (Scheme 3II).
Both groups of enzymes use two consecutive steps of nucleo-
philic displacement at phosphorus for a complete substrate-
product cycle, with formation of a phosphoryl-enzyme inter-
mediate (group I), or a cyclic phosphate (group II), respectively.
In addition, a third group of enzymes, represented by inositol
1-phosphate phosphatase,43,44 and the nonspecific-phospholipase
C45 uses an activated water molecule as a nucleophile for a direct
attack at the phosphorus atom (Scheme 3III). To date, of the
above three groups of enzymes only the first two have been
reported to catalyze transesterification with exogeneous alcohols,
most likely involving the reaction of such alcohols with the
enzyme-phosphoester intermediate, or the cyclic phosphate. This
is a consequence of the transesterification mechanisms used by
group I and II enzymes in their natural reactions. For example,
in the case of group II enzymes, the externally added alcohol
can undergo an exchange with an alcohol of a ternary complex
at the end of the first catalytic step (Scheme 4). The exogeneous
alcohol can then use the binding site formerly occupied by the
natural alkoxy leaving group. In addition, the protonation status
of the active site general base-general acid pair is reversed at
the end of the first step (i.e., B1 is protonated, and B2 is
deprotonated, Schemes 3II and 4), as compared to the initial
situation, enabling nucleophilic activation of the incoming
alcohol. In the resulting transesterification reactions the en-
zymes therefore catalyze conversion of a diester into a diester,
an energetically benign process. In the case of group III
enzymes, an exogeneous alcohol molecule would have to be
accommodated in place of a water molecule (Scheme 3III), an
unlikely scenario due to a greater steric size of an alcohol as
compared to water. Alternatively, for the reverse reaction of
this group of enzymes, the phosphate monoester product (in
the case of diesterase) would have to be converted into a
phosphate diester, an energetically disadvantageous process due
to a higher solvation energy of a monoester as compared to a
In summary, we have shown a potential of PI-PLC in
synthesis of alkyl inositol phosphodiesters. One of the possible
areas of application of the produced compounds is in studies
of structure-activity relationship of inositol-related enzymes
such as PI-synthase,55 or inositol 1-phosphate phosphatase.7 Both
of these enzymes take part in inositol recycling, and thus they
could constitute targets of antisignaling drug design.
Experimental Section
1H NMR spectra were recorded in D2O with a Bruker AM-400, AC-
250, AC-200, and a Varian XL-300 spectrometers and were indirectly
referenced to TMS. 31P NMR spectra were recorded in D2O with a
Bruker AC-250 and AC-200 spectrometers and were indirectly
referenced to 85% phosphoric acid. MS spectra were obtained with a
Hewlett Packard 5989B electrospray mass spectrometer. HPLC
(46) Bryant, F. R.; Benkovic, S. J. Biochemistry 1979, 18, 2825-2827.
(47) Cummins, J. H.; Potter, B. V. L. Eur. J. Biochem. 1987, 162, 123-
128.
(38) Bruzik, K. S.; Tsai, M.-D. Biochemistry 1984, 23, 1656-1661.
(39) Kanfer, J. N. Can. J. Biochem. 1980, 58, 13709-1380.
(40) Thompson, J. E.; Venegas, F. D.; Raines, R. T. Biochemistry 1994,
33, 7408-7414.
(48) Heller, D. AdV. Lipid. Res. 1978, 16, 267-326.
(49) Nakamura, J.; Luttimer, S. A.; Greene, D. A. Diabetologia 1994,
37, 1147-1153.
(50) Bondeson, J.; Sundler, R. Biochim. Biophys. Acta 1987, 899, 258-
(41) Hershlag, D. J. Am. Chem. Soc. 1994, 26, 11631-11635.
(42) Breslow, R. Acc. Chem. Res. 1991, 24, 317-324.
(43) Bone, R.; Frank, L.; Springer, J. P.; Pollack, S. J.; Osborne, S.;
Atack, J.R.; Knowles, M. R.; McAllister, G.; Ragan C. I.; Broughton, H.
B.; Baker, R.; Fletcher, S. R. Biochemistry 1994, 33, 9460-94671
(44) Pollack, S. J.; Knowles, M. R.; Atack, J. R.; Broughton, H. B.;
Ragan, C. I.; Osborne, S. A.; McAllister, G. Eur. J. Biochem. 1993, 217,
281-287.
264.
(51) Thompson, J. E.; Raines, R. T. J. Am. Chem. Soc. 1994, 116, 5467-
5468.
(52) Cuchillo, C. M.; Pares, X.; Guasch, A.; Barman, T.; Travers, F.;
Nogues, M. V. FEBS Lett. 1993, 3, 207-210.
(53) Boix, E.; Nogues, M. V.; Schein, C. H.; Benner, S. A.; Cuchillo,
C. M. J. Biol. Chem. 1994, 269, 2529-2534.
(54) Steyaert, J.; Engelborghs, Y. Eur. J. Biochem. 1995, 233, 140-
144.
(45) Hansen, S.; Hough, E.; Svensson, L. A.; Wong, Y.-L.; Martin, S.
F. J. Mol. Biol. 1993, 234, 179-187.
(55) Antonsson, B. E. Biochem. J. 1994, 297, 517-522.

Synthesis of Phosphodiesters with Phospholipase C
J. Am. Chem. Soc., Vol. 118, No. 33, 1996 7687
Scheme 4
analyses were performed with a Hitachi L-6200A solvent delivery
system, and elutions were analyzed by the Hitachi HPLC Manager
software. For separation of nonchromophoric inositol phosphate esters
high-performance anion-exchange chromatography (HPAEC) was
employed. The Carbopack PA1 column (4.6 × 250 mm) equipped
with a guard column (4.6 × 50 mm) (Dionex) was eluted with a linear
gradient of sodium acetate (40-200 mM during 30 min) in isocratic
100 mM sodium hydroxide at the 1 mL/min flow rate. Elutions were
monitored with the Dionex pulsed amperometric detector (PAD-2) with
sensitivity set at 1 µA, and the ionization potentials set as E1 0.05 V,
E2 ) 0.6 V, E3 ) -0.6 V, and the sampling time set at 400 ms.
Separation and quantitation of chromogenic samples was performed
using the reverse phase chromatography and employing D-4500 UV-
vis photo-diode array detector (PDA) set at the 800 ms sampling time
and 7 nm spectral resolution. The reverse phase column Microsorb
MV-C18 (4.6 × 250 mm, 300 Å pore size) (Rainin) was eluted with
a solvent system consisting of 0.1% TFA in acetontrile/water (3:1,
component A) and 0.1% TFA in water (component B). The linear
gradient of 10% to 40% of component A during 30 min was applied at
the 1 mL/min flow rate. The reported chromatograms are for the 280
nm wavelength, unless indicated otherwise. Analytical thin-layer
chromatography (TLC) was performed on silica gel 60 F254 aluminum
plates (Merck). Plates were visualized under UV light (254 nm) or
with a 10% ethanol solution of phosphomolybdic acid. Dowex 1X8-
200 and 50WX8-200 ion-exchange resins were used for preparative
anion and cation exchange chromatography, respectively.
B. cereus PI-PLC was from Boehringer, and PI-PLC from B.
thuringiensis was expressed in E. coli as described earlier.56 Except
for some experiments shown in Figure 1 all reactions were performed
with the enzyme from B. thuringiensis. All compounds synthesized
have the 1D-configuration of the inositol moiety. Synthesis of
biotinylated pentaethylene glycol will be described elsewhere (Kubiak
and Bruzik, unpublished). All other chemicals and solvents were from
Aldrich or Sigma, and were used without additional purification.
Inositol 1,2-Cyclic Phosphate. (A) PI fraction of soybean phos-
pholipid obtained as described earlier29 (1 g, ca. 1.2 mmol) was
dispersed in the solution of sodium deoxycholate (20 mM, 100 mL),
containing EDTA (1 mM), and the pH was adjusted to 7.5 with 1 N
HCl. PI-PLC (20 µg) was added, and the reaction was stirred gently
at room temperature. The progress of the reaction was examined by
TLC (chloroform-methanol-NH4OH, 7:3:0.2) monitoring disappear-
ance of the PI spot, and by HPAEC monitoring formation of IcP and
IP. After all PI had been used up, but little IP had been formed, the
pH of the reaction mixture was adjusted to 5.0 with HCl to precipitate
SDC as a free acid. The mixture was extracted three times with
chloroform-methanol (5:2, 100 mL). Phases were separated by
centrifugation, and the final aqueous phase was freeze-dried. The
residue was chromatographed on the formate form-Dowex 1X8-200
anion-exchange resin using a step-gradient of ammonium formate. The
residual unextracted sodium deoxycholate was eluted off with water,
and the IcP product was eluted with 50 mM ammonium formate. This
fraction was freeze-dried to the give the final IcP product (27%
assuming 100% purity of the starting material). 1H NMR δ 4.51 (tr,
H-2, J ) 4.8 Hz, 1H), 4.12 (ddd, H-1, J ) 4.8, 7.86, 20.2 Hz, 1H),
3.60 (tr, H-6, J ) 8.4 Hz, 1H), 3.54 (ddd, H-3, J ) 2.0, 3.8, 10.0 Hz,
1H), 3.44 (tr, H-4, J ) 9.6, 1H), 3.09 (tr, H-5, J ) 9.8, 1H). In a
larger scale preparation the cleavage of 20 g of phosphoinositide fraction
performed analogously gave 1.2 g of purified IcP.
(B) The soybean lecithin (12 g, Sigma) was solubilized in ether-
water mixture (1:1, 800 mL). To this heterogeneous mixture was added
with PI-PLC (30 µg), and the mixture was stirred at room temperature.
The progress of the cleavage was followed by TLC, monitoring the
formation of diacylglycerol (hexane-acetone, 8.5:1.5; Rf 0.3). The
increase in diacylglycerol stopped after ca. 12 h. The mixture was
extracted three times with chloroform-methanol (1:1 v/v, 500 mL),
and the aqueous phase was concentrated to dryness under vacuum. The
residue was chromatographed on the anion exchange column (Bio-
Rad AG 1X4, 25 × 200 mm, formate form) collecting the 250 mM
ammonium formate fraction (300 mL). This fraction was concentrated
and re-evaporated several times with water to remove ammonium
formate to give the essentially pure IcP (1.1 g), containing ca. 16% of
ammonium formate. This product was used for further reactions
without additional purification.
Enzymatic Synthesis of O-Alkyl Inositol Phosphates: General
Procedure. IcP (10-20 mg) was dissolved in deionized water (0.5
mL) containing a desired primary alcohol (80 mM-6.0 M). The pH
of the reaction solvent was adjusted to 6.5-7.5 with a dilute HCl and
NH4OH. PI-PLC (50 µL, containing 10 µg of the protein) was added,
and the progress of the reaction was followed by HPAEC, HPLC, or
31P NMR. All reactions were run at ambient temperature. The reactions
were stopped by adding 1 N HCl to pH 3, when the concentration of
the phosphodiester product reached maximum. The mixture was loaded
onto the Dowex 50WX8-200 cation-exchange column (6 × 60 mm),
previously equilibrated with ammonium formate, and 10 mL of solution
was collected and freeze-dried. The residue was redissolved in water
(0.5 mL) and chromatographed on Dowex 1X8-200 anion exchange
column (5 × 100 mm), formerly equilibrated with ammonium formate
solution. The first fraction (15 mL) eluted off with water contained
only an alcohol (in the case of nonionic alcohols). The product was
eluted off typically with 30-50 mM ammonium formate (10-15 mL).
This fraction was evaporated to dryness and analyzed by HPAEC, 31
and H NMR, and ES-MS. The products were frequently found to
contain significant amount of ammonium formate, in which case they
were redissolved in water and reconcentrated again.
P
1
With products of transesterification of certain alcohols (such as PEG-
biotin) no suitable separation/detection method could be found to
monitor the progress of transesterification. In this case, the reaction
was continued until the concentration of IcP dropped to ca. 10% of its
original value (as observed by HPAEC). The mixture was then acidified
to pH 3, left for 2-3 h, and the IP byproduct was separated on the
Dowex anion exchange column as described above. The eluates were
monitored by HPAEC, and fractions containing IP were discarded. All
fractions eluted prior to IP were pooled and concentrated. The residue
was redissolved in small amounts of water, and the product was
precipitated with acetone.
Control Experiments. (A) IcP (5 mM) in Tris-HCl buffer (0.1 M,
pH 7.4) was stored at room temperature over the period of 4 days. (B)
The analogous sample containing additionally 1.0 M glycerol was stored
during 6 days. Under both conditions IcP was stable as determined
by 31P NMR (within 5% detection limit). (C) The PI-PLC preparation
(ca. 1 mg of protein per 1 mL) was dialyzed against doubly distilled
water, and the resulting enzyme solution (ca. 3.5 µg of protein) was
used to hydrolyze IcP (10 mM) in 70 mM MOPS-Na buffer (pH 7.0)
(56) Koke, J. A.; Yang, M.; Henner, D. J.; Volwerk, J. J.; Griffith, O.
H. Prot. Expression Purif. 1991, 2, 51-58.
(57) Estevez, V. A.; Prestwich, G. D. J. Am. Chem. Soc. 1991, 113,
9885-9887.
(58) Schaeffer, R.; Nehls-Sahabandu, M.; Grabowsky, B.; Dehlinger-
Kremer, M.; Schulz, I.; Mayr, G. W. Biochem. J. 1990, 272, 817-825.
(59) Hirata, M.; Watanabe, Y.; Ishinatsu, T.; Yanaga, F.; Koga, T.; Ozaki,
S. Biochem. Biophys. Res. Commun. 1990, 168, 379-386.

7688 J. Am. Chem. Soc., Vol. 118, No. 33, 1996
Bruzik et al.
1
containing 10 mM EDTA-Na. 31P NMR spectra revealed formation
of only inositol 1-phosphate (within 2% detection limit).
D-Mannito-1-phospho-(1-myo-inositol) (13): H NMR δ 4.07 (tr,
H-2, J ) 2.7 Hz, 1H), 4.0 (m, H′-1, 1H), 3.89 (m, H′-1, 1H), 3.78 (dtr,
H-1, J ) 10.8, 2.8 Hz, 1H), 3.6 (m, 3H), 3.58-3.54 (m, 3H), 3.51-
3.41 (m, 3H), 3.35 (dd, H-3, J ) 10, 2.7 Hz, 1H), 3.12 (tr, H-5, J )
9.3 Hz, 1H); 31P NMR δ 0.99 ppm; ESMS m/z 423.0.
sn-Glycero-3-phospho-(1′-myo-inositol) (4). The solution of 1,2-
dipalmitoyl-sn-glycero-3-phospho-1′-myo-inositol (17 mg, synthesized
as described earlier8 in butanol-methanol-water, 3:5:1) was added
with methylamine/toluene (30%, 0.4 mL) and kept at 55 °C during 3
h. The mixture was diluted with water and extracted with chloroform,
and the aqueous phase was evaporated to dryness. The residue was
redissolved in water and passed through H-form Dowex 50X8-200,
and the eluate was freeze-dried to give the pure diester 4.
L-Idito-1-phospho-(1-myo-inositol) (14): 1H NMR δ 4.23 (tr, H-2,
J ) 2.6 Hz, 1H), 4.02-3.9 (m, H-1′, H-2′, H-1, 4H), 3.84 (m, H-3′,
1H), 3.78-3.6 (m, 6H), 3.53 (dd, H-3, J ) 10.0 Hz, 1H), 3.31 (tr,
H-5, J ) 9.5 Hz, 1H); 31P NMR δ 0.63 ppm; ESMS m/z 423.0.
O-1-(3-Aminopropyl) 1-myo-inositol phosphate (15): 1H NMR δ
4.04 (tr, H-2, J ) 2.4 Hz, 1H), 3.85 (trd, H-1′, overlapped, 2H), 3.75
(ddd, H-2, J ) 2.9, 9.7, 12.4 Hz, 1H), 3.55 (tr, H-6, J ) 9.6 Hz, 1H),
3.46 (tr, H-4, J ) 9.3 Hz, 1H), 3.35 (dd, H-3, J ) 2.5, 10.1 Hz, 1H),
3.13 (tr, H-5, J ) 9.2 Hz, 1H), 2,96 (tr, H-3′, J ) 7.3 Hz, 2H), 1.82 (p,
H-2′, J ) 6.6 Hz, 2H); 31P NMR δ 0.68 ppm; ESMS m/z 317.0.
Cholinephospho-(1-myo-inositol) (16): 1H NMR δ 4.28 (m, H-1′A,
H-1′B, 2H), 4.15 (tr, H-2, J ) 2.8 Hz, 1H), 3.87 (ddd, H-1, J ) 2.8,
8.4, 9.9 Hz, 1H), 3.67 (tr, H-6, J ) 9.7 Hz, 1H), 3.58 (m, H-2′, 2H),
3.56 (tr, H-4, 9.3 Hz, 1H), 3.47 (dd, H-3, J ) 2.7, 10 Hz, 1H), 3.24
(tr, H-5, J ) 9.3 Hz, 1H), 3.13 (s, Me, 9H); 31P NMR δ -1.06; ESMS
m/z 329.9.
4: 1H NMR (500 MHz, signal assignment assisted by 1H-1H COSY)
δ 4.45 (tr, H-2, J ) 2.8 Hz, 1H), 4.2-4.17 (H-1, H-2′, H-3′′, m, 4H),
3.94 (tr, H-6, J ) 9.7 Hz, 1H), 3.8 (dd, H-1′A, J ) 4.5, 11.9 Hz, 1H),
3.84 (tr, H-4, J ) 9.8 Hz, 1H), 3.79 (dd, H-1′B, J ) 6.0, 11.8 Hz, 1H),
3.74 (dd, H-3, J ) 2.8, 10.0 Hz, 1H), 3.52 (tr, H-5, J ) 9.4 Hz, 1H);
13C NMR δ 76.42 (d, J ) 6.1 Hz), 73.93, 72.2, 71.32 (d, J ) 6.5 Hz),
71.15, 70.73, 70.67 (d, J ) 6.7 Hz), 66.52 (d, J ) 5.5 Hz), 62.05; 31
NMR δ 0.14 ppm; ESMS m/z 333.8.
P
O-1-(2-Amino-2-hydroxymethylene-3-hydroxypropyl) 1-myo-
1
inositol phosphate (5): H NMR (500 MHz) δ 4.12 (tr, H-2, J ) 2.8
Hz, 1H), 3.91 (m, H-1′A,B, 2H), 3.85 (ddd, H-1, J ) 2.8, 8.1, 10.9
Hz, 1H, 3.62 (tr, H-6, J ) 9.6 Hz, 1H), 3.6 (m, H-3′, H-4′, 4H), 3.49
(dd, H-4, J ) 9.0, 9.9 Hz, 1H); 3.42 (dd, H-3, J ) 2.8, 9.9 Hz, 1H),
3.20 (tr, H-5, J ) 9.3 Hz, 1H); 31P NMR δ -1.56 ppm.
L-Serine-3-phospho-(1-myo-inositol) (17): 1H NMR δ 4.09 (tr, H-1′,
J ) 5.4 Hz, 2H), 4.04 (tr, H-2, J ) 2.7 Hz, 1H), 3.81 (tr, H-2′, J )
4.29 Hz, 1H), 3.75 (ddd, H-1, J ) 12.00, 9.00, 3.00 Hz, 1H), 3.53 (tr,
H-6, J ) 9.8 Hz, 1H), 3.41 (dd, H-4, J ) 18.6, 9.3 Hz, 1H), 3.35 (dd,
H-3, J ) 10.02, 1.8 Hz, 1H), 3.12 (tr, H-5, J ) 9.3 Hz, 1H); 31P NMR
δ -0.74 ppm; ESMS m/z 346.
O-1-(6-Hydroxyhexyl) 1-myo-inositol phosphate (6): 1H NMR
(D2O) δ 4.23 (tr, H-2, J ) 2.5 Hz, 1H), 3.91 (tr, H-1′, partially
overlapped with H-1, 2H), 3.90 (ddd, H-2, partially overlapped with
H-1′, 1H), 3.73 (tr, H-6, J ) 10.0 Hz, 1H), 3.64 (tr, H-4, J ) 9.5 Hz,
1H), 3.59 (tr, H-6′, J ) 6.5 Hz, 2H), 3.55 (dd, H-3, J ) 3.0, 10.0 Hz,
1H), 3.31 (tr, H-5, J ) 9.5 Hz, 1H), 1.63 (p, H-2′, J ) 7.0 Hz, 2H),
1.55 (p, H-5′, J ) 7.0 Hz, 2H), 1.38-1.35 (m, H-3′, H-4′, overlapped,
4H); 31P NMR δ 0.84; ESMS m/z 359.8.
O-Methyl 1-myo-inositol phosphate (7): 1H NMR δ 4.15 (tr, H-2,
J ) 2.8 Hz, 1H), 3.84 (ddd, H-1, J ) 2.8, 8.2, 9.9 Hz, 1H), 3.65 (tr,
H-6, J ) 9.7 Hz, 1H), 3.56 (dd, H-4, J ) 9.3 Hz, 10 Hz, 1H), 3.52
(dd, Me, J ) 10.8 Hz, 3H), 3.46 (dd, H-3, J ) 2.8, 10.0 Hz, 1H), 3.23
(tr, H-5, J ) 9.2 Hz, 1H); 31P NMR δ 1.1; ESMS 273.6.
O-n-Propyl 1-myo-inositol phosphate (8): 1H NMR δ 4.16 (tr, H-2,
J ) 2.6 Hz, 1H), 3.85 (ddd, H-1, J ) 2.8, 8.3, 10.0 Hz, 1H), 3.76 (q,
H-1′, J ) 6.4 Hz, 2H), 3.66 (tr, H-6, J ) 9.6 Hz, 1H), 3.57 (tr, H-4,
J ) 9.8 Hz, 1H), 3.47 (dd, H-3, J ) 2.7, 9.8 Hz, 1H), 3.24 (tr, H-5, J
) 9.3 Hz, 1H), 1.55 (q, H-2′, J ) 7.3 Hz, 2 H), 0.83 (tr, Me, J ) 7.4
Hz, 3H); 31P NMR δ 0.29; ESMS m/z 301.0 (M - H+).
(Pentaethyleneglyco)phospho-1-myo-inositol (18): 1H NMR δ 4.24
(tr, H-2, 1H), 4.05 (m, CH2OP, 2H), 3.94 (ddd, H-1, 1H), 3.68-3.77
(m, OCH2CH2O, 16H), 3.63 (m, H-4, H-6, CH2OH, 4H), 3.53 (dd, H-3,
1H), 3.31 (tr, H-5, 1H); 31P NMR δ 0.49 ppm; ESMS m/z 479.0 (M -
H+).
(Biotinylpentaethyleneglyco)phospho-1-myo-inositol (19): 1H NMR
δ 4.60 (dd, 1H), 4.42 (dd, 1H), 4.25 (m, CH2OCO, H-2, 3H), 4.05 (m,
CH2OP, 2H), 3.94 (ddd, H-1, 1H), 3.60-3.80 (m, OCH2CH2O, 16H),
3.52 (dd, H-3, 1H), 3.78 (m, 2H), 2.96 (dd, 1H), 2.75 (d, 1H), 2.42 (tr,
2H), 1.32-1.80 (m, 6H); 31P NMR δ 0.44 ppm; ESMS m/z 706.
Glucose-6-phospho-(1-myo-inositol) (20): 5.20 (d, J ) 4.8 Hz),
4.62 (d, J ) 8.0 Hz), 4,23 (tr, J ) 2.7 Hz, H-2, 1H), 4.16 (ddd, J )
5.3, 9.6, 1 Hz), 4.11 (tr, J ) 4.9 Hz), 4.06 (q, J ) 5.3 Hz), 3.93 (m,
H-1, 1H), 3.71 (tr, J ) 9.7 Hz, 1H), 3.61 (tr, J ) 9.7 Hz, 1H), 3.53-
3.43 (m, 3H), 3.29 (tr, J ) 9.4 Hz, 1H), 3.25 (m); 31P NMR δ 0.65
ppm; ESMS m/z 421.4.
1
Uridine-5′-phospho-(1-myo-inositol) (22): H NMR δ 7.89 (d, J
O-n-Butyl 1-myo-inositol phosphate (9): 1H NMR δ 4.20 (tr, H-2,
J ) 2.6 Hz, 1H), 3.88 (trd, H-1′, partially overlapped with H-1, 2H),
3.89 (ddd, H-1, partially overlapped with H-1′, 1H), 3.71 (tr, H-6, J )
9.6 Hz, 1H), 3.61 (tr, H-4, J ) 9.8 Hz, 1H), 3.50 (dd, H-3, J ) 2.8,
10.0 Hz, 1H), 3.28 (tr, H-5, J ) 9.4 Hz, 1H), 1.57 (p, H-2′, J ) 6.7
Hz, 2H), 1.33 (s, H-3′, J ) 7.8 Hz, 2H), 0.87 (tr, H-4′, J ) 7.2 Hz,
3H); 31P NMR δ 0.86 ppm; ESMS m/z 315.5.
O-Allyl 1-myo-inositol phosphate (10): 1H NMR δ 5.95 (m, H-2′,
J ) 5.28 Hz, 1H), 5.33 (dd, H-3′A, J ) 1.1, 17.1 Hz, 1H), 5.19 (dd,
H-3′B, J ) 1.1, 10.5 Hz, 1H), 4.40 (m, 2H), 4.20 (tr, H-2, J ) 2.8 Hz,
1H), 3.90 (ddd, H-1, J ) 2.8, 8.4, 11.1 Hz, 1H), 3.70 (tr, H-6, J ) 9.6
Hz, 1H), 3.61 (tr, H-4, J ) 9.8 Hz, 1H), 3.50 (dd, H-3, J ) 2.8, 10.0
Hz, 1H), 3.28 (tr, H-5, J ) 9.4 Hz, 1H); 31P NMR δ 0.63 ppm; ESMS
m/z 299.4.
O-(2-Hydroxyethyl) inositol phosphate (11): 1H NMR δ 4.18 (tr,
H-2, J ) 2.6 Hz, 1H), 3.88 (ddd, H-1 partially overlapped with H-1′,
1H), 3.68 (m, H-2′, 2H), 3.66 (tr, H-6 partially overlapped with H-2′,
J ) 9.6 Hz, 1H), 3.56 (dd, H-4, J ) 9.4, 9.9 Hz, 1H), 3.47 (dd, H-3,
J ) 2.7, 9.9 Hz, 1H), 3.24 (tr, H-5, J ) 9.3 Hz, 1H); 31P NMR (D2O)
δ 0.14 ppm; ESMS m/z 303.4.
O-1-[(2S)-Hydroxypropyl)] 1-myo-inositol phosphate (12): 1H
NMR δ 4.17 (tr, H-2, J ) 2.9 Hz, 1H), 3.92 (qdd, H-2′ partially
overlapped with H-1, 1H), 3.88 (ddd, H-1, J ) 3.0, 8.5, 10.2 Hz, 1H),
3.81 (ddd, H-1′A, J ) 3.4, 5.7, 10.6 Hz, 1H), 3.66 (tr, H-6 partially
overlapped with H-1′B, J ) 9.3 Hz, 1H), 3.65 (ddd, H-1′B, J ) 6.3,
7.0, 10.6 Hz, 1H), 3.56 (tr, H-4, J ) 9.5 Hz, 1H), 3.46 (dd, H-3, J )
2.8, 10.0 Hz, 1H), 3.24 (tr, H-5, J ) 9.3 Hz, 1H), 1.08 (d, H-3′, J )
6.4 Hz, 3H); 31P NMR δ 0.63 ppm; ESMS m/z 317.7.
) 8.0 Hz, 1H), 5.92 (d, H-1′, J ) 4.7 Hz, 1H), 5.90 (d, J ) 8.0 Hz,
1H), 4.31 (m, H-5′, 2H), 4.2 (m, H-2 overlapped with H-4′ and H-2′,
3H), 4.09 (m, H-3′, 1H), 3.92 (m, H-1, 1H), 3.71 (tr, H-6, J ) 9.6 Hz,
1H), 3.61 (tr, H-4, J ) 9.8 Hz, 1H), 3.50 (dd, H-3, J ) 2.8, 6.0 Hz,
1H), 3.27 (tr, H-5, J ) 9.4 Hz, 1H); 31P NMR δ -0.47 ppm; ESMS
m/z 485.
Transesterification of Ser-Tyr-Ser-Met with IcP. The solution
(50 µL) of tetrapeptide (1 mg) in water was treated with PI-PLC solution
(4 µL, 1 mg/mL). The progress of the reaction was monitored by RP-
HPLC on C18 column using gradient elution as specified earlier. The
reaction was stopped after 48 h and the mixture subjected to LC-ESMS,
using chromatographic conditions as described earlier.
Acknowledgment. We would like to thank Dr. Richard van
Breemen of UIC for generating ES MS spectra, Mr. Robert J.
Kubiak for preparation of biotinylated pentaethylene glycol, and
Dr. T. L. Rosenberry of Case Western University for generous
gift of B. thuriupiensis PI-PLC. This work was supported by
Research Grant GM 30327 from National Institutes of Health
(M.-D.T., S.R., and K.S.B.) and by the Department of Medicinal
Chemistry and Pharmacognosy of UIC (K.S.B. and Z.G.).
Supporting Information Available: 1H and 31P NMR
spectra and ESMS data for compounds 2-23 synthesized in
this work (27 pages). See any current masthead page for
ordering and Internet access instructions.
JA9616084