Mapping phorbol ester binding domains of protein kinase C (PKC): The design, synthesis and biological activity of novel phorbol ester dimers

Article abstract of DOI:10.1055/s-1999-3656

The design and synthesis of a new class of protein kinase C (PKC) activators, phorbol ester dimers, is described. These dimers were found to bind to and activate PKC with affinities depending dramatically on their tether length and composition. In two cas

Full text of DOI:10.1055/s-1999-3656

PAPER
1401
Mapping Phorbol Ester Binding Domains of Protein Kinase C (PKC):
The Design, Synthesis and Biological Activity of Novel Phorbol Ester Dimers
*
a
a
a
b
Paul A.Wender, Michael F.T. Koehler, Dennis L. Wright, Kazuhiro Irie
a
Department of Chemistry, Stanford University, Stanford, California 94305–5080, USA
Fax +1(650)7250259; E-mail: Wender@leland.stanford.edu
b
Department of Applied Life Sciences, School of Agriculture, Kyoto University, Kyoto 606–8502, Japan
Received 8 April 1999
binding subunits would be linked together with a tether of
appropriate length and composition. The two subunits
could thus bind to individual CRDs or to one CRD and a
Abstract: The design and synthesis of a new class of protein kinase
C (PKC) activators, phorbol ester dimers, is described. These
dimers were found to bind to and activate PKC with affinities de-
pending dramatically on their tether length and composition. In two secondary site. To the extent that the CRDs or secondary
cases, the binding affinities of these novel compounds exceeded binding sites differ among the isozymes, isozyme selec-
that of phorbol dibutyrate.
tive binding would be possible. Our interest in this ap-
Key words: bioorganic chemistry, dimerizations, esterifications, proach to PKC regulation was further stimulated by
antitumor agents, lipids
studies on a PKC-g knockout mouse which suggest a cru-
cial role for this isozyme in mediating neuropathic pain.6
Over the past two decades, the protein kinase C isozymes
(
PKCs) have emerged as one of the most important fami-
1
lies of enzymes involved in cellular regulation. These
serine/threonine kinases have been implicated in a diverse
array of biological processes ranging from learning to tu-
mor promotion and are being pursued as therapeutic tar-
gets in the treatment of cancer, diabetes and neuropathic
pain. It appears that this diversity of function derives from
the fact that at least 11 PKC isozymes are expressed and
that individual isozymes can have different functions.
Therefore, the identification of isozyme selective agents is
of great fundamental and therapeutic importance.
Spacer Domain
phorbol
CRD 1
phorbol
CRD 2
Protein Kinase C
The PKCs are grouped according to their activator and co-
2
factor requirements: the conventional isozymes (a, bI,
bII, g) are activated by diacyl glycerol (DAG) and calci-
Figure 1 Design Concept for Isozyme Selective Compounds. Two
separate ligands for CRD1 and CRD2 are covalently linked through a
um, the novel isozymes (d, e, q) are also activated by tether which serves to properly position the ligands relative to the
DAG but are calcium independent, and the atypical (z, l) CRDs (or secondary binding sites) as well as to the membrane.
PKCs require neither DAG nor calcium for activation.
The conventional and novel PKCs bind DAG and other
The goals of the initial study reported herein were to syn-
activators using their first conserved (C1) region, which
thesize a series of dimeric PKC activators and to deter-
contains two homologous cysteine rich domains (CRD1
mine if they bound and activated PKC in a fashion
and CRD2). These CRDs are also referred to individually
different from the corresponding monomeric phorbol es-
as C1A and C1B.
ters as a prelude to and screen for more complex isozyme
In 1995, we demonstrated that a machine synthesized 51-
selective functional assays. Initially, it was decided to ex-
amine ligands which potently bind PKC as monomers to
better enable us to detect weak binders as we refined our
designs. With this in mind, we selected the phorbol esters,
which are known to be nanomolar binders of the CRDs,
and additionally have several functional groups which
could serve to link two monomers together. It should be
noted that this concept can also be generally applied to
other PKC activators to produce homodimeric (e.g. DAG-
mer corresponding to PKC-g
, (i.e. PKC-g C1B),
1
01–151
binds phorbol esters with affinities and phospholipid re-
3
quirements comparable to the natural protein. These
studies have more recently been extended to include the
3
3
synthesis and [ H]-phorbol 12,13 dibutyrate ([ H]-PDBu)
binding assay of both CRDs of all conventional and novel
4
PKCs. Intriguingly, in all but one of the isozymes, only
the second CRD binds phorbol esters with high affinity. In
the exception, PKC-g, the first and second CRDs bind
linker-DAG) or heterodimeric (e.g. DAG-linker-phorbol)
5
phorbol esters with comparably high affinities. This find-
conjugates. 7
ing provides the conceptual basis for the design of a new
class of potentially isozyme selective agents in which two
Synthesis 1999, No. SI, 1401–1406 ISSN 0039-7881 © Thieme Stuttgart · New York

1
402
P. A. Wender et al.
PAPER
OH
OH
13
12
OH
H
OAc
H
H
H
HO2C-(spacer)-CO2H, Et3N,
2,4,6-trichlorobenzoyl chloride;
9
Ac2O, Et3N, CH2Cl2/THF
83%
OH
OH
4
3
then 2, DMAP, PhCH3, 83-90%
O HO
O HO
OH
OAc
20
Phorbol (1)
2
[
spacer]
3
3
3
3
3
a: spacer=(CH ) , R=Ac
b: spacer=(CH2)10, R=Ac
2 6
O
H
O
O
H
O
c: spacer=(CH ) , R=Ac
2 12
d: spacer=(CH2)20, R=Ac
HClO4
MeOH
55-85%
e: spacer=CH O(CH ) O(CH ) OCH , R=Ac
OAc
H
OAc
H
2
2 2
2 2
2
4
4
4
a: spacer=(CH ) , R=H
b: spacer=(CH ) , R=H
c: spacer=(CH2)12, R=H
2 6
OH
OH
2 10
O HO
O HO
4d: spacer=(CH2)20, R=H
e: spacer=CH O(CH ) O(CH ) OCH , R=H
OR
OR
4
2
2 2
2 2
2
Scheme Synthesis of phorbol ester dimers 4a–e using a modular synthetic approach.
We previously reported a comprehensive analysis of the the diacid to the bis-mixed anhydride by the method of
structural features of phorbol esters that are required for Yamaguchi10 and direct exposure to the 13,20-diacetate
recognition at the PKC regulatory domain. According to routinely provided the dimers 3a–e in excellent yields
this analysis, the heteroatom triad whose spatial orienta- (83–90%). This dimerization protocol is general and sev-
tions correspond best with PKC binding are the oxygens eral different types of linkers have been successfully em-
at C4 (or C3), C9, and C20 as well as a lipophilic domain ployed including those containing rigid groups or
at C12 which is proposed to associate with the mem- heteroatom functionality in the tether.
8
brane. Based on this model, we sought to link two phor-
Final deblocking of the C20 acetate group is then required
bol moieties by esterification at the C12 position of
to generate the desired phorbol dimers. The high inherent
phorbol thereby introducing the crucial hydrophobic re-
reactivity of this position allows application of an acid
gion at C12 while leaving the putative pharmacophoric at-
catalyzed transesterification (HClO /MeOH) to deliver
4
oms free to interact with the protein. This modular
the unprotected dimers in just three steps overall from
approach provides dimeric activators wherein all three
phorbol. The resultant first generation dimers are pseudo-
functional domains, the two CRD binding ligands and the
C2 symmetric dimers in which both binding units are
spacer connecting them, can be readily varied. Our initial
identical. However, this synthetic scheme is easily ex-
synthetic efforts were directed at the incorporation of link-
tended to allow for the linking of different activators at the
ers of different length, while using phorbol esters on both
ends of the tether.
sides as our CRD binding domains (Scheme).
As a first test of whether these dimers would exhibit novel
Phorbol possesses five hydroxyl groups which can be se-
activities, we elected to examine their affinity in a conven-
lectively modified by careful choice of reaction condi-
3
tional binding assay based on the inhibition of [ H]-PDBu
9
tions. The allylic hydroxyl at C20 and the tertiary C13
binding to a PKC isozyme mixture isolated from a rat
hydroxyl group typically show the highest level of reac-
tivity, allowing for their direct conversion of phorbol to
the 13,20-bisacetate 2 by exposure to acetic anhydride and
triethylamine. The hydroxyl positioned at C12 is suffi-
ciently more reactive than the C4 and C9 hydroxyl groups
that dimer formation would be expected to occur at the
C12 position without further protection of the diacetate 2.
11
brain homogenate. In general, those dimers with largely
lipophilic linkers including polymethylene chains,
alkynes, or aromatic rings appeared highly active in pre-
liminary screens yielding apparent inhibition constants in
the low nanomolar region. Interestingly, compound 4e, in-
corporating a glycol based tether, failed to displace PDBu
at concentrations as high as 100 nM in these screens. This
Several methods for diester formation were examined in- is consistent with the tether subunit serving as a lipophilic
cluding the use of diacids activated with DCC or 2-io- anchor, facilitating binding of the phorbol domains at ei-
dopyridinium salts, as well as direct condensation with the ther end.
corresponding diacid chlorides. Although these methods
The simplest polymethylene derivatives 4a–d exhibited
did deliver the dimeric phorbol esters, the reactions were
the most potent binding, and they were thus selected for
somewhat capricious and typically yielded the dimers in
further evaluation. Standard competition binding assays
only moderate yields (40–50%). However, conversion of
3
employing [ H]-phorbol-12,13-dibutyrate as the reporter
Synthesis 1999, No. SI, 1401–1406 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
Mapping Phorbol Ester Binding Domains of Protein Kinase C (PKC)
1403
esterification at their C12 positions are able to competi-
tively bind to protein kinase C. Indeed, these are among
1
4
the tightest binders of PKC measured. Second, a signif-
icant relationship was observed between the length of the
tether employed and the binding constant that pointed to
an optimal length of 10–12 methylene groups. The differ-
ence of just 4 methylenes from dimer 4a to dimer 4b in-
creased the binding affinity 45 fold, suggesting that a
second binding site is being accessed.
An important issue related to this design is whether the
binding of these high affinity ligands actually results in
the activation or inhibition of PKC. To address this point,
a standard assay for enzymatic activity involving the in-
corporation of a radiolabeled phosphate onto a serine/
threonine residue of a suitable substrate was conducted.15
Under standard conditions and concentrations identical to
those used with phorbol-12-myristate-13-acetate, the
Figure 2 Incorporation of 32P into PKC substrate MBP
presence of 0.1 mM Ca , PS and 100 nM dimers 4a–d
in the
4–14)
(
2
+
dimers were found to promote phosphorylation of a mye-
1
2
molecule yielded dose-response curves for the 6 and 20
(4–14)
lin basic protein peptide fragment, MBP.
This indi-
methylene derivatives with K values in the low nanomo-
i
cates that their binding of PKC was accompanied by a
normal catalytic response (Figure 2) and the dimers can be
considered as activators of the enzyme.
lar region. However, dose-response curves generated for
the 10 and 12 methylene linkers produced data which, de-
spite good internal consistency, did not fit satisfactorily to
the theoretical curves for competitive binding, especially
at very low concentrations of inhibitor. In addition, it was
found that the observed IC values for these two dimers
This study represents the first synthesis and evaluation of
a new class of exceptionally potent dimeric activators of
PKC based on regulatory domain binding motifs. Prelim-
inary biological studies show that these compounds bind
to PKC with affinities varying in a manner dependent on
their tether length, and, in two cases, exceeding the affin-
ity of phorbol dibutyrate. They additionally activate the
enzyme in a manner comparable to phorbol esters. The
modular synthesis can deliver the compounds quickly
while retaining a high degree of efficiency, as is required
for combinatorial synthesis. The relationship between
tether length and PKC affinity as well as the molecules’
extreme potency suggest that these compounds are engag-
ing multiple binding sites on the enzyme. Synthesis of this
type of dimeric compound therefore provides a method
for mapping distances between binding sites. It is not yet
clear whether the enhanced binding is attributable to asso-
ciation of the phorbol ester subunits of the dimer with the
C1A and C1B subdomains or with one of these and an as
yet unrecognized region of PKC. However, to the extent
that these sites differ in distance or affinity in the different
isozymes, this study provides the basis for a rational ap-
proach to the design of PKC isozyme selective binders.
Efforts to differentiate these possibilities and to determine
5
0
was dependent on the concentration of PKC employed in
the assay.
Both of these observations are consistent with the occur-
rence of ligand depletion, which arises when high affinity
ligands, which must be tested at extremely low concentra-
tions, are used at concentrations so low that a significant
portion of that ligand is sequestered by its receptor in the
assay. As a result, the assumption that the amount of free
ligand in the assay solution is equal to the amount added
breaks down. Ligand depletion can be suppressed by
minimizing the concentration of the receptor in the assay
system, and so the binding assays were performed with
the lowest concentration of PKC possible while maintain-
ing good signal to noise ratios. Binding assays conducted
1
3
under these optimized conditions delivered K values in
i
the picomolar region for these designed activators. Com-
petition binding assays resulted in values of 4.5 nM, 0.1
nM, 0.4 nM and 2.7 nM for dimers 4a, 4b, 4c and 4d re-
spectively (Table). Under identical assay conditions, the
K for PDBu was determined to be 1.31 ± 0.24 nM.
d
These studies provide several important findings. First, the isozyme binding affinities of these new ligands are in
they demonstrate that phorbol esters dimerized through progress.
Oxygen- and moisture-sensitive reactions were carried out in oven
dried flasks sealed with rubber septa under a positive pressure of dry
Table 1 Inhibition Constants for the Binding of Phorbol Ester
Dimers to Rat Brain PKC
N or Ar. Correspondingly sensitive liquids were transferred via sy-
2
Dimer
n
K (nM) ± SD
ringe or cannula through rubber septa. Organic solutions were con-
centrated on a Büchi rotary evaporator connected to a ChemGlass
diaphragm pump, followed by exposure to high vacuum (<1 torr)
for at least 15 minutes.
i
4
4
4
4
a
b
c
6
4.5 ± 0.1
10
12
20
0.1 ± 0.06
0.4 ± 0.07
2.7 ± 0.02
All commercially available reagents were used without further pu-
d
rification unless otherwise noted. Et O and THF were distilled from
2
Synthesis 1999, No. SI, 1401–1406 ISSN 0039-7881 © Thieme Stuttgart · New York

1
404
P. A. Wender et al.
PAPER
Na–benzophenone ketyl prior to use. MeCN, CH Cl , diisopropy-
38.9, 34.5, 29.4, 29.2, 28.9, 25.6, 25.1, 23.8, 21.1, 20.9, 16.7, 14.4,
10.1.
2
2
lamine, Et N and pyridine were distilled from CaH . Toluene was
3
2
distilled over sodium metal. Flash chromatography grade hexanes
were distilled from technical grade hexanes.
IR(neat): n = 3405(m), 2931(m), 1724(s), 1626(w), 1376(s),
1
260(s).
HRMS (FAB) Calcd for C H O Na, 1113.5398; found,
6
0
82 18
Phorbol 13,20-Diacetate 2 (89%)
1
113.5456, TLC R = 0.26 (hexane/EtOAc: 1/1)
f
To a stirred solution of phorbol•EtOH (50.0 mg, 0.122 mmol) in
CH Cl (1 mL) and THF (1 mL) under N was added Et N (230 µL,
C14 Dimer-20 Acetate 3c
2
2
2
3
1
.65 mmol) followed by Ac O (0.15 ml, 1.59 mmol) and the solu-
From 1,14 tetradecanedioic acid diacid and phorbol 13,20-diace-
2
tion was allowed to stir for 8 h. The reaction was concentrated under
reduced pressure to give a dark orange oil which was immediately
chromatographed (1:1 hexane: EtOAc) to give the 13, 20 diacetate
tate, (90%)
1
H NMR (500 MHz, CDCl ): d = 7.63 (m, 1H), 5.74 (d, 1H, J =4
3
Hz), 5.56 (br s, 1H), 5.43 (d, 1H, J =10.5 Hz), 4.48 (ab quartet, 2H,
J =12.5 Hz), 3.27 (m, 2H), 2.5 (ab quartet, 2H, J =19 Hz), 2.4 (8
lines, 2H), 2.14 (m, 1H), 2.11 (s, 3H), 2.08 (s, 3H), 1.81 (m, 3H),
1.6 (m, 6H), 1.31 (m, 1H), 1.28 (m, 3H), 1.26 (s, 3H), 1.23 (s, 3H),
1.09 (d, J =5 Hz,1H), 0.9 (m, 4H).
2
as an amorphous white solid (89%).
1
H NMR (CDCl ): d = 7.58 (s, 1H), 5.7 (d, 1H, J = 4.6 Hz), 4.5 (dd,
3
2
H, J = 4.5, 12 Hz), 3.9 (dd, 1H, J = 3.6, 9.5), 3.21 (app t, 1H, J = 5
Hz), 3.14 (br s, 1H), 2.85 (s, 1H), 2,7 (s, 1H), 2.4 (dd, 2H, J = 18, 6
Hz), 2.1 (s, 3H), 2.0 (m, 1H), 1.6 (br s, 1H), 1.2 (s, 3H), 1.2 (m, 4H).
13
C (125 MHz, CDCl ): d = 208.7, 173.7, 173.6, 170.7, 160.8,
3
1
3
C (CDCl ): d = 208, 174.3, 171.1, 160.5, 136.2, 133.1, 132.5, 78.3,
135.5, 132.8, 132.6, 78.1, 76.5, 73.5, 69.4, 65.5, 56, 42.8, 39.3,
38.9, 36.6, 36, 34.5, 29.4, 29.2, 28.9, 25.6, 25.1, 23.8, 21.1, 20.9,
16.7, 14.4, 10.1.
3
7
2
7.3, 73.2, 69.4, 68.2, 56.5, 45.1, 39.2, 38.9, 35.1, 25.5, 23.4, 21,
0.7, 16.5, 14.9, 10.1.
IR(neat): n = 3403(m), 2926(m), 2855(w),1728(s), 1628(w),
Dimer Formation; General Procedure
1
377(m), 1260(s).
To a solution of Et N (4 equivalents) in toluene (1mL/12 mg diace-
3
HRMS (FAB) Calcd for C H O Na, 1141.5712; found,
6
2
86 18
tate) was added the diacid (0.6 equivalents) and the resulting solu-
tion was allowed to stir until it became homogenous. At that time,
1
141.5671, TLC R = 0.26 (hexane/EtOAc: 1/1)
f
2
,4,6-trichlorobenzoyl chloride (1.4 equivalents) was added drop-
C22 Dimer-20 acetate 3d
wise via syringe and the solution was allowed to stir for 1 h at r.t.
over which time it took on a light yellow color. A solution of the di-
acetate (1 equivalent) in toluene (1 mL/12 mg diacetate) was pre-
pared and to this was added DMAP (1:1 (w/w) with the diacetate)
and a few microliters of CH Cl to dissolve the diacetate. This solu-
tion was then added dropwise via syringe to the preformed anhy-
dride and the solution became highly turbid. The reaction was
monitored by TLC and was typically complete within 1 h. Upon
completion, the solution was applied directly to a column of silica
gel and the column was eluted with hexane: EtOAc (1:1) and the
product containing fractions concentrated to give the dimer (85–
From 1,22-docosanedioic acid and phorbol 13,20-diacetate, (83%)
1
H NMR (500 MHz, CDCl ): d = 7.63 (m, 1H), 5.74 (d, 1H, J = 4
3
Hz), 5.57 (br s, 1H), 5.43 (d, 1H, J = 10 Hz), 4.48 (ab quartet, 2H, J
=
12.5 Hz), 3.28 (m, 2H), 2.51 (ab quartet, 2H, J = 19 Hz), 2.34 (8
2
2
lines, 2H), 2.14 (m, 1H), 2.11 (s, 3H), 2.08 (s, 3H), 1.81 (m, 3H),
1
1
.6 (m, 14H), 1.31 (m, 1H), 1.28 (m, 3H), 1.26 (s, 3H), 1.23 (s, 3H),
.09 (d, J = 5 Hz, 1H), 0.9 (m, 4H).
1
3
C (125 MHz, CDCl ): d = 208.7, 173.7, 173.6, 170.7, 160.8,
3
1
3
2
35.5, 132.8, 132.7, 78.1, 76.4, 73.5, 69.4, 65.5, 56.1, 42.9, 39.3,
8.9, 36, 34.6, 29.7, 29.67, 29.65, 29.4, 29.3, 29, 25.6, 25.1, 23.8,
1.1, 20.9, 16.7, 14.4, 10.1.
9
5%) as either a colorless oil or a foam.
IR(neat): n = 3408(m), 2925(m), 2854(m), 1731(s), 1377(s),
260(s).
C8 Dimer-20 Acetate 3a
From suberic acid and phorbol 13,20-diacetate, (87%)
1
1
HRMS (FAB) Calcd for C H O Na, 1253.6963; found,
70 102 18
H NMR (500 MHz CDCl ): d = 7.62 (s, 1H), 5.72 (d, 1H, J =4.1
3
1
253.7000, TLC R = 0.35 (hexane/EtOAc: 1/1)
Hz), 5.56 (br s, 1H), 5.42 (d, 1H, J =10.5 Hz), 4.48 (ab quartet, 2H,
J =12.5 Hz), 3.25 (m, 2H), 2.52 (ab quartet, J =19 Hz), 2.3 (8 lines,
f
2
1
0
1
4
1
H), 2.11 (s, 3H), 2.08 (s, 3H), 1.81 (s, 3H), 1.6 (m, 2H), 1.38 (m,
PEG-Dimer-C20 acetate 3e
From the PEG diacid chloride and phorbol 13,20-diacetate, (54%)
H),1.24 (s, 3H), 1.23 (s, 3H), 1.38 (m, 2H), 1.09 (d, J =5.5 Hz,1H),
.9 (m, 4H). 13C (125 MHz CDCl ): d = 209.4, 174.4, 174.1, 171.5,
3
To a solution of phorbol 13,20 diacetate (3.0 mg. 6.7mM) in 600 µL
61.4, 136.3, 133.6, 133.3, 78.7, 77.3, 74.2, 70.1, 66.2, 56.7, 43.6,
0, 39.6, 36.8, 35.1, 29.3, 26.3, 25.6, 24.6, 21.8, 21.7, 17.4, 15.1,
0.8.
of CH Cl was added a solution of the bis-acid chloride (0.8 mg in
2
2
8
0 µL of CH Cl ) followed by a small quantity of DMAP. The so-
2 2
lution was allowed to stir under N at r.t. for 5h and then quenched
2
IR(neat): n = 3403(m), 2929(m), 1723(s), 1628(w), 1376(s),
260(s).
by the addition of sat. NaHCO . The layers were separated and the
3
1
aqueous layer extracted (3 x 5 mL) with CH Cl and the organic lay-
2
2
er concentrated. Chromatography of the resultant oil (50%
EtOAc:hexane to 70% EtOAc:hexane) to give the dimer (2.0 mg,
HRMS (FAB) Calcd for C H O Na, 1057.4772; Found,
5
6
74 18
1
057.4787, TLC R =0.17 (hexane/EtOAc: 1/1)
f
5
4%).
C12- Dimer-20 Acetate 3b
From 1,12-dodecanedioic acid and phorbol 13,20-diacetate, (85%)
1
H NMR (500 MHz, CDCl ): d = 7.58 (m, 1H), 5.685 (d, 1H, J = 5
3
Hz), 5.5 (br s, 1H), 5.48 (d, 1H, J = 10.5 Hz), 4.4–4.0 (m, 4H), 3.79
(s, 2H), 3.55 (m, 4H), 3.27 (m, 2H), 2.5 (ab quartet, 2H, J = 20 Hz),
1
H NMR (500 MHz, CDCl ): d = 7.58 (s, 1H), 5.69 (d, 1H, J =4 Hz),
3
5
.51 (br s, 1H), 5.38 (d, 1H, J =10.5 Hz), 4.43 (ab quartet, 2H, J
2
4
.2 (m, 1H), 2.14 (s, 3H), 2.07 (s, 3H), 1.79 (m, 3H), 1.6–1.5 (m,
H), 1.31–1.21 (m, 4H), 1.1 (d, J = 5 Hz,1H), 0.9 (m, 1H).
=12.5 Hz), 3.23 (m, 2H), 2.5 (ab quartet, 2H, J =19.2 Hz), 2.29 (8
lines, 2H), 2.14 (m, 1H), 2.07 (s, 3H), 2.03 (s, 3H), 1.76 (m, 3H),
1
3
C (125 MHz, CDCl ): d = 209.4, 174.4, 171.5, 171.1, 161.1,
3
1
1
.6 (m, 7H), 1.31 (m, 1H), 1.28 (m, 3H), 1.26 (s, 3H), 1.19 (s, 3H),
.04 (d, J =5.2 Hz,1H), 0.9 (m, 2H).
1
5
36.5, 133.7, 133.1, 78.7, 74.1, 71.6, 71.6, 70, 69.2, 66, 64.4, 56.7,
4.1, 43.5, 39.9,36.9, 30.4, 26.6, 24.5, 21.8, 21.7, 17.4, 15.1, 10.8.
1
3
C (125 MHz, CDCl ): d = 208.7, 173.7, 173.6, 170.7, 160.8,
3
IR(neat): n = 3403(m), 2926(m), 1722(s), 1629(m), 1458(w),
378(s), 1260(s).
1
35.5, 132.8, 132.6, 78.1, 76.5, 73.5, 69.4, 65.5, 56, 42.9, 39.3,
1
Synthesis 1999, No. SI, 1401–1406 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
Mapping Phorbol Ester Binding Domains of Protein Kinase C (PKC)
1405
HRMS (FAB) Calcd for C H O , 1082.4824; found, 1082.4723,
C22 Dimer 4d
5
6
74 21
TLC R = 0.25 (hexane/EtOAc: 1/1)
(79%)
f
1
C20 Deprotection; General Procedure
H NMR (500 MHz, CDCl ): d = 7.6 (s, 1H), 5.71 (d, 1H, J =4 Hz),
3
5
.59 (br s, 1H), 5.43 (d, 1H, J =10 Hz), 4.03 (ab quartet, 2H, J =13
To a solution of the dimer in MeOH (3.2 mL / 10 mg of dimer) was
added a solution of perchloric acid in MeOH (prepared by adding
Hz), 3.27 (m, 2H), 2.55 (ab quartet, 2H, J =18.5 Hz), 2.34 (6 lines,
2
H), 2.17 (m, 1H), 2.11 (s, 3H), 1.81 (m, 3H), 1.6 (m, 14H), 1.31
m, 1H), 1.28 (m, 3H), 1.26 (s, 3H), 1.23 (s, 3H),.91 (d, J =5
Hz,1H), 0.9 (m, 4H).
2
8 µL of concentrated perchloric acid to 10 mL of MeOH) in one
(
portion (3.2 mL/ 10 mg dimer) and the flask was wrapped in alumi-
num foil and was allowed to stir under a positive pressure of N for
2
1
3
4
8 h. Upon completion, a small quantity of solid NaHCO was add-
C (125 MHz, CDCl ): d = 209.7, 174.5, 174.4, 161.5, 141.2,
3
3
ed and allowed to stir for 1 h. The solution was then filtered through
a 1 inch pad of silica gel and the filter cake was washed repeatedly
133.6, 129.8, 78.9, 77.2, 74.4, 68.7, 66.3, 56.8, 43.5, 39.7, 39.4,
39.2, 36.9, 35.3, 30.36, 30.31, 30.3, 29.9, 29.7, 26.3, 25.9, 24.5,
21.8, 20.9, 17.5, 15.1, 10.8.
(
5–6 times) with EtOAc. The solution was concentrated to give a
white residue which was purified by silica gel chromatography
EtOAc as eluent) to give the dimer (65–75%) as a colorless oil or
IR(neat): n = 3404(m), 2925(m), 1716(s), 1628(w), 1377(s),
1
(
262(s).
as a foam.
HRMS (FAB) Calcd for C H O Na, 1169.6752; found,
1
6
6
98 16
C8 Dimer 4a
169.6703 TLC R = 0.63 (100% EtOAc).
f
(
85%)
PEG-Dimer 4e
(
1
H NMR (500 MHz, CDCl ): d = 7.61 (s, 1H), 5.7 (d, 1H, J = 5 Hz),
3
55%)
5
1
.57 (br s, 1H), 5.41 (d, 1H, J = 10 Hz), 4.04 (ab quartet, 2H, J =
2.5 Hz), 3.26 (m, 2H), 2.54 (ab quartet, 2H, J = 19 Hz), 2.35 (8
1
H NMR (500 MHz, CDCl ): d = 7.58 (m, 1H), 5.685 (d, 1H, J = 4.5
3
Hz), 5.57 (br s, 1H), 5.48 (d, 1H, J =10.5 Hz), 4.4–4.0 (m, 4H), 3.79
s, 2H), 3.55 (m, 4H), 3.27 (m, 2H), 2.5 (ab quartet, 2H, J = 20 Hz),
.2 (m, 1H), 2.14 (s, 3H),1.79 (m, 3H), 1.6–1.5 (m, 4H), 1.31–1.21
m, 4H), 1.1 (d, J = 5 Hz, 1H), 0.9 (m, 1H).
lines, 2H), 2.11 (s, 3H), 1.81 (d, J = 1.5 Hz, 3H), 1.6 (m, 2H), 1.38
m, 1H),1.27 (s, 3H), 1.23 (s, 3H), 1.22 (s, 3H), 1.09 (d, J = 5
Hz,1H), 0.9 (m, 4H).
(
2
(
1
(
1
3
C (125 MHz, CDCl ): d = 208.9, 173.7, 173.4, 160.7, 140.5,
3
3
C (125 MHz, CDCl ): d = 209.4, 174.6, 171.4, 161, 141.4, 131.6,
1
2
32.8, 129, 78.1, 73.7, 67.9, 65.6, 56, 42.9, 38.9, 39.5, 36.2, 34.5,
9.7, 28.6, 25.6, 24.8, 23.8, 21.1, 16.8, 14.4, 10.1.
3
1
3
29.5, 78.8, 77.1, 75.7, 75.6, 74.3, 69.1, 68.6, 56.8, 56.7, 43.6,
9.7,39.2, 37.1, 23.6, 21.8, 17.4, 15.1, 10.8.
IR(neat): n = 3396(m), 2925(m), 1716(s), 1628(w), 1377(s),
261(s).
IR(neat): n = 3412(m), 2932(m), 1718(s), 1629(m), 1438(w),
377(s), 1261(s).
1
1
HRMS (FAB) Calcd for C H O Na, 973.4561; found, 973.4601,
5
2
70 16
HRMS (FAB) Calcd for C H O , 1021.4409; found, 1021.4409,
TLC R = 0.28 (100% EtOAc).
TLC R = 0.26 (100% EtOAc).
52 70 19
f
f
C12 Dimer 4b
(
59%)
PKC Binding Assay Protocol
A. Preparation
Filters (30–35) were prepared by soaking for 30 min in a solution
obtained by dissolving 3 mL of 10% polyethyleneimine in 100 mL
of distilled water and swirling periodically.
1
H NMR (500 MHz, CDCl ): d = 7.56 (s, 1H), 5.651 (d, 1H, J =4.8
3
Hz), 5.6 (br s, 1H), 5.52 (d, 1H, J =10.5 Hz), 4.0 (ab quartet, 2H, J
=
13.1 Hz), 3.22 (m, 2H), 2.49 (ab quartet, 2H, J =18.5 Hz), 2.43 (br
s, 1H), 2.31 (m, 2H), 2.13 (m, 1H), 2.07 (s, 3H), 1.74 (m, 3H), 1.53
(
(
1
m, 6H), 1.31 (m, 1H), 1.27 (m, 3H), 1.26 (s, 3H), 1.23 (s, 3H), 1.06
d, J =5 Hz,1H), 0.9 (m, 4H).
Wash buffer was prepared by diluting 20 mL of 1 M Tris pH=7.4 to
a volume of 500 mL and then cooling the resultant solution on ice
for at least 30 min.
3
C (125 MHz, CDCl ): d = 209, 173.8, 173.7, 160.7, 140.5, 132.8,
3
1
2
29.1, 78.2, 76.4, 73.7, 68, 65.5, 56, 42.8, 38.9, 38.5, 36.2, 34.6,
9.3, 29.1, 28.9, 25.6, 25.1, 23.8, 21.1, 16.8, 14.4, 10.1.
Phosphatidyl serine was prepared by concentrating 400 µL of 10
(
mg/ml) PS (Avanti Polar Lipids) in CHCl with a stream of N ,
3
2
IR(neat): n = 3396(m), 2925(m), 1716(s), 1628(w), 1377(s),
261(s).
then suspending it in 4 mL of Tris buffer followed by sonication at
0 °C (4 x 30 sec; 30 sec on then 30 sec off; power = 6; 40% duty
cycle).
1
HRMS (FAB) Calcd for C56H78O16Na1029.5187; found,
029.5227, TLC Rf=0.41 (100% EtOAc).
1
PKC (rat brain, purified by DEAE-cellulose chromatography) is
prepared by dilution of the stock solution by adding 50 µL of stock
to 10 mL of PKC buffer. This buffer was prepared from 1 mL of 1
M Tris 7.4, 2 mL of 1 M KCl, 30 µL of 0.1 M CaCl , 40 mg of BSA,
and dilution to 20 mL total volume.
C14 Dimer 4c
(
78%)
1
2
H NMR (500 MHz, CDCl ): d = 7.6 (s, 1H), 5.71 (d, 1H, J = 5 Hz),
3
5
1
2
1
1
.6 (br s, 1H), 5.42 (d, 1H, J = 10.5 Hz), 4.05 (ab quartet, 2H, J =
3 Hz), 3.26 (m, 2H), 2.5 (ab quartet, 2H, J = 19 Hz), 2.64 (br s, 1H),
.35 (m, 2H), 2.17 (m, 1H), 2.11 (s, 3H), 1.78 (d, 3H, J = 1.5 Hz),
.6 (m, 6H), 1.31 (m, 1H), 1.28 (m, 3H), 1.26 (s, 3H), 1.23 (s, 3H),
.1 (d, J = 5 Hz 1H), 0.9 (m, 4H).
The compound to be assayed was diluted to the desired concentra-
tions in absolute EtOH.
B. PKC Binding Assay
Each data point is obtained in triplicate and, in addition, a zero
point, where no competitor was added, and a non-specific point,
where all specific binding was eliminated through addition of ex-
cess phorbol myristate acetate, are also acquired.
1
3
C (125 MHz, CDCl ): d = 209, 173.8, 173.7, 160.8, 140.6, 132.8,
3
1
2
29.1, 78.2, 76.5, 73.6, 68, 65.6, 56, 42.8, 38.9, 38.4, 36.1, 34.6,
9.6, 29.5, 29.2, 28.9, 25.6, 25.1, 23.8, 21.1, 16.7, 14.4, 10.1
6
0 µL of PS vesicles, 200 µL of diluted PKC, and 20 µL total vol-
ume of the compound to be assayed are added to an Eppendorf tube
1.5 mL). The tubes used to measure non-specific binding receive 5
IR(neat): n = 3405(m), 2925(m), 1717(s), 1457(w), 1376(s),
261(s).
1
(
HRMS (FAB) Calcd for C H O Na, 1057.5500; found,
5
8
82 16
µL of a 1 mM PMA solution (in 5% DMSO/H O). To the solution
was added 20 µL of tritiated PDBu (30 nM, final concentration = 2
2
1
057.5406, TLC R = 0.42 (100% EtOAc).
f
Synthesis 1999, No. SI, 1401–1406 ISSN 0039-7881 © Thieme Stuttgart · New York

1
406
P. A. Wender et al.
PAPER
nM, DuPont NEN) and each tube was momentarily vortexed to ob-
tain good mixing. The samples are then incubated for 30 min at
dry before being placed in a scintillation vial without scintillation
fluid and counted.
3
7° C and the reaction was stopped by cooling on ice for at least 5
min. The contents of each tube was then filtered by applying the
mixture uniformly to the filter disk that is placed on a Hirsch funnel.
The tube was rinsed with 500 µL of ice-cold 20 mM Tris•HCl, pH
Acknowledgment
We thank Professor Daria Mochly-Rosen for providing laboratory
facilities for and invaluable assistance in performing our binding
and kinase assays. Support of this work through a grant (CA31841)
provided by the National Institutes of Health and a grant-in-aid for
scientific research from the Ministry of Education, Science, Culture
and Sports, Japan (K.I.) is also gratefully acknowledged. HRMS
analyses were performed at the UC Riverside Mass Spectrometry
Facility. Fellowship support from the Eli Lilly Corporation
7
.4 buffer and the wash was applied to the filter disk. The disk was
then washed by slowly dripping 5 mL of the ice-cold buffer solution
onto the filter disk. The disk was then placed in a scintillation vial
and 3 mL of CytoScint fluid was added. The filteres were soaked for
at least 4 h and were then quantitated in a scintillation counter. The
data were then processed by averaging the three points, plotted and
the theoretical curve fitted to the data using the standard non-linear
regression algorithm in the KaleidaGraph v 3.0 package running on
a Macintosh system. The following equation is used to fit the curve:
(M.F.T.K.) is gratefully recognized.
References
IC(50) = B = B +(B B ) / 1+10(
log(dimer))–log(IC50)
NS
max– NS
(
(
(
1) Nishizuka, Y. Science 1986, 233, 305.
2) Dekker, L. V.; Parker, P. J. Trends Biochem. Sci. 1994, 19, 73.
3) Wender, P. A.; Irie, K.; Miller, B.L. Proc. Natl. Acad. Sci.
USA 1995, 92, 239.
B = avg CPM at concentration (dimer)
BNS = CPM at saturating concentration of dimer (variable)
Bmax = CPM when [dimer]=0
(
4) Irie, K.; Oie, K.; Nakahara, A.; Yanai, Y.; Ohigashi, H.;
Wender, P. A.; Fukuda, H.; Konishi, H.; Kikkawa, U. J. Am.
Chem. Soc. 1998, 120, 9159.
[dimer] = dimer concentration
(
5) Quest, A. F.; Bell, R.M. J. Biol. Chem. 1991, 266, 18330.
Irie, K.; Yanai, Y.; Oie, K.; Ishizawa, J.; Nakagawa, Y.;
Ohigashi, H.; Wender, P. A.; Kikkawa, U. Bioorg. & Med.
Chem. 1997, 5, 1725.
IC50 = concentration of dimer at which 50% inhibition is observed
The K value was obtained form the IC50 as follows:
i
(
(
6) Malmberg, A. B.; Chen, C.; Tonegawa, S.; Basbaum, A.I.
Science 1997, 278, 279.
7) During the course of our investigations, the following
example of these homodimeric species appeared in the
literature:
K = IC / 1+{[PDBu]/K (PDBu)}
i
50
d
Kinase Assay Protocol
The following stock solutions were required for the kinase assay:
1
) Lipid was prepared by placing 12 µL phosphatidyl serine in
Sodeoka, M.; Arai, M. A.; Adachi, K.; Uotsu, K.; Shibasaki,
M. J. Am. Chem. Soc. 1998, 120, 457.
CHCl (Avanti Polar Lipids, 10 mg/ml), and 24 µL phosphatidyl
3
choline in CHCl (Avanti Polar Lipids, 20 mg/ml) in a glass test
(8) Wender, P. A.; Koehler, K.F.; Sharkey, N. A.; Dell’Aquila,
M.L.; Blumber, P. M. Proc. Natl. Acad. Sci. USA 1986, 83,
4214.
Wender, P. A.; Cribbs, C. M. Adv. Med. Chem. 1992, 1, 1.
(9) Hecker, E.; Schmidt, R. Fortschr. Chem. Org. Naturst. 1974,
31, 377.
3
tube. A stream of dry N was directed over the solutions until the or-
2
ganic solvents were removed. The remaining lipids were then sus-
pended in 1 ml of 20mM Tris, pH 7.4 by vigorously pipetting the
solution. This suspension was then sonicated at 0 °C (4 x 30 sec; 30
sec on then 30 sec off; power = 6; 40% duty cycle).
) Substrate solution was prepared by mixing MBP4–14 (Sigma), 750
µL; 10 mM CaCl , 150 µL; and 100 µL of 20 mM Tris, pH 7.4.
(10) Yamaguchi, M. J. Synth. Org. Chem. Jpn. 1980, 38, 22.
2
(
11) Bollag, G. E.; Roth, R. A.; Beaudoin, J.; Mochly-Rosen, D.;
2
Koshland, D.E. Proc. Natl. Acad. Sci. USA 1986, 83, 5822.
3
3
) The ATP solution is prepared with 75 µL of 1 M MgCl , 25 µL
2
.75 mM ATP, 1.5 ml 20mM Tris, pH 7.4, and 1.5 µL g P ATP.
(12) Wender, P. A.; DeBranbander, J.; Harran, P. G.; Jimenez, J.-
M.; Koehler. M.F.T.; Lippa, B.; Park, C.-M.; Siedenbiedel,
C.; Pettit, G.R. Proc. Natl. Acad. Sci. USA 1998, 95, 6624.
(13) Goldstein, A.; Barrett, R. W. Mol. Pharm. 1987, 31, 603.
(14) Sharkey, N. A.; Blumberg, P.M. Carcinogenesis 1986, 7, 677.
Sharkey, N. A.; Blumberg, P.M. Cancer Res. 1985, 45, 19.
(15) Epand, R. M. Analytical Biochemistry 1994, 218, 241.
3
2
The assay was performed in a Nunc V96 polypropylene 96 well
plate designed to prevent protein adhesion to its surface. Into each
well is placed 15 µL of the lipid vesicles, 5 µL of substrate solution,
2
0 µL of test compound, PKC solution (10 µL, 1:3 dilution form rat
brain prep), and ATP solution (25 µL). The reactions were incubat-
ed at r.t. for 30 min and then stopped by the addition of 100 µL of
1
00 mM H PO . 150 µL of each stopped reaction is removed and
3 4
2
blotted on 1 cm Whatman P81 cation exchange paper. After air
drying, the filters were washed three times with deionized water and
once with absolute EtOH. The papers were then again allowed to air
Article Identifier:
1437-210X,E;1999,0,SI,1401,1406,ftx,en;xC02299SS.pdf
Synthesis 1999, No. SI, 1401–1406 ISSN 0039-7881 © Thieme Stuttgart · New York