Conformationally constrained analogues of diacylglycerol (DAG). 28. DAG-dioxolanones reveal a new additional interaction site in the C1b domain of PKCδ

J. Med. Chem. 2007, 50, 3465-3481

3465

Conformationally Constrained Analogues of Diacylglycerol (DAG). 28. DAG-dioxolanones

Reveal a New Additional Interaction Site in the C1b Domain of PKCδ

Yongseok Choi,^‡,#,†Yongmei Pu,^§,#Megan L. Peach,^|,#Ji-Hye Kang,^‡Nancy E. Lewin,^§Dina M. Sigano,^‡Susan H. Garfield,

Peter M. Blumberg,*^,§and Victor E. Marquez*^,‡

Laboratory of Medicinal Chemistry, Center for Cancer Research, National Cancer InstitutesFrederick, National Institutes of Health,

Frederick, Maryland 21702, Laboratory of Cancer Biology and Genetics and Laboratory of Experimental Carcinogenesis, Center for Cancer

Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, and Basic Research Program, SAIC-Frederick,

Inc., NCIsFrederick, Frederick, Maryland 21702

ReceiVed March 7, 2007

Diacylglycerol (DAG) lactones have provided a powerful platform for structural exploration of the interactions

between ligands and the C1 domains of protein kinase C (PKC). In this study, we report that

DAG-dioxolanones, novel derivatives of DAG-lactones, exploit an additional point of contact (glutamine

27) in their binding with the C1b domain of PKCδ. Mutation of this point of contact to glutamate selectively

impairs binding of the DAG-dioxolanones compared to that of the corresponding DAG-lactones (1200- to

3000-fold versus 35- to 55-fold, respectively). The differential response of this mutated C1b domain to the

DAG-dioxolanones relative to the DAG-lactones provides a unique tool to probe the role of the C1b domain

in PKCδ function, where the response to the DAG-lactones affords a positive control for retained function.

Using this approach, we show that the C1b domain of PKCδ plays the predominant role in the translocation

of PKCδ to the membrane in the presence of DAG.

Introduction

one of multiple families of proteins that have appropriate C1

domains to act as DAG receptors.⁸The other DAG responsive

families include the PKD family,⁹a distinct class of serine/

threonine kinases; the chimaerin family,¹⁰inhibitors of p21Rac;

the munc-13 family,¹¹proteins involved in synaptic vesicle

priming; the RasGRP family,^12,13guanyl nucleotide exchange

factors for Ras and Rap1; and the DAG kinase family,^14,15which

functions to abrogate DAG signaling, thus providing negative

feedback on DAG signaling pathways. Of these C1 domain-

containing families, the PKC family is the best-studied mediator

in the DAG signaling pathway and an important target for drug

development.

The lipophilic second messenger diacylglycerol (DAG^a),

produced through hydrolysis of phosphatidylinositol 4,5-bis-

phosphate (PIP₂) by the activation of receptor-coupled phos-

pholipase C (PLC) or indirectly from phosphatidylcholine via

phospholipase D,¹plays a central role in signal transduction of

numerous physiological and pathological events including

proliferation, differentiation, apoptosis, angiogenesis, and drug

resistance.^2-6These diverse functions of the DAG signal are

realized through binding with high affinity to a class of zinc

finger structures, the so-called “C1 domains”. C1 domains were

first identified in PKC as the interaction site for DAG and

phorbol esters.⁷It is now recognized, however, that PKC is only

The PKC family of serine/threonine kinases consists of 10

members that are divided into three classes: (i) the classical

PKCs (cPKCs R, âI, âII, γ); (ii) the novel PKCs (nPKCs δ, ꢀ,

η, θ); the atypical PKCs (aPKCs ú, ι/λ). Both the classic and

novel PKCs contain two tandem C1 domains (C1a and C1b) in

* To whom correspondence should be addressed. For P.M.B.: phone,

301-496-3189; fax, 301-496-8709; e-mail, blumberp@dc37a.nci.nih.gov.

For V.E.M.: phone, 301-846-5954; fax, 301-846-6033; e-mail, marquezv@

mail.nih.gov.

their N-terminal regulatory region that bind to DAG in a Ca²⁺

-

^‡National Cancer InstitutesFrederick.

dependent (cPKCs) or -independent manner (nPKCs). Besides

C1 domains, the regulatory region of the cPKCs and nPKCs

also contains a pseudosubstrate domain, which can interact with

the catalytic site of the C-terminal kinase domain to inhibit the

kinase activity. Binding of DAG to the C1 domain completes a

hydrophobic surface on the C1 domain, favoring its interaction

with the membrane and causing the membrane translocation of

the whole PKC protein.¹⁶This DAG-C1 domain-membrane

interaction is coupled to a conformational change in PKC,

causing the release of the pseudosubstrate domain from the

catalytic site and thereby activating the enzyme.⁸

^†Current address: Laboratory of Molecular Therapeutics, School of

Biosciences and Biotechnology, Korea University, Anam-dong Seongbuk-

gu, Seoul 136-701, Korea. E-mail: ychoi@korea.ac.kr.

^#These authors contributed equally to this work.

^§Laboratory of Cancer Biology and Genetics, National Cancer Institute.

^|SAIC-Frederick, Inc.

Laboratory of Experimental Carcinogenesis, National Cancer Institute.

^aAbbreviations: CAN, ceric ammonium nitrate; CHO, Chinese hamster

ovary; DAG, diacylglycerol; DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene;

DMAP, dimethylaminopyridine; DMS, dimethyl sulfide; DNA, deoxy-

ribonucleic acid; FABMS, fast atom bombardment mass spectra; GFP, green

fluorescent protein; GST, glutathione S-transferase; HAART, highly active

antiretroviral therapy; HIV/AIDS, human immunodeficiency virus/acquired

immune deficiency syndrome; IL-6, interleukin 6; IPTG, isopropyl-O-^D-

thiogalactopyranoside; LB, Luria broth; LDA, lithium diisopropylamide;

LiHMDS, lithium hexamethyldisilylazide; MsCl, methanesulfonyl chloride;

NMR, nuclear magnetic resonance; OD, optical density; PCC, pyridinium

chlorochromate; PCR, Polymerase chain reaction; PDBu, phorbol 12,13-

dibutyrate; PIP₂, phosphatidylinositol 4,5-bisphosphate; PKC, protein kinase

C; PKD, protein kinase D; PLC, phospholipase C; PMA, phorbol 12-

myristate-13-acetate; PMP, p-methoxyphenyl; pTsOH, p-toluenesulfonic

acid; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electro-

phoresis; SEM, standard error of the mean; THF, tetrahydrofuran; TLC,

thin layer chromatography.

Multiple classes of high-affinity ligands for the C1 domains

of PKCs have been described. These ligands include the

diterpenes such as the phorbol esters, macrocyclic lactones such

as the bryostatins, polyacetates such as aplysiatoxin, or indole

alkaloids such as teleocidin.¹⁷Unfortunately, with the exception

of the indole alkaloids, the complicated structures of these

ligands have largely prevented their further development through

medicinal chemistry. Taking advantage of the DAG derivatives,

Published on Web 06/26/2007

3466 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

Scheme 1^a

Table 1. Inhibitory Dissociation Constants (K_i) of DAG-lactones and

Corresponding DAG-dioxolanones for PKCR^a

^a(a) NaH, BnOH, ∆; (b) PCC, 4 Å molecular sieves, CH₂Cl₂; (c)

allylmagnesium bromide, THF, -78 °C; (d) BH₃‚SMe₂, THF, -78 °C; (e)

PCC, CH₂Cl₂; (f) LiHMDS, RCHO, THF, -78 °C; (g) (i) Et₃N, MsCl,

CH₂Cl₂, 0 °C f room temp; (ii) DBU; (h) CAN, CH₃CN, H₂O, 0 °C; (i)

Et₃N, DMAP, R′C(O)Cl, CH₂Cl₂, 0 °C f room temp; (j) BCl₃, CH₂Cl₂,

-78 °C.

DAG-lactone

(log P ) 4.27)

DAG-dioxolanones

(log P ) 3.82)

K_i(nM)

Z-1²⁴

Z-2

Z-3

2.90 ( 0.35

3.49 ( 0.57

3.07 ( 0.40

E,Z-4^b

6.41 ( 0.76

19.83 ( 0.52

5.78 ( 0.95

in which the flexibility of the structure has been constrained to

reduce the entropic loss due to binding,¹⁷our group has

developed a series of DAG-lactones with K_ivalues for binding

to PKC approaching those of phorbol esters.¹⁷These structures

have provided promising lead compounds for selective actions

mediated by subsets of the DAG-responsive proteins.^18,19They

likewise have proven to be powerful tools for probing the

detailed molecular interactions underlying ligand binding to the

C1 domains.¹⁷

In this study, we describe the characterization of dioxolanone

analogues of the DAG-lactones. These compounds were pre-

dicted by molecular modeling to incorporate an additional site

of interaction within the binding cleft of the C1 domain.

Consistent with this prediction, we show that these compounds

are orders of magnitude more sensitive to disruption of this

predicted site of interaction through mutation of the C1 domain.

5^c

E,Z-6^b

^aThe K_ivalues were measured by competitive binding of the compounds

for the receptor against [³H]PDBu. Each value is an average of three

independent experiments, expressed as the mean ( SEM. The partition

coefficients (log P) were calculated according to the atom-based program

MOE SLog P.²³^bMixture of E and Z isomers. ^cSingle geometric isomer

(E or Z).

Scheme 2^a

Chemistry

As illustrated in Scheme 1, target DAG-lactones (1-3, Table

1) were synthesized in racemic form starting from commercially

available glycidyl-4-methoxyphenyl ether, according to meth-

odology previously developed in our laboratory.^20-22As before,

the DAG-lactones were successfully separated into both E- and

Z-isomers. The vinyl proton of the Z-isomers displayed a

characteristic multiplet at δ 6.20-6.24, while the corresponding

signal for the E-isomer appeared more downfield at δ 6.72-

6.78. The final target DAG-lactones appear in Table 1.

Our strategy for assembling the functionalized DAG-diox-

olanone targets focused on constructing a central 1,3-dioxolan-

4-one skeleton (XII) that could be elaborated in a manner similar

to the DAG-lactones (Scheme 2). Accordingly, reaction of

common intermediate II with commercially available 3-chloro-

1,2-propanediol under acidic conditions gave X. Base-catalyzed

elimination, followed by ozonolysis of the resultant double bond

gave XII, the dioxolanone equivalent to DAG-lactone IV. Aldol-

type condensation of XII with either 3-isobutyl-5-methylhexanal

or isovaleryl aldehyde gave XIII, which was not purified but

immediately subjected to elimination conditions to give XIV.

Deprotection of the PMP group followed by acylation led to

XVI, and upon debenzylation, the desired racemic 1,3-dioxolan-

4-one analogues (XVII, 4-6) were obtained. In the case of the

DAG-dioxolanones, all the final compounds were obtained as

inseparable mixtures of E- and Z-isomers except for 5, which

was obtained as a single isomer whose geometric stereochem-

istry could not be determined. The final target DAG-dioxol-

anones appear in Table 1.

^a(a) 3-Chloro-1,2-propanediol, pTsOH‚H₂O, C₆H₆, ∆; (b) KO^tBu, THF,

∆; (c) (i) O₃, MeOH/CH₂Cl₂, -78 °C; (ii) DMS, room temp; (d) (i)

LiHMDS, THF, -78 °C; (ii) RCHO; (e) (i) Et₃N, MsCl, CH₂Cl₂, 0 °C f

room temp; (ii) DBU; (f) CAN, CH₃CN, H₂O; (g) Et₃N, DMAP, R′C(O)Cl,

CH₂Cl₂, 0 °C f room temp; (h) BCl₃, CH₂Cl₂, -78 °C.

Results

Binding Potencies of the DAG-dioxolanones Were Main-

tained at the Level of Their Corresponding DAG-lactones.

The potencies of the three DAG-dioxolanones prepared were

first evaluated for binding to PKCR, which has been our routine

assay for the DAG-lactones. The binding affinities of the parent

DAG-lactones were measured for comparison, and for simplicity

only the data for the Z-isomers are included (Table 1). The

differences in potencies between Z- and E-isomers for DAG-

lactones are small and vary from 1.1- to 1.8-fold, generally in

favor of the Z-isomer. All the DAG-dioxolanones were potent

ligands, with K_ivalues 2- to 6-fold weaker than those of their

corresponding DAG-lactones. These results indicate that the

introduction of the oxygen in the ring system to form the

dioxolanone ring was well tolerated. However, the anticipated

formation of an additional hydrogen bond did not lead to an

overall enhancement in binding, implying that the additional

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3467

Figure 1. Docking prediction of interactions between C1b domain of PKCδ and DAG-dioxolanones 4-6 (Z-isomers). The amino acid sequence

of the C1b domain of PKCδ is shown in the upper panel. The residues that form hydrogen bonds with the DAG-dioxolanones are marked in red.

The added oxygen in the ring structure forms a new hydrogen bond with the glutamine residue (GLN 257), as shown in the yellow circles. The

indicated glutamine residue is underlined both in the upper panel and in the lower pictures. The DAG-dioxolanone structure is shown in green.

binding energy of this hydrogen bond was counterbalanced by

a poorer overall fit into the C1 domain (vide infra).

Table 2. Inhibitory Dissociation Constants (K_i) of the Dioxolanones

(4-6) and Their Corresponding DAG Lactones (Z-1, Z-2, Z-3) for the

Wild-Type C1b Domain of PKCδ (wt δC1b) and the Glutamate Mutant

(Q27E)^a

Prediction from Computer Modeling of an Additional Site

of Interaction in the PKCδ C1b Domain upon Conversion

of the DAG-lactone to the Corresponding DAG-dioxolanone.

DAG-lactones have provided a powerful platform for probing

the interactions of ligands with the C1 domains. Our modeling

studies of these binary complexes have identified two alternative

binding modes, designated as sn-1 or sn-2, depending on the

identity of the carbonyl engaged in binding to the protein. The

sn-1 binding mode is defined as that in which the sn-1 carbonyl

is hydrogen-bonded to the C1 domain, and for the alternative

sn-2 binding mode, it is the sn-2 carbonyl that appears directly

engaged in hydrogen bonding to the protein. Although from

the modeling perspective both binding modes appear to form

an identical network of hydrogen bonds as was observed with

phorbol 13-O-acetate, we have determined that DAG-lactones

bind preferentially in the sn-2 binding mode.^25,26This binding

mode is additionally enhanced by the presence of highly

branched R-alkylidene side chains, which lie adjacent to the

sn-2 carbonyl. In such binding mode the orphan sn-1 carbonyl

binds at the membrane interface.^25,26

K

_wt/δC1b(nM)

K_Q27E/δC1b(nM)

K_Q27E/δC1b/K_wt/δC1b

4^b

0.82 ( 0.07

1.15 ( 0.07

2.82 ( 0.36

0.96 ( 0.09

0.90 ( 0.07

2.13 ( 0.15

0.33 ( 0.05

2510 ( 220

44.8 ( 4.5

3460 ( 210

53.0 ( 3.7

1720 ( 160

73.72 ( 5.3

6.2 ( 1.0

3070

39

1230

55

1910

35

19

Z-1^c

5^b

Z-2^c

6^b

Z-3^c

PDBu

^aThe K_ivalues were measured by competitive binding of the compounds

for the receptors against [³H]PDBu. Each value is an average of three

independent experiments, expressed as the mean ( SEM. Ratios of K values

(third column) of Mutant to K values of wild type (K_ifor DAG-lactones

and DAG-dioxolanones; K_dfor PDBu) indicate the change in potency.

^bDAG-dioxolanones. ^cDAG-lactones.

an overall enhancement in binding because it was counterbal-

anced by the poorer sn-1 binding mode involving the confor-

mationally flexible sn-1 carbonyl rather than the conformation-

ally rigid lactone carbonyl, thus voiding the entropic advantage

of the sn-2 binding mode.

We performed computational docking studies to predict the

binding mode and the interactions formed between the PKCδ

C1b domain and the new DAG-dioxolanone compounds. The

computer modeling demonstrated that the added oxygen could

form an additional hydrogen bond in the binding cleft of the

δC1b domain with the side chain of the glutamine residue at

position 257 (numbered as position 27 in the isolated δC1b

domain) only in the less favorable sn-1 binding mode. This

glutamine residue, although important for the structural integrity

of the C1 domain,²⁷is not predicted to interact with the DAG-

lactones directly. In the sn-1 binding mode, four amino acids

are predicted to form hydrogen bonds with the DAG-dioxol-

anones (marked in red in the upper panel of Figure 1). Thr242,

Leu251, and Gly253 form three hydrogen bonds with the

hydroxyl group and the sn-1 carbonyl group respectively on

the side chain of the DAG-lactone backbone. The added oxygen

in the ring system (marked with a yellow dashed circle in each

image in Figure 1) forms the fourth hydrogen bond with the

Gln257 residue (underlined in the figure). As discussed in the

previous section, this additional hydrogen bond did not lead to

The Glutamine Residue of the Isolated δC1b Plays an

Important Role in Binding by the DAG-dioxolanones but

Not by the Corresponding DAG-lactones. To verify the role

of Gln257 in the interaction with the DAG-dioxolanones, we

used site-directed mutagenesis to convert the glutamine (Gln)

to glutamate (Glu). For ease of analysis, we performed the site

directed mutagenesis on the isolated C1b domain of PKCδ

(where the glutamine is in position 27 relative to the start of

the C1b domain, with the mutant thus designated Q27E). The

C1b mutant (Q27E/δC1b) was expressed and purified as a GST-

fusion protein from E. coli. Scatchard analysis was performed

to determine the dissociation constant (K_d) of the Q27E/δC1b

mutant for binding to [³H]PDBu. The K_dvalue we obtained

was 6.2 ( 1.0 nM (three experiments, mean ( SEM). Compared

with the K_dof the wild-type δC1b for PDBu (0.33 ( 0.05 nM,

three experiments, mean ( SEM), the binding potency of the

Q27E/δC1b mutant was decreased about 20-fold (Table 2). Such

an effect on binding of [³H]PDBu was not unexpected, given

the important role of this residue in the overall folding of the

C1 domain.²⁷Although this residue has been predicted not to

3468 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

Figure 2. Dose response curves of the DAG-dioxolanones (A-C) and their corresponding DAG lactones (D-F) for binding to isolated wild-type

C1b domain (wt/δC1b) and the C1b mutant (Q27E/δC1b) of PKCδ. The curves were obtained from the competitive binding of these compounds

in the presence of [³H]PDBu. Each curve is from a single experiment. All experiments were performed at least two additional times, with similar

results: (A, D) dose response curves for the DAG-dioxolanone 4 and its corresponding DAG-lactone Z-1; (B, E) dose response curves for DAG-

dioxolanone 5 and its corresponding DAG-lactone Z-2; (C, F) dose response curves for DAG-dioxolanone 6 and its corresponding DAG-lactone

Z-3.

interact with phorbol esters directly, it forms bridging hydrogen

bonds at the point where the large â sheet of C1b comes apart

at one end of the binding groove.¹⁶However, by the replacement

with a very similar amino acid, glutamate, the folding could

evidently be maintained for binding to PDBu with an affinity

in the 10^-9M range. For comparison, all binding activity for

PDBu was lost upon mutation of Gln27 to Gly or Trp, and

binding affinities were reduced to 22 or 105 nM upon mutation

to Val or Thr, respectively.²⁷

absence the DAG-dioxolanones would bind correspondingly

more weakly in the sn-1 binding mode.

Does the Glutamine Residue (Q27) in the Isolated δC1a

Domain Play a Similar Role in the DAG-dioxolanone

Interaction? Glutamine residue Q27 is highly conserved among

different C1 domains including the C1a and C1b domains of

PKCδ (see Figure 4A), but it is becoming increasingly clear

that different C1 domains show substantial differences in their

interactions with ligands.^19,28Does the dioxolanone ring also

form a new interaction with the δC1a domain through the

glutamine residue at the indicated position? To answer this

question, we mutated the glutamine residue at position 27 of

the isolated δC1a domain into glutamate as had been done for

the δC1b domain. Then we examined the binding affinity of

the C1a mutant (Q27E/δC1a) for the DAG-dioxolanone 4. The

result was clearly different from what we observed with the

Q27E/δC1b mutant. First, the binding affinity of the Q27E/

δC1a mutant for PDBu was greatly reduced. The wild-type δC1a

potently bound PDBu with a K_dof 2.04 ( 0.24 nM (Table 3).

However, after mutation, the K_dincreased to 383 ( 15 nM for

a 200-fold decrease of the binding potency. Compared to the

C1b mutant (Q27E/δC1b) (K_d) 6.2 ( 1.0 nM, Table 2), the

δC1a mutation (Q27E/δC1a) had a much greater effect on the

phorbol ester binding, suggesting that the glutamine residue

(Q27) plays a much more critical role in the protein folding of

δC1a. Second, the binding affinities of the Q27E/δC1a mutant

for the DAG-dioxolanone and the corresponding DAG lactone

were decreased to a similar degree. As shown in Table 3, both

the DAG-dioxolanone 4 and the DAG-lactone Z-1 bound the

wild-type δC1a with high potency (K_ivalues of 4.31 ( 0.13

and 1.21 ( 0.11 nM, respectively). However, after mutation,

the K_ivalues increased to 1567 ( 58 and 476 ( 6 nM,

respectively. There were 390- and 360-fold decreases of the

The K_ivalues of the DAG-dioxolanones 4-6 for the Q27E/

δC1b mutant were measured and compared with those for the

wild-type δC1b domain. Dramatic differences were observed

between the wild-type δC1b and the Q27E/δC1b mutant in the

binding to the DAG-dioxolanones. All the DAG-dioxolanones

bound to the Q27E/δC1b mutant with a much weaker binding

affinity compared with the wild type (Figure 2A-C). For

example, 4 bound to the wild-type C1b with a K_iof 0.82 (

0.07 nM (three experiments, mean ( SEM), whereas for the

mutant the K_iwas 2510 ( 220 nM (three experiments, mean (

SEM) (Figure 2A, Table 2), corresponding to a 3000-fold

decrease in binding potency. The other two DAG-dioxolanones

5 and 6 (Figure 2B,C; Table 2) behaved similarly.

In contrast to the dramatic decreases in binding potencies of

the DAG-dioxolanones to the Q27E/δC1b mutant, the binding

affinities of the corresponding DAG lactones Z-1, Z-2, and Z-3

showed much more modest decreases of 35- to 55-fold, similar

to the 20-fold decrease in potency for PDBu (Figure 2D-F,

Table 2). Our results support the prediction from the molecular

modeling that the Gln will provide a unique interaction with

the ring oxygen of the dioxolanone. They are likewise consistent

with the prediction from the studies with the wild-type C1

domain, that the lack of enhanced potency in the presence of

the additional hydrogen bond to the Gln implies that in its

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3469

Figure 3. Representative structures from molecular dynamics simulations of C1 domains in the membrane environment. The center of the bilayer

is indicated by a solid line, and the interfacial region is located between the two dashed lines. Hydrogen bonds to the ligands are indicated by

gold-colored lines: (A) DAG-lactone Z-1 bound sn-2 to PKCδ C1b; (B) DAG-dioxolanone Z-4 bound sn-1 to PKCδ C1b; (C) DAG-lactone Z-1

bound sn-1 to Q27E/PKCδ C1b; (D) DAG-dioxolanone Z-4 bound sn-2 to Q27E/PKCδ C1b; (E) DAG-lactone Z-1 bound sn-2 to PKCδ C1a; (F)

DAG-dioxolanone Z-4 bound sn-1 to PKCδ C1a; (G) DAG-lactone Z-1 bound sn-1 to Q27E/PKCδ C1a; (H) DAG-dioxolanone Z-4 bound sn-2 to

Q27E/PKCδ C1a.

binding potencies of the DAG-dioxolanone and the correspond-

ing DAG-lactone, respectively, for the Q27E/δC1a mutant. The

approximately equal loss of the binding potency for PDBu, the

DAG lactone, and the DAG-dioxolanone implies that the

glutamine residue (Q27) in the δC1a domain has a greater effect

on the domain structure and function than the corresponding

mutation in the δC1b domain, since it abrogates binding with

all ligands and not just the DAG-dioxolanones which form

specific interactions with residue Q27.

Modeling the Dynamic Behavior of C1 Domains in the

Membrane Environment. The sequences of PKCδ C1b and

C1a are 42% identical, and the zinc-binding residues and

residues essential for ligand binding are completely conserved.

This suggests that the backbone structures of these two C1

domains are very similar. An examination of the static structure

of δC1b compared to a homology model of δC1a does not offer

any insight of why mutation of Q27 in these two domains has

such widely different effects on ligand binding. To address this

issue, we performed molecular dynamics simulations of DAG-

lactone Z-1 and the corresponding DAG-dioxolanone 4 (Z-

configuration) bound in the sn-1 and sn-2 orientations to the

PKCδ C1a and C1b, Q27E/δC1a, and Q27E/δC1b domains.

We have shown previously²⁵that interactions between the C1

domain, its ligands, and the membrane environment are essential

to understanding the structure and function of PKC. To take

this into account, the simulations used an implicit solvent model

that approximates the solvation effects of a lipid bilayer

membrane, including the heterogeneous dielectric environment

of the membrane interfacial region where water interacts with

the lipid headgroups and glycerol backbones.²⁹

then underwent molecular dynamics simulation for 10 ns, over

the course of which an ensemble of 100 snapshots of its structure

was collected. We then calculated the average hydrogen-bonding

interaction energy between the C1 domain and the DAG-lactone

or DAG-dioxolanone for each ensemble (Table 4).

In both the wild-type δC1b and δC1a domains, DAG-lactones

bind more strongly in the sn-2 orientation but DAG-dioxol-

anones prefer to bind sn-1. This is consistent with our earlier

results from simulated annealing simulations in vacuum²⁶and

our predictions from the docking studies suggesting that the

DAG-dioxolanones can form an extra hydrogen bond to the C1

domain in the sn-1 binding mode. With the mutant Q27E/δC1b

and Q27E/δC1a domains, however, the situation is exactly

reversed: DAG-lactones prefer the sn-1 binding mode, and

DAG-dioxolanones prefer to bind sn-2. In the DAG-lactones,

although they do not form a stable hydrogen bond with Gln27,

the sn-2 binding mode is favored in part because of transient

weak interactions between atom Nꢀ of Gln27 and the sn-1 ester

oxygen atoms. These become unfavorable repulsive interactions

when Gln27 is mutated to Glu and the side chain nitrogen is

replaced by an oxygen, so the sn-1 binding mode in which no

polar atoms in the ligand are close to Glu27 is now preferred.

With the DAG-dioxolanones, the hydrogen bond at the close

contact between the extra ring oxygen and atom Nꢀ of Gln27

becomes a heavy liability when Gln27 is mutated to Glu, and

the sn-2 binding mode is an improvement, although here as with

the DAG-lactones, there will still be unfavorable interactions

with Glu27 and the sn-1 ester oxygens.

Representative structures from each simulation, with ligands

in their favored binding modes, are shown in Figure 3. In all

cases the C1 domains remain close to their starting configuration

with the binding site loops buried in the interfacial region on

one side of the membrane. The wild-type δC1b domain (Figure

3A,B) retains its crystal conformation, and the ligand remains

A starting configuration for each system was generated by

translating and rotating the C1 domains to a position as

hypothesized,¹⁶with the top part of the C1 domain near the

binding site inserted partially into the membrane. Each system

3470 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

Figure 4. Confocal microscopy comparing membrane translocation of the wild-type full length PKCδ-GFP and its C1b mutant (Q257E/PKCδ-

GFP) by the dioxolanone 4 in living CHO-K1 cells: (A) amino acid sequence alignment of the C1a and C1b domains of PKCδ, where the indicated

glutamine residue is shown in green; (B) time serial images showing the dynamic translocation of wild-type PKCδ-GFP to the plasma membrane

and internal membranes in living CHO-K1 cells after the application of 4 (20 µM); (C) time serial images showing the distribution of the C1b

mutant of PKCδ-GFP (Q257E) after the application of 4 (20 µM), with no membrane translocation observed; (D) time serial images showing the

dynamic translocation of the C1b mutant (Q257E) after the application of PMA (1 µM); (E) time serial images showing the dynamic translocation

of the C1b mutant (Q257E) after the application of Z-1 (20 µM), the DAG-lactone corresponding to 4. The minutes in each panel represent the time

of drug treatment. Each experiment was performed at least three times with similar results.

Table 3. Comparison of the Binding Affinities of the DAG

Table 4. Hydrogen-Bonding Interaction Energies (kcal/mol) between

PKCδ C1 Domains and DAG Ligands.

Dioxolanone 4, the DAG-lactone Z-1 and PDBu for the Wild-Type C1a

Domain of PKCδ (wt/δC1a), and the Glutamate Mutant (Q27E/δC1a)^a

δC1b

δC1a

Q27E/δC1b Q27E/δC1a

K_i(nM)

K_d(nM)

DAG-lactone Z-1

bound sn-1

DAG-lactone Z-1

bound sn-2

DAG-dioxolanone Z-4 -112.54 -109.48

bound sn-1

-103.18 -99.09

-104.28

-100.41

-104.52

-124.21

-109.89

phorbol ester

PDBu

DAG-lactone

DAG-dioxolanone

-106.68 -109.81

-76.76

Z-1

4

Q27E/δC1a

wt/δC1a

K_Q27E/K_wt

383 ( 15

2.04 ( 0.24

188

476 ( 6

1.21 ( 0.11

393

1567 ( 58

4.31 ( 0.13

364

-101.03

-102.18

DAG-dioxolanone Z-4 -109.62 -105.38

bound sn-2

^aThe K_ivalues were measured by competitive binding of the compounds

for the receptors against [³H]PDBu. Each value is an average of three

independent experiments, expressed as the mean ( SEM.

infra). A similar structure with a slightly opened binding site is

observed with the mutant Q27E/δC1b domain bound to DAG-

lactone (Figure 3C), although here the second binding site loop

(residues 20-25) changes shape somewhat to allow residue

Glu27 to be fully solvated outside the interfacial region. With

the remaining mutant structures (Figure 3D,G,H), the binding

site is severely distorted, with the second loop folded open to

lie nearly horizontal to the plane of the bilayer. The binding

site is a full 6 Å wider than in the δC1b domain crystal structure.

stably bound. The wild-type δC1a domain structure is also stable

(Figure 3 E,F), although the binding site is opened about 1 Å

wider as measured from tip to tip across the top, and it does

not penetrate as far into the membrane as the δC1b domain.

This may explain why the δC1a domain ligand binding affinity

is generally slightly lower (Table 2, 3) and why it has different

membrane translocating abilities than the δC1b domain (vide

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3471

These changes explain why the Q27E/δC1b domain has lost

the ability to bind DAG-dioxolanones, and the Q27E/δC1a

domain has lost all ligand binding ability. The changes in

structure are driven by the necessity of solvating the mutant

Glu27 residue, which is pulled out of the interfacial region of

the bilayer by repulsive interactions between Glu27 and oxygen

atoms in the DAG-dioxolanones and, in the case of the δC1a

domain, a lower membrane affinity in general.

Treatment. To probe the role of the C1a domain, we mutated

the corresponding glutamine residue in the δC1a domain (Q185)

to glutamate to generate the C1a mutant (Q185E/PKCδ-GFP).

We found that, in contrast to the behavior of the C1b mutant

(Q257E/PKCδ-GFP), membrane translocation of the C1a mutant

(Q185E/PKCδ-GFP) was easily detected after the application

of DAG-dioxolanone 4 (20 µM) (Figure 5B), as was likewise

observed for PMA (Figure 5C) and the DAG-lactone Z-1 (Figure

5D). This result argues that the C1a domain is not necessary

for translocation of PKCδ by either phorbol ester or DAG

analogues in the presence of a functional C1b domain. Although

the C1a/DAG-dioxolanone interaction had been inhibited in the

C1a mutant, the strong C1b/DAG-dioxolanone interaction drove

the whole PKCδ molecule to the membrane. However, it should

be noted that without the binding to the C1a domain, the plasma

membrane localization of the whole PKCδ molecule seemed

to be less persistent and clear. The C1a mutant was more likely

to translocate to the internal membranes. This behavior was most

evident in the PMA and the DAG-lactone-treated C1a-mutant-

expressing cells (Figure 5C,D). As shown in Figure 5C, the

C1a mutant translocated to both the plasma membrane and the

nuclear membrane in response to PMA (1 µM), but the signal

at the plasma membrane became much weaker several minutes

after the translocation. In Figure 4D, we showed that with PMA

the C1b mutant only translocated to the plasma membrane,

where it persisted. A similar pattern could be seen for the

translocation of the C1a mutant (Figure 5D) and the C1b mutant

(Figure 4E) in response to the DAG-lactone. These results

suggest that the C1a and C1b domains of PKCδ may play

different roles in the ligand interaction and the enzyme activation

by targeting the enzyme to different subcellular locations.

Roles of the Individual C1 Domains in Membrane Trans-

location of Intact PKCδ in Response to DAG-dioxolanone

Treatment in Living CHO Cells. (a) Mutating the Glutamine

Residue (Q257) in the C1b Domain of Full-Length PKCδ

Prevented Its Membrane Translocation in Response to

DAG-dioxolanone Treatment. The dramatic effect of the Q27E

mutation on the binding to the C1b domain of the DAG-

dioxolanones, with a much more modest effect on the binding

by phorbol esters or the DAG-lactones, now provides us with

a powerful tool for probing the role of this domain in PKCδ

function. Previously, an important role for the C1b domain had

been indicated by using a P11G mutation, which dramatically

reduces PDBu binding affinity.^27,30An underlying problem with

interpretation, however, is that the overall effects of this mutation

on PKC conformation cannot be assessed. In contrast, with the

Q27E mutation in the C1b domain of PKCδ, a positive control

is provided by PDBu or by the DAG-lactone, which still retains

an adequate level of affinity, whereas binding to the DAG-

dioxolanone should be lost.

We first studied the role of the C1b domain in PKCδ

membrane translocation. We mutated the glutamine residue

(Q257) of C1b in PKCδ-GFP to generate a C1b mutant of PKCδ

(Q257E/PKCδ-GFP). Then, we monitored the response of the

Q257E/PKCδ-GFP mutant to DAG-dioxolanone 4 in living

CHO cells using real time confocal microscopy and compared

this response to that to phorbol ester or a DAG-lactone. As

shown in Figure 4B, 4 (20 µM) translocated the wild-type

PKCδ-GFP to both the plasma membrane and the internal

membranes within 5 min after DAG-dioxolanone administration.

However, no membrane translocation was observed in the C1b

mutant (Q257E/PKCδ-GFP) even after 1 h of DAG-dioxolanone

treatment (Figure 4C) or with higher drug concentrations (data

not shown). In contrast, treatment with 1 µM PMA induced

clear plasma membrane translocation of the Q257E/PKCδ-GFP

mutant immediately (within 1 min) (Figure 4D), confirming that

the PKCδ remained capable of translocation in the presence of

ligands that could still interact with the mutant C1b domain.

Likewise, when we applied the DAG-lactone Z-1 (20 µM) to

the C1b mutant-expressing cells, membrane translocation of the

Q257E/PKCδ-GFP mutant was observed within 10 min after

administration of the compound (Figure 4E). These results

indicate that without ligand binding to δC1b, δC1a alone could

not translocate the intact PKCδ to the membranes in response

to DAG-dioxolanone, implying that the C1a domain does not

play the major role in PKCδ redistribution. The concentration

of compound 4 (20 µM) was chosen to give a prompt, complete

response to ligand addition. Typically, to achieve this goal,

amounts used by researchers are appreciably higher than the

measured K_ivalues determined in vitro, which represent ideal

conditions for supporting the interactions. In the current studies,

the objective was to show the differential pattern of response,

so complete dose response curves, which are quite demanding

of instrument time, were not performed.

Discussion

As a central regulatory mechanism for a diversity of cellular

functions, such as cell proliferation, differentiation, and apop-

tosis, the diacylglycerol signaling pathway has gained wide-

spread attention in the biological and pharmaceutical fields.³¹

Protein kinase C (PKC), the major receptor for the DAG signal

and one of the important factors in tumorigenesis,²has thus

emerged as an attractive target for cancer therapy and a range

of other conditions. Several strategies have been used to develop

the PKC modulators. The first strategy has been inhibitors

targeted to the kinase domain of the PKCs. For example,

LY333531, a selective inhibitor of PKCâ enzymatic activity,

is currently in clinical trials for treatment of vascular complica-

tions of diabetes.^32,33A second strategy has been to use antisense

oligonucleotides to manipulate the expression of the PKCs. For

example, the PKCR antisense oligonucleotide LY900003/ISIS

3521 is currently in clinical trials.³⁴A third strategy has been

the use of compounds targeted to the C1 domains, where they

are able to induce selective responses. Bryostatin 1, which is

in clinical trials as a cancer chemotherapeutic agent, has a unique

biphasic dose response for down-regulation and protection of

PKCδ.^35,3612-Deoxyphorbol 13-acetate is antitumor-promot-

ing³⁷and is under consideration for use in combination with

HAART therapy for treatment of HIV/AIDS.³⁸Ingenol 3-an-

gelate, which induces biphasic induction of IL-6,³⁹is in clinical

trials for actinic keratosis and nonmelanotic skin cancer. A

general rationale for this approach is that different PKC isoforms

or different families of DAG receptor proteins may have

antagonistic effects, depending on the system, and thus, for

example, blocking PKCꢀ, an isoform that induces proliferation

and is antiapoptotic, may be equivalent to stimulation of PKCδ,

which is antiproliferative and proapoptotic.

(b) Mutating the Corresponding Glutamine Residue

(Q185) in the C1a Domain of PKCδ Did Not Abolish Its

Membrane Translocation in Response to DAG-dioxolanone

3472 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

Figure 5. Confocal microscopy comparing membrane translocation of the C1a mutant (Q185E/PKCδ-GFP) of PKCδ-GFP by DAG-dioxolanone

4, its corresponding DAG- lactone Z-1 and PMA in living CHO-K1 cells: (A) time serial images showing the dynamic translocation of wild-type

PKCδ-GFP to the plasma membrane and internal membranes in living CHO-K1 cells after the application of 4 (20 µM); (B) time serial images

showing the dynamic membrane translocation of the C1a mutant (Q185E) after the application of 4 (20 µM); (C) time serial images showing the

dynamic translocation of the C1a mutant (Q185E) after the application of PMA (1 µM); (D) time serial images showing the dynamic translocation

of the C1a mutant (Q185E) after the application of corresponding DAG-lactone Z-1 (20 µM). The minutes in each panel represent the time of drug

treatment. Each experiment was performed three times with similar results.

Although natural products such as the phorbol esters, indole

alkaloids, aplysiatoxin, or bryostatin have impressive potency

for C1 domains, their intricate structures and multiple chiral

centers have complicated their exploitation as lead structures

for medicinal chemistry. Unfortunately, although the endogenous

ligand for C1 domains, namely, sn-1,2-diacylglycerol, has a

relatively simple structure with only a single chiral center, its

conformational flexibility is associated with relatively weak

potency. Starting with the DAG structure, our group has been

able to develop DAG-lactones with in vitro potencies approach-

ing those of the phorbol esters using a combination of

conformational constraint and appropriate substitution to opti-

mize interactions with the lipid bilayer and the rim of the C1

domain binding cleft.¹⁷The DAG-lactones have provided a

powerful template for structural modification to probe the

interactions of ligands with the C1 domains. Further insights

have emerged from molecular modeling, based on the X-ray

structure of the complex between phorbol 13-acetate and the

C1b domain of PKCδ¹⁶as well as from solution NMR studies.⁴⁰

As part of our effort to further enhance ligand potency and

C1 domain selectivity, we have sought to identify additional

potential points of interaction between DAG-lactones and the

C1 domains. We described here a novel class of DAG

derivatives, DAG-dioxolanones, that were predicted from

computer modeling to possess such an additional interaction

site. The DAG-dioxolanones have an oxygen replacing the CH₂

in the ring structure of the DAG lactones (Table 1). The

modeling predicted that this extra oxygen in the DAG-

dioxolanones should be able to form an additional hydrogen

bond with the glutamine residue in the C1b domain of PKCδ

(Q257 in the numbering scheme for the intact PKCδ, Q27 in

the numbering scheme of the isolated C1b domain as shown in

Figure 1) only in the sn-1 binding mode involving the sn-1

carbonyl on the side chain. We were able to verify the

importance of this predicted interaction site by mutating the

indicated glutamine residue in the isolated δC1b domain. It has

been reported that the side chain of the glutamine residue is

very important for the geometry of the binding cleft of the C1

domain.⁴¹We therefore had tried to minimize the side effect of

the mutation by replacing the glutamine residue (Q) with

glutamate (E). Fortunately, with the glutamate residue, the

binding affinity of the mutated C1b domain (Q27E/δC1b) for

[³H]PDBu was maintained at a level with K_din the 10^-9

M

range (6.2 ( 1.0 nM), reflecting a relatively weak influence

(20-fold increase in the K_dvalue compared with the K_dof 0.33

( 0.05 nM for the wild-type) of the mutation on its binding

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3473

capability for the phorbol esters. Encouragingly, our measure-

ments on the DAG-dioxolanones demonstrated that the mutation

dramatically decreased the binding affinity. For the three DAG-

dioxolanones 4-6 that we had studied, the K_ivalues had

increased from 10^-9M for the wild-type δC1b to 10^-6M for

the δC1b mutant. However, the corresponding DAG-lactones

only showed a weak decrease (35- to 55-fold of increase in the

K_ivalues) in their binding potency for the Q27E/δC1b mutant

(Table 2). It is thus quite clear from the above results that, as

predicted by the computer modeling, the glutamine residue of

the δC1b domain does play an important role in the interaction

with the DAG-dioxolanone, and this interaction is weak or

nonexistent for the DAG-lactones or phorbol esters.

mutations that abrogate binding activity (e.g., P11G, numbered

according to the sequence of the isolated C1 domain) is the

lack of a positive control to distinguish the effect of the alteration

itself from the lack of recognition of ligand. Because the Q27E

mutation selectively blocks response to the DAG-dioxolanone

but not phorbol ester or DAG-lactone, comparison of the

response of the mutated PKCδ to these ligands provides a test

for the function of the C1b domain. We showed (Figures 4 and

5) that when we abolished the C1b/dioxolanone interaction by

mutating the indicated glutamine residue (Q257) in the C1b

domain, the whole PKCδ enzyme no longer translocated to the

membranes in response to DAG-dioxolanone 4 (Figure 4C).

Since DAG-dioxolanone 4 bound with good affinity to the wild-

type C1a domain of PKCδ (K_ivalue of 4.31 ( 0.13 nM, Table

3), the lack of translocation of the C1b mutant (Q257E/PKCδ)

by the DAG-dioxolanone indicates that the C1a-dioxolanone

interaction by itself was unable to drive translocation of the

whole PKCδ protein to the membranes, at least over the ligand

range examined. The C1b/dioxolanone interaction would thus

appear to be the major force for driving the membrane

translocation of the whole PKCδ enzyme. This conclusion was

consistent with our results with the C1a mutant of PKCδ

(Q185E/PKCδ). Although this aspect of the analysis is less

robust, in that the mutation markedly reduces the potency of

the C1a domain for the other ligands as well, depriving us of a

positive control (Table 3), the retention of translocation of the

mutated PKCδ enzyme again argues that the C1a/ligand

interaction is not critical for the membrane translocation of the

whole PKCδ enzyme in response to these ligands.

Close examination of the data is consistent, in any case, with

a secondary contribution of the C1a domain of PKCδ to the

membrane site to which PKCδ translocates. In response to 1

µM PMA, the C1a mutant (Q185E) (functional C1b domain,

reduced affinity at the C1a domain for PMA) translocated to

both plasma membrane and the nuclear membrane (Figure 5C),

whereas the C1b mutant (Q257E) (functional C1a domain,

reduced affinity at the C1b domain for PMA) only translocated

to the plasma membrane in response to PMA (Figure 4D). No

nuclear membrane translocation could be seen even after 1 h

of the PMA treatment (data not shown). For the native PKCδ,

DAG-lactones cause translocation to both plasma membrane

and the nuclear membrane (data not shown). In response to the

DAG-lactone, the C1b mutant only translocated to the plasma

membrane (Figure 4 E), whereas the C1a mutant only translo-

cated to the nuclear membrane (Figure 5 D). The results

presumably reflect the reduced affinity of the mutated C1

domain for the phorbol ester or DAG-lactone, with the C1a

domain favoring plasma membrane localization and the C1b

domain being less selective. They also emphasize that both

domains influence translocation of PKCδ, even if the C1b

domain plays the predominant role.

One objective of strategies for introducing additional points

of interaction would be to further enhance potency. This

objective was not met with the current DAG-dioxolanones

because of their exclusive preference for the sn-1 binding mode

that counterbalanced the entropic advantage typically observed

for DAG-lactones that bind in the sn-2 binding mode. Simply

put, the extra oxygen in DAG-dioxolanones forms an effective

hydrogen bond with glutamine (Q27E) only when it binds in

the sn-1 mode. Therefore, despite this additional hydrogen bond

(Figure 1), the binding affinities of these new DAG derivatives

were not increased relative to the corresponding DAG-lactones,

neither for the intact PKCR (Table 1) nor for the isolated C1b

domain of PKCδ (Table 2). Rather, there was a several-fold

decrease in their binding affinities (Tables 1 and 2). Since the

mutation of the glutamine residue (Q27E/δC1b) much more

strongly reduced the binding potencies of the DAG-dioxolanone

than of the DAG lactones, a plausible explanation is that the

δC1b domain is able to retain binding to the DAG-lactones even

after mutation, whereas interactions between the mutated Glu27

residue and the DAG-dioxolanone are so unfavorable that

binding is lost.

The DAG-responsive PKC and PKD isoforms possess twin

C1 domains. Their relative contributions to PKC function,

despite considerable analysis, still remain unresolved, with

different methodological approaches yielding different conclu-

sions. For example, in terms of binding affinity, initial studies

showed that both C1 domains of PKCγ bound PDBu with high

affinity, indicating that the C1a and C1b of PKCγ are function-

ally equivalent.^7,42By expressing the truncated PKCR in the

yeast Saccharomyces cereVisiae, Riedel’s group found that the

presence of either C1 domain allowed phospholipid and phorbol

ester regulation of PKCR enzymatic activity, implying an equal

function of the C1a and C1b domains in PKCR.⁴³In contrast,

later investigations suggested that the two C1 domains function

differently in PKC activation. Cho and co-workers described

that the hydrophobic residues in the C1a domain were essential

for the membrane penetration and activation of PKCR, whereas

those in the C1b domain were not directly involved in these

processes.⁴⁴Using dimeric bisphorbols, Newton’s group dem-

onstrated that both C1 domains of PKCâ were oriented for

potential membrane interaction but only one C1 domain bound

ligand in a physiological context.⁴⁵With synthetic C1 peptides,

Irie’s group reported different binding affinities of the C1a and

C1b domains of different PKC isoforms for various ligands.^46,47

Our previous work has demonstrated differential roles of the

C1a and C1b domains in translocation and down-regulation of

PKCδ, with different responses for different ligands.^48,49

It has been suggested elsewhere that phorbol esters and DAG

interact with different C1 domains, with the phorbol ester

interacting with the C1b domain and the DAG interacting with

the C1a domain.⁵⁰Our results argue that this is not of major

importance for translocation of PKCδ and that the C1b domain

plays the predominant role for the DAG-lactones as well as for

the phorbol esters.

The high conservation of phorbol ester binding affinity across

species as diverse as nematodes, drosophila, and man had

suggested that different C1 domains should show very similar

characteristics for ligand recognition. Emerging evidence from

multiple groups, however, emphasizes impressive diversity

among the properties of different C1 domains.^45,46,48For

The combination of the DAG-dioxolanones and the Q257E

mutation provides a powerful tool for addressing the role of

ligand binding to the C1 domain in PKC functioning. An

underlying problem with deletion of an entire C1 domain or of

3474 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

each construct was finally confirmed by sequence analysis (DNA

Minicore, Center for Cancer Research, NCI, NIH).

example, we described that the RasGRP-selective DAG-lactone

130C037 bound with high affinity to the C1b domain of PKCδ

but with orders of magnitude lower affinity to the C1a domain

of PKCδ or to either C1 domain of PKCR.¹⁹The substantially

different effect of the mutation Q27E for the C1a domain of

PKCδ, where it had no differential effect on binding of the

DAG-dioxolanone compared to the DAG-lactone or the phorbol

ester, provides yet another example of this diversity.

In a summary, our results described here again illustrate the

power of the DAG-lactone template as a platform for exploration

of ligand-C1 domain interactions. Understanding of these

interactions is central to the rational design of ligands selective

for PKC isoforms or other members of the families of proteins

with C1 domains that mediate signaling downstream of the

lipophilic second messenger DAG. These proteins represent

promising therapeutic targets, given both their biological func-

tions and the actions of agents currently under development.

After construction, the recombinant plasmids of individual C1

domains of PKCδ were transformed into BL-21-Gold (DE3) E.

coli competent cells (Stratagene, La Jolla, CA). The expression of

the GST-fusion proteins was induced by the addition of 0.5 mM

isopropyl-O-D-thiogalactopyranoside (IPTG) when the OD of the

LB medium reached 0.5-0.7. The bacteria were harvested after 4

h of induction at 37 °C. The expressed GST-tagged C1 protein

was purified using a B-PER GST spin purification kit according to

the manufacturer’s instructions (Pierce Biotechnology, Inc., Rock-

ford, IL). The purity of the protein was verified by SDS-PAGE

and staining with Coomasie Blue. The protein concentration was

measured using the Bio-Rad protein assay kit according to the

manufacturer’s instructions (Bio-Rad, Hercules, CA). The purified

GST-C1 protein was stored in 30% glycerol at -70 °C.

Site-Directed Mutagenesis of the C1 Domains of PKCδ.

Mutagenesis of the glutamine residue in the C1a and C1b domains

of PKCδ was performed in the isolated GST-δC1a, GST-δC1b,

and the intact PKCδ-GFP proteins using the QuikChange II site-

directed mutagenesis kit (Stratagene) according to the manufac-

turer’s instructions. The C1 domain containing pGEX plasmid or

the intact PKCδ containing pEGFP plasmid was used as the

template for the mutagenesis reactions. The glutamine/glutamate

mutagenic primers for the δC1b domain were 5′-GGGACTGGT-

GAAGGAGGGATTAAAGTGTGAAG-3′ (sense) and 5′-CTTCA-

CACTTTAATCCCTCCTTCACCAGTCCC-3′ (antisense). The mu-

tagenic primers for the δC1a domain were 5′-GGGCCTCAA-

CAAGGAAGGCTACAAATGCAGG-3′ (sense) and 5′-CCTG-

CATTTGTAGCCTTCCTTGTTGAGGCCC-3′ (antisense). The mu-

tated codons are underlined. The mutation was confirmed by DNA

sequence analysis (DNA Minicore, Center for Cancer Research,

NCI, NIH). The mutated plasmids were used to express the

corresponding GST or GFP fusion proteins, which were purified

or used in the experiments as described elsewhere in this study.

[³H]PDBu Binding Assay. Enzyme-ligand interactions were

assessed in terms of the ability of the ligand to displace bound

[20-³H]phorbol 12,13-dibutyrate (PDBU) from a recombinant single

isozyme (PKCR) in the presence of phosphatidylserine.^52-56[³H]-

PDBu binding to the recombinant full-length PKCR or the

individual C1 domains was measured using the polyethylene glycol

precipitation assay developed in our laboratory as described

elsewhere.⁵²Briefly, the assay mixture (250 µL) contained 50 mM

Tris-HCl (pH 7.4), 100 µg/mL phosphatidylserine, 4 mg/mL bovine

immunoglobulin G, [³H]PDBu, and various concentrations of

competing ligand. Incubation was carried out at 37 °C for 5 min

(for full-length PKC isoforms) or 18 °C for 10 min (for C1

domains). Samples were chilled on ice for 7 min, and 200 µL of

35% polyethylene glycol in 50 mM Tris-HCl (pH 7.4) was added.

The samples were mixed and incubated on ice for an additional 10

min. The tubes were centrifuged in a Beckman Allegra 21R

centrifuge at 4 °C (12 200 rpm, 15 min). A 100 µL aliquot of the

supernatant was removed for the determination of the free

concentration of [³H]PDBu, and the pellet was carefully dried. The

tip of the centrifuge tube containing the pellet was cut off and

transferred to a scintillation vial for the determination of the total

bound [³H]PDBu. Cytoscint (ICN, Costa Mesa, CA) was added to

both an aliquot of the supernatant and the pellet. Radioactivity was

determined by scintillation counting. Specific binding was calculated

as the difference between total and nonspecific binding. Standard

Scatchard analysis was performed to determine the dissociation

constants (K_d) of the individual C1 domains, and the inhibitory

dissociation constants (K_i) were calculated using our standard

method as described previously.⁵²

Experimental Section

Materials and General Procedures. All reagents for chemical

syntheses were commercially available. [20-³H]Phorbol 12,13-

dibutyrate ([³H]PDBu) (20 Ci/mmol) was purchased from Perkin-

Elmer (Boston, MA). PDBu and phorbol 12-myristate-13-acetate

(PMA) were purchased from LC Laboratories (Woburn, MA).

Phosphatidyl-L-serine was purchased from Avanti Polar Lipids

(Alabaster, AL). Reagents for expression and purification of

glutathione S-transferase (GST) fusion proteins were obtained from

Pierce Biotechnology, Inc. (Rockford, IL). Cell culture medium

and reagents and the DNA primers were obtained from Invitrogen

(Carlsbad, CA). Recombinant mouse PKCR was expressed and

partially purified as described elsewhere.⁵¹Melting points were

determined on a MelTemp II apparatus, Laboratory Devices, and

are uncorrected. Column chromatography was performed on silica

gel 60, 230-400 mesh (E. Merck or Bodman Industries), and

analytical TLC was performed on Analtech Uniplates silica gel GF.

¹H and ¹³C NMR spectra were recorded on a Varian Unity Inova

instrument at 400 and 100 MHz, respectively. Spectra are referenced

to the solvent in which they were run (7.26 ppm for CDCl₃).

Positive-ion fast atom bombardment mass spectra (FABMS) were

obtained on a VG 7070E-HF double-focusing mass spectrometer

operated at an accelerating voltage of 6 kV under the control of a

MASPEC-II data system for Windows (Mass Spectrometry Ser-

vices, Ltd.). Either glycerol or 3-nitrobenzyl alcohol was used as

the sample matrix, and ionization was effected by a beam of xenon

atoms generated in a saddle-field ion gun at 8.0 ( 0.5 kV. Nominal

mass spectra were obtained at a resolution of 1200, and matrix-

derived ions were background-subtracted during data system

processing. Elemental analyses were performed by Atlantic Micro-

lab, Inc., Norcross, GA. The partition coefficients (log P) were

calculated according to the atom-based program MOE SLog P.²³

Construction and Purification of GST-Fused C1 Domains of

PKCr and PKCδ in E. coli. The C1a and C1b domains of PKCδ

were generated by polymerase chain reaction (PCR) using the

Platinum Pfx DNA polymerase (Invitrogen). The full-length cDNA

clone of murine PKCδ was used as the template. The following

oligonucleotides were used as the PCR primers to pull out the

targeted C1 domains: (1) forward and reverse primers for δC1a

were 5′-AAACAGGCCAAGATCCACTACA-3′ and 5′-GGTGTC-

CCGGCTATTGGT-3′; (2) forward and reverse primers for δC1b

were 5′-CAGAAAGAACGCTTCAACATCG-3′ and 5′-GGCCT-

CAGCCAAGAGCTTT-3′. The blunt-ended PCR products were

ligated into a pCR-Blunt vector using the Zero Blunt PCR cloning

kit (Invitrogen). The pCR-Blunt vector was digested with EcoRI

(New England BioLabs, Inc., Beverly, MA) to produce adhesive

ends of the C1 fragment. This fragment was then ligated into the

appropriate glutathione S-transferase (GST)-containing vectors (i.e.,

pGEX-5X-1, pGEX-5X-2, pGEX-5X-3) (Amersham Biosciences,

Piscataway, NJ), using the EcoRI restriction sites with the sequence

of the insert in the intended reading frame. The DNA sequence of

Expression and Imaging of the GFP-Tagged PKC Proteins

in Live CHO Cells. CHO-K1 cells (obtained from ATCC,

Manassas, VA) were cultured at 37 °C in F-12 (HAM) nutrient

containing 10% FBS, penicillin (50 units/mL), and streptomycin

(0.05 mg/mL) in a 5% CO₂atmosphere. The plasmid DNA of GFP-

fused PKC proteins was transfected into the CHO-K1 cells using

Lipofectamine 2000 (Invitrogen) according to the manufacturer’s

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3475

instructions. The expression of the fluorescent protein was detected

24 h after transfection. Confocal fluorescent images were collected

with a Bio-Rad MRC 1024 confocal scan head (Bio-Rad) mounted

on a Nikon microscope with a 60× planapochromat lens. Excitation

at 488 nm was generated by a krypton-argon gas laser with a 522/

32 emission filter for green fluorescence. For kinetics of PKC-

GFP translocation in live cells, cells plated on a 40 mm round

coverslip were enclosed in a Bioptechs Focht Chamber System

(Bioptechs, Butler, PA). The chamber was inverted and attached

to the microscope stage with a custom stage adapter. A temperature

controller set at 37 °C was connected, and medium was perfused

through the chamber with a Lambda microperfusion pump.

Sequential images of the same cell were collected at various time

points using LaserSharp Software (Bio-Rad).

Docking of DAG-dioxolanones into the C1b Domain of

PKCδ. Structures for all compounds were built in SYBYL⁵⁷and

minimized with the MMFF94 force field and partial charges.⁵⁸

Docking was then performed using FlexX⁵⁹through its SYBYL

module into the crystal structure of the C1b domain of PKCδ.¹⁶

The binding site was defined as residues 238-243, 250-254, and

257. The ring structure in the ligand was treated flexibly, and all

other options were set to their default values.

Homology Modeling of the C1a Domain of PKCδ and the

Q27E Mutants. A homology model of the C1a domain of PKCδ

was built on the backbone coordinates of the crystal structure of

the C1b domain.¹⁶Side chains for the model were constructed using

the program SCWRL,⁶⁰which uses a backbone-dependent rotamer

library to place residues in their most likely conformation given the

backbone φ-ψ angles at that position. Residues homologous to

PKCδ were left unchanged from their crystallographic positions. The

model was refined with a small energy minimization, with phorbol

13-O-acetate left in position from the crystal structure to prevent

the binding site loops from closing during the minimization. Harmon-

ic positional restraints on the backbone atoms were gradually relaxed

over the course of the minimization to eliminate steric clashes in

the side chains without inducing deformations in the backbone.

Structures for the mutant Q27E C1a and C1b domains were built

by simply replacing the glutamine residue at position 27 in the

structure with a glutamate residue, without any subsequent mini-

mization or other refinement.

Molecular Dynamics with an Implicit Solvent Bilayer. Simu-

lations were run in CHARMM,⁶¹version c32b2, using the all-

hydrogen force field parameter set for proteins⁶²with backbone

dihedral corrections.⁶³Additional parameters for the DAG-lactones

and DAG-dioxolanones were developed by analogy with existing

parameters for lipids and furanose sugars and were checked against

ab initio geometry-optimized structures calculated at the RHF/6-

31G* theory level.

3.0 Å in order to keep the ligand in the binding site and to preserve

its sn-1 or sn-2 orientation. Without these restraints, ligands tended

to drift in and out of the binding site and switch binding orientations

(data not shown).

Each system first underwent 25 steps of steepest descent

minimization with the ligand atoms and protein backbone atoms

fixed in place, followed by 100 steps of adopted-basis Newton-

Raphson minimization with harmonic constraints on the backbone

atoms and finally 200 steps of adopted-basis Newton-Raphson

minimization with no constraints. The systems were then heated

in 10 K increments every 0.2 ps from a starting temperature of 50

to 300 K and equilibrated at 300 K for 100 ps. After equilibration,

data were collected over the course of 10 ns, with coordinates

written to disk at 100 ps intervals to give an ensemble of 100

structure snapshots for each system.

The hydrogen-bonding energy for each system was calculated

in CHARMM by first defining four subsets consisting of ligand

donor atoms, ligand acceptor atoms, receptor donor atoms, and

receptor acceptor atoms. Each potential donor-acceptor pair was

tested to see whether it was within allowable hydrogen-bonding

geometry, i.e., with an H‚‚‚A distance of 2.7 Å or less and a

D-H‚‚‚A angle greater than 90°. If so, the nonbonded (electrostatic

and van der Waals) interaction energy between the two atoms was

calculated. The sum of these interaction energies for all the

snapshots in the ensemble was then divided by 100 to give the

average hydrogen-bonding energy for the ensemble.

E- and Z-5-[(4-Methoxyphenoxy)methyl]-3-(3-methylbutyl-

idene)-5-[(phenylmethoxy)methyl]-4,5-dihydrofuran-2-one (VI,

R ) CH₂CH(CH₃)₂). Lithium bis(trimethylsilyl)amide (LHMDS,

5.4 mL, 5.4 mmol, 1 M in THF) was added dropwise to a -78 °C

stirring solution of IV^20,21(920 mg, 2.7 mmol) in THF (10 mL).

After 10 min, the mixture was treated with isovaleryl aldehyde (0.29

mL, 2.7 mmol) and stirred at the same temperature for 20 min.

The reaction mixture was then quenched by slow addition of

saturated aqueous NH₄Cl and extracted with Et₂O (3×). The

combined organic layers were then washed with water (1×) and

brine (1×), dried over MgSO₄, and concentrated in vacuo. The

residue was purified by flash column chromatography on silica gel

with EtOAc/hexanes (1:4) to give V (R ) CH₂CH(CH₃)₂), which

was used immediately without further purification or characteriza-

tion. This compound was taken up in CH₂Cl₂(20 mL) and treated

with Et₃N (1.1 mL, 8.1 mmol) and methanesulfonyl chloride (0.31

mL, 4.1 mmol). After the mixture was stirred for 10 min at room

temperature, DBU (1.2 mL, 8.1 mmol) was added and the mixture

was stirred overnight. Concentration followed by purification by

silica gel chromatography with EtOAc/hexanes (1:6) gave Z-VI

(R ) CH₂CH(CH₃)₂, 260 mg, 23%) and E-VI (R ) CH₂CH(CH₃)₂,

512 mg, 46%) as a mixture of isomers that were separated via

column chromatography.

The implicit solvent was modeled using the GBSW module

(generalized Born with a simple switching function²⁹). A low-

dielectric, solvent-inaccessible planar slab perpendicular to the z-axis

was set up at z ) 0. The thickness of the slab was 25.0 Å, and the

width of the smoothing region between the membrane and water

environments was 5.0 Å. In this smoothing region, a polynomial

function is used to scale the solvent (water) accessibility from 0

inside the membrane to 1 outside, which gives an approximation

of the dielectric characteristics of a real membrane bilayer. For the

polar portion of the solvation energy, the external solvent dielectric

constant was set to 80.0 and the internal dielectric constant of the

protein was set to 1.0, with a smoothing length of 0.6 Å at the

boundary between the two. Optimized atomic Born radii⁶⁴were

used to define the dielectric boundary at the protein surface. The

surface tension coefficient for the nonpolar portion of the solvation

Z-VI (R ) CH₂CH(CH₃)₂). ¹H NMR (CDCl₃) δ 7.27-7.36 (m,

5 H, PhCH₂OCH₂), 6.83 (s, 4 H, MeOC₄H₄OCH₂), 6.24 (tt, 1 H, J

) 7.8, 2.3 Hz, CdCHCH₂CH(CH₃)₂), 4.60 (AB q, 2 H, J ) 12.1

Hz, PhCH₂OCH₂), 4.03 (AB q, J ) 9.8 Hz, 2 H, MeOC₄H₄OCH₂),

3.76 (s, 3 H, MeOC₄H₄OCH₂),3.68 (AB q, J ) 10.2 Hz, 2 H,

PhCH₂OCH₂), 2.91 (AB qm, J ) 16.5, 2 H, H-4_ab), 2.64-2.68 (m,

2 H, CdCHCH₂CH(CH₃)₂), 1.75 (sept, 1 H, J ) 6.7 Hz,

CdCHCH₂CH(CH₃)₂), 0.96 and 0.97 (d, J ) 6.7 Hz, 6 H, Cd

CHCH₂CH(CH₃)₂). FAB-MS (m/z, relative intensity): 411 (MH⁺,

45), 410 (M^•+, 74), 91 (100), Anal. (C₂₅H₃₀O₅) C,H.

E-VI (R ) CH₂CH(CH₃)₂). ¹H NMR (CDCl₃) δ 7.24-7.34 (m,

5 H, PhCH₂OCH₂), 6.80 (s, 4 H, MeOC₄H₄OCH₂), 6.77 (tt, 1 H, J

) 7.7, 2.8 Hz, CdCHCH₂CH(CH₃)₂), 4.57 (AB q, 2 H, J ) 12.1

Hz, PhCH₂OCH₂), 4.03 (AB q, J ) 9.8 Hz, 2 H, MeOC₄H₄OCH₂),

3.74 (s, 3 H, MeOC₄H₄OCH₂), 3.67 (AB q, J ) 10.2 Hz, 2 H,

PhCH₂OCH₂), 2.91 (AB qm, J ) 17.1, 2 H, H-4_ab), 2.01-2.08 (m,

2 H, CdCHCH₂CH(CH₃)₂), 1.80 (sept, J ) 6.7 Hz, 1 H,

CdCHCH₂CH(CH₃)₂), 0.933 and 0.930 (d, J ) 6.7 Hz, 6 H, Cd

CHCH₂CH(CH₃)₂). FAB-MS (m/z, relative intensity): 411 (MH⁺,

45), 410 (M^•+, 74), 91 (100), Anal. (C₂₅H₃₀O₅) C,H.

energy was 0.04 kcal mol^-1Å^-2

.

We began with structures for DAG-lactone Z-1 and DAG-

dioxolanone Z-4, in both the sn-1 and sn-2 binding modes, docked

into the PKCδ C1b domain, the homology model of the PKCδ

C1a domain, and models of the mutant Q27E C1b and C1a domains,

for a total of 16 systems. Weak constraints were included to hold

the distance between the atoms involved in the three conserved

hydrogen bonds between the ligands and the protein at less than

Z- and E-5-[(4-Methoxyphenoxy)methyl]-3-[5-methyl-3-(2-

methylpropyl)hexylidene]-5-[(phenylmethoxy)methyl]-4,5-dihy-

3476 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

drofuran-2-one (VI, R ) CH₂CH(CH₂CH(CH₃)₂)₂. These com-

pounds have been previously reported.²¹

CH₂CH(CH₃)₂, 392 mg, 1.2 mmol) was combined with 5-methyl-

3-(2-methylpropyl)hexanoyl chloride (245 mg, 1.2 mmol) to give

E-VIII (R ) CH₂CH(CH₃)₂, R′ ) CH₂CH(CH₂CH(CH₃)₂)₂, 265

mg, 47%) as a colorless oil. ¹H NMR (CDCl₃) δ 7.26-7.35 (m, 5

H, PhCH₂OCH₂), 6.75 (tt, J ) 7.7, 2.8 Hz, 1 H, CdCHCH₂CH-

(CH₃)₂), 4.55 (s, 2 H, PhCH₂OCH₂), 4.21 (s, 2 H, (CH₃)₂CHCH₂)₂-

CHCH₂CO₂CH₂), 3.55 (AB q, J ) 10.0 Hz, 2 H, PhCH₂OCH₂),

2.73 (AB qm, 2 H, J ) 17.1 Hz, H-3_ab), 2.18-2.20 (m, 2 H, (CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 2.04 (app t, 1 H, CdCHCH₂CH(CH₃)₂),

1.91 (app p, 2 H, (CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 1.79 (app sept,

1 H, CdCHCH₂CH(CH₃)₂), 1.51-1.64 (m, 2 H, (CH₃)₂CHCH₂)₂-

CHCH₂CO₂CH₂), 1.02-1.14 (m, 4 H, (CH₃)₂CHCH₂)₂CHCH₂CO₂-

CH₂), 0.93 and 0.92 (d, J ) 6.6 Hz, 6 H, CdCHCH₂CH(CH₃)₂),

0.84-0.87 (m, 12 H, (CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂). FAB-MS

(m/z, relative intensity): 473 (MH⁺, 24), 91 (100). Anal. (C₂₉H₄₄O₅‚

0.5H₂O) C, H.

Z-{4-[5-Methyl-3-(2-methylpropyl)hexylidene]-5-oxo-2-[(phen-

ylmethoxy)methyl]-2-(2,3-dihydrofuryl)}methyl 3-Methylbu-

tanoate (Z-VIII, R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH-

(CH₃)₂). According to the general procedure, Z-VII (R )

CH₂CH(CH₂CH(CH₃)₂)₂, 298 mg, 0.76 mmol) was combined with

isovaleryl chloride (0.14 mL, 1.1 mmol) to give Z-VIII (R ) CH₂-

CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂, 309 mg, 86%) as a

colorless oil. ¹H NMR (CDCl₃) δ 7.27-7.36 (m, 5 H, PhCH₂OCH₂),

6.20 (tt, 1 H, J ) 7.5, 2.2 Hz, CdCHCH₂CH(CH₂CH(CH₃)₂)₂, 4.56

(s, 2 H, PhCH₂OCH₂), 4.22 (AB q, 2 H, J ) 11.9 Hz, (CH₃)₂-

CHCH₂CO₂CH₂), 3.54 (AB q, 2 H, J ) 9.9 Hz, PhCH₂OCH₂),

2.82 (AB qm, 2 H, J ) 16.4 Hz, H-3_a,b), 2.65-2.69 (m, 2 H, Cd

CHCH₂CH(CH₂CH(CH₃)₂)₂), 2.17-2.19 (m, 2 H, CH₃)₂CHCH₂-

CO₂CH₂), 2.00-2.12 (m, 1 H, (CH₃)₂CHCH₂CO₂CH₂), 1.57-1.70

(m, 3 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂, 0.99-1.11 (m, 4 H, Cd

CHCH₂CH(CH₂CH(CH₃)₂)₂, 0.94 and 0.93 (s, 6 H, (CH₃)₂CHCH₂-

CO₂CH₂),0.84-0.87 (m, 12 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂).

FAB-MS (m/z, relative intensity): 473 (MH⁺, 27), 91 (100). Anal.

(C₂₉H₄₄O₅) C, H.

E-{4-[5-Methyl-3-(2-methylpropyl)hexylidene]-5-oxo-2-[(phen-

ylmethoxy)methyl]-2-(2,3-dihydrofuryl)}methyl 3-Methylbu-

tanoate (E-VIII, R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH-

(CH₃)₂). According to the general procedure, E-VII (R )

CH₂CH(CH₂CH(CH₃)₂)₂, 87 mg, 0.22 mmol) was combined with

isovaleryl chloride (0.04 mL, 0.33 mmol) to give E-VIII (R ) CH₂-

CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂, 103 mg, 100%) as a

colorless oil. ¹H NMR (CDCl₃) δ 7.27-7.36 (m, 5 H, PhCH₂OCH₂),

6.77 (tt, J ) 7.6, 2.8 Hz, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂, 4.55

(s, 2 H, PhCH₂OCH₂), 4.23 (AB q, J ) 11.9 Hz, 2 H, (CH₃)₂-

CHCH₂CO₂CH₂), 3.55 (AB q, J ) 9.9 Hz, 2 H, PhCH₂OCH₂),

2.73 (AB qm, 2 H, J ) 17.1 Hz, H-3_ab), 2.16-2.18 (m, 2 H, (CH₃)₂-

CHCH₂CO₂CH₂), 2.01-2.16 (m, 3 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂and (CH₃)₂CHCH₂CO₂CH₂), 1.57-1.73 (m, 3 H, Cd

CHCH₂CH(CH₂CH(CH₃)₂)₂, 1.06-1.12 (m, 4 H, CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂, 0.94 and 0.92 (s, 6 H, (CH₃)₂CHCH₂CO₂CH₂),

0.87 and 0.85 (m, 12 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂). FAB-

MS (m/z, relative intensity): 473 (MH⁺, 25), 91 (100). Anal.

(C₂₉H₄₄O₅) C, H.

General Procedure for Removal of the Benzyl Ether Group.

A solution of BCl₃(3 equiv, 1 M in CH₂Cl₂) was added to a -78

°C stirring solution of VIII (1 equiv) in CH₂Cl₂(10 mL/mmol),

and the reaction was monitored by TLC. Upon completion, the

reaction mixture was quenched with saturated aqueous NaHCO₃,

warmed to room temperature, and extracted with CH₂Cl₂(2×).

Concentration in vacuo followed by purification by silica gel

chromatography with EtOAc/hexanes gave IX.

General Procedure for Removing the PMP Group. Ceric

ammonium nitrate (CAN, 3 equiv) was added to a 0 °C stirring

solution of VI (1 equiv) in CH₃CN/H₂O (4:1, 9 mL/mmol). After

15 min, the reaction was quenched with aqueous saturated NaHCO₃,

extracted with EtOAc, dried over MgSO₄, and concentrated in

vacuo. Purification by silica gel chromatography with hexanes/

EtOAc (2:1) gave VII, which was carried through to the next step

without further characterization.

Z-5-(Hydroxymethyl)-3-(3-methylbutylidene)-5-[(phenyl-

methoxy)methyl]-4,5-dihydrofuran-2-one (Z-VII, R ) CH₂CH-

(CH₃)₂). According to the general procedure, Z-VI (R ) CH₂CH-

(CH₃)₂, 260 mg, 0.63 mol) was combined with CAN (1.0 mg, 1.89

mmol) to give Z-VII (R ) CH₂CH(CH₃)₂, 196 mg, 100%) as a

colorless oil. ¹H NMR (CDCl₃) δ 7.27-7.36 (m, 5 H, PhCH₂OCH₂),

6.20 (tt, J ) 7.8, 2.3 Hz, 1 H, CdCHCH₂CH(CH₃)₂), 4.55 (AB q,

2 H, J ) 11.9 Hz, PhCH₂OCH₂), 3.66 and 3.74 (dd, J ) 12.1, 6.4

Hz, 2 H, HOCH₂), 3.56 (AB q, 2 H, J ) 10.1 Hz, PhCH₂OCH₂),

2.82 (app q, 2 H, CdCHCH₂CH(CH₃)₂), 2.53-2.67 (m, 2 H, H-4_ab),

2.41 (t, J ) 6.6 Hz, 2 H, HOCH₂), 1.81 (sept, J ) 6.7 Hz, 1 H,

CdCHCH₂CH(CH₃)₂), 0.92 and 0.91 (d, J ) 6.7 Hz, 6 H, Cd

CHCH₂CH(CH₃)₂).

E-5-(Hydroxymethyl)-3-(3-methylbutylidene)-5-[(phenyl-

methoxy)methyl]-4,5-dihydrofuran-2-one (E-VII, R ) CH₂CH-

(CH₃)₂). According to the general procedure, E-VI (R ) CH₂CH-

(CH₃)₂, 512 mg, 0.1.2 mol) was combined with CAN (1.9 mg, 3.6

mmol) to give E-VII (R ) CH₂CH(CH₃)₂, 392 mg, 100%) as a

white solid: mp 44-45 °C; ¹H NMR (CDCl₃) δ 7.24-7.34 (m, 5

H, PhCH₂OCH₂), 6.72 (tt, J ) 7.7, 2.9 Hz, 1 H, CdCHCH₂CH-

(CH₃)₂), 4.54 (s, 2 H, PhCH₂OCH₂), 3.74 (dd, J ) 12.1, 6.5 Hz, 1

H, HOCHH), 3.64 (dd, J ) 12.1, 6.4 Hz, 2 H, HOCHH), 3.56 (AB

q, J ) 10.1 Hz, 2 H, PhCH₂OCH₂), 2.80 (t, J ) 6.5, 1 H, HOCH₂),

2.73 (m, 2 H, CdCHCH₂CH(CH₃)₂), 2.02-2.06 (m, 2 H, H-4_ab),

1.78 (sept, J ) 6.7 Hz, 1 H, CdCHCH₂CH(CH₃)₂), 0.92 and 0.91

(d, J ) 6.7 Hz, 6 H, CdCHCH₂CH(CH₃)₂).

Z-5-(Hydroxymethyl)-3-[5-methyl-3-(2-methylpropyl)hexyl-

idene]-5-[(phenylmethoxy)methyl]-4,5-dihydrofuran-2-one (Z-

VII, R ) CH₂CH(CH₂CH(CH₃)₂)₂). According to the general

procedure, Z-VI (R ) CH₂CH(CH₂CH(CH₃)₂)₂, 379 mg, 0.77 mol)

was combined with CAN (1.26 g, 2.3 mmol) to give Z-VII (R )

CH₂CH(CH₂CH(CH₃)₂)₂, 298 mg, 98%). This compound has been

previously reported.²¹

E-5-(Hydroxymethyl)-3-[5-methyl-3-(2-methylpropyl)hexyl-

idene]-5-[(phenylmethoxy)methyl]-4,5-dihydrofuran-2-one (E-

VII, R ) CH₂CH(CH₂CH(CH₃)₂)₂). According to the general

procedure, E-VI (R ) CH₂CH(CH₂CH(CH₃)₂)₂, 135 mg, 0.27 mol)

was combined with CAN (448 mg, 0.81 mmol) to give E-VII (R

) CH₂CH(CH₂CH(CH₃)₂)₂, 87 mg, 81%). This compound has been

previously reported.²¹

General Procedure for Acylation. A solution of VII (1 equiv)

in CH₂Cl₂(5 mL/mmol) was treated with Et₃N (3 equiv), acid

chloride (1.5 equiv), and catalytic DMAP. After being stirred for

10 min, the mixture was concentrated in vacuo and the residue

was purified by silica gel chromatography with EtOAc/hexanes (1:

7) to give VIII.

Z-{4-(3-Methylbutylidene)-5-oxo-2-[(phenylmethoxy)methyl]-

2-(2,3-dihydrofuryl)}methyl 5-Methyl-3-(2-methylpropyl)hex-

anoate (Z-VIII, R ) CH₂CH(CH₃)₂, R′ ) CH₂CH(CH₂CH-

(CH₃)₂)₂). According to the general procedure, Z-VII (R )

CH₂CH(CH₃)₂, 196 mg, 0.63 mmol) was combined with 5-methyl-

3-(2-methylpropyl)hexanoyl chloride (128 mg, 0.63 mmol) to give

Z-VIII (R ) CH₂CH(CH₃)₂, R′ ) CH₂CH(CH₂CH(CH₃)₂)₂, 127

mg, 43%) as a colorless oil, which was contaminated with a small

amount of the E-isomer that was formed under the reaction

conditions. FAB-MS (m/z, relative intensity): 473 (MH⁺, 38), 91

(100). Anal. (C₂₉H₄₄O₅‚0.5H₂O) C, H.

Z-{2-(Hydroxymethyl)-4-[5-methyl-3-(2-methylpropyl)hexyl-

idene]-5-oxo-2-(2,3-dihydrofuryl)}methyl 3-Methylbutanoate (Z-

2). According to the general procedure, Z-VIII (R ) CH₂CH(CH₂-

CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂, 300 mg, 0.63 mol) was combined

with BCl₃(1.9 mL, 1.9 mmol) to give Z-2 (193 mg, 79%) as a

1

E-{4-(3-Methylbutylidene)-5-oxo-2-[(phenylmethoxy)methyl]-

2-(2,3-dihydrofuryl)}methyl 5-Methyl-3-(2-methylpropyl)hex-

anoate (E-VIII, R ) CH₂CH(CH₃)₂, R′ ) CH₂CH(CH₂CH-

(CH₃)₂)₂). According to the general procedure, E-VII (R )

colorless oil. H NMR (CDCl₃) δ 6.24 (tt, J ) 7.5, 2.2 Hz, 1H,

CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 4.21 (AB q, J ) 11.9 Hz, 2 H,

(CH₃)₂CHCH₂CO₂CH₂), 3.69 (dd, J ) 12.1, 6.8 Hz, 1 H, CHHOH),

3.62 (dd, J ) 12.2, 6.2 Hz, 1 H, HOCHHOH), 2.90 (dq, J ) 16.2,

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3477

2.2 Hz, 1 H, H-3_a), 2.63-2.76 (m, 3 H, H-3_band CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂), 2.47 (br t, J ) 6.3 Hz, 1 H, HOCH₂), 2.21 (d,

J ) 6.9 Hz, 2 H, (CH₃)₂CHCH₂CO₂CH₂), 2.03-2.12 (m, 1 H,

(CH₃)₂CHCH₂CO₂CH₂), 1.57-1.69 (m, 3 H, CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂), 1.08 (td, J ) 7.0, 2.3 Hz, 4 H, CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂), 0.94 (d, J ) 6.6 Hz, 6 H, (CH₃)₂CHCH₂CO₂-

CH₂), 0.84 and 0.85 (d, J ) 2.3 Hz, 12 H, CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂). FAB-MS (m/z, relative intensity): 383 (MH⁺,

72), 57 (100). Anal. (C₂₂H₃₈O₅) C, H.

¹H NMR (CDCl₃) δ 7.27-7.34 (m, 5 H, PhCH₂OCH₂), 6.81-6.88

(m, 4 H, MeOC₄H₄OCH₂), 4.60-4.61 (m, 2 H, PhCH₂OCH₂),

4.44-4.53 (m, 1 H, H-4), 4.23-4.30 (m, 1 H, MeOC₄H₄OCHH),

3.98-4.05 (m, 3 H, MeOC₄H₄OCHH and H-5_a,b), 3.77 (s, 3 H,

MeOC₄H₄OCH₂), 3.51-3.68 (m, 4 H, CHCH₂Cl and PhCH₂OCH₂).

FAB-MS (m/z, relative intensity): 378 (M^+•, 61.3), (91, 100). Anal.

(C₂₀H₂₃ClO₅) C, H.

4-Methoxy-1-({4-methylene-2-[(phenylmethoxy)methyl](1,3-

dioxolan-2-yl)}methoxy)benzene (XI). A mixture of X (652 mg,

t

E-{2-(Hydroxymethyl)-4-[5-methyl-3-(2-methylpropyl)hexyl-

idene]-5-oxo-2-(2,3-dihydrofuryl)}methyl 3-Methylbutanoate (E-

2). According to the general procedure, E-VIII (R ) CH₂CH(CH₂-

CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂, 103 mg, 0.22 mol) was combined

with BCl₃(0.66 mL, 0.66 mmol) to give E-2 (60 mg, 73%) as a

1.72 mmol) and BuOK (386 mg, 3.44 mmol) in THF (30 mL)

was refluxed for 19 h. After cooling to room temperature, the

mixture was washed with water, dried over MgSO₄, and concen-

trated in vacuo. Purification by silica gel chromatography (hexanes/

EtOAc, 15:1) gave XI (472 mg, 80%) as a colorless oil. ¹H NMR

(CDCl₃) δ 7.27-7.32 (m, 5H, PhCH₂OCH₂), 6.81-6.90 (m, 4 H,

MeOC₄H₄OCH₂), 4.62-4.63 (m, 4 H, PhCH₂OCH₂and MeOC₄H₄-

OCH₂), 4.44-4.45 (m, 1 H, CdCHH), 4.10 (s, 2 H, PhCH₂OCH₂),

3.94-3.96 (m, 1 H, CdCHH), 3.77 (s, 3 H, MeOC₄H₄OCH₂), 3.72

(AB d, J ) 1.0 Hz, 2 H, H-5_a,b). FAB-MS (m/z, relative intensity):

343 (MH⁺, 8), 91 (100). Anal. (C₂₀H₂₂O₅) C, H.

1

colorless oil. H NMR (CDCl₃) δ 6.78 (tt, J ) 7.5, 2.8 Hz, 1 H,

CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 4.22 (AB q, J ) 11.9 Hz, 2 H,

(CH₃)₂CHCH₂CO₂CH₂), 3.67 (AB q, J ) 12.1 Hz, 2 H, HOCH₂),

2.61-2.83 (m, 3 H, HOCH₂and H-3_a,b), 2.20 (d, J ) 6.8 Hz, 2 H,

(CH₃)₂CHCH₂CO₂CH₂), 2.02-2.13 (m, 3 H, (CH₃)₂CHCH₂CO₂-

CH₂and CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 1.71 (p, J ≈ 6.8 Hz, 1

H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 1.60 (sept, J ≈ 6.9 Hz, 2 H,

CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 1.05-1.12 (m, 4 H, CdCHCH₂-

CH(CH₂CH(CH₃)₂)₂), 0.95 (d, J ) 6.6 Hz, 6 H, (CH₃)₂CHCH₂-

CO₂CH₂), 0.86 (d, J ) 6. 6 Hz, 12 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂). FAB-MS (m/z, relative intensity): 383 (MH⁺, 100). Anal.

(C₂₂H₃₈O₅) C, H.

2-[(4-Methoxyphenoxy)methyl]-2-[(phenylmethoxy)methyl]-

1,3-dioxolan-4-one (XII). O₃was bubbled into a solution of XI

(3.38 g, 9.87 mmol) in CH₂Cl₂/MeOH (20 mL/20 mL) for 10 min

and then degassed with N₂at -78 °C. To this was added dimethyl

sulfide (1.45 mL, 19.7 mmol) at -78 °C, and the mixture was stirred

at room temperature for 2 h. The resulting mixture was concentrated

in vacuo, and the residue was dissolved in CH₂Cl₂, washed with

brine, dried over MgSO₄, and concentrated in vacuo. Purification

by silica gel chromatography (hexanes/EtOAc, 4:1) gave XII (2.42

g, 71%) as a colorless oil. ¹H NMR (CDCl₃) δ 7.29-7.38 (m, 5H,

PhCH₂OCH₂), 6.80-6.86 (m, 4 H, MeOC₄H₄OCH₂), 4.63 (s, 2 H,

PhCH₂OCH₂), 4.46 (AB q, J ) 14.6 Hz, 2 H, PhCH₂OCH₂), 4.17

(AB q, J ) 10.7 Hz, 2 H, MeOC₄H₄OCH₂), 3.80 (AB q, J ) 11.1

Hz, H-5_a,b), 3.77 (s, 3 H, MeOC₄H₄OCH₂). FAB-MS (m/z, relative

intensity): 383 (M + K⁺, 8), 344 (MH^•+, 76), 91 (100). HRMS

(FAB) calcd for C₁₉H₂₀O₆(MH⁺): 34.1260. Found: 344.1266.

2-[(4-Methoxyphenoxy)methyl]-5-(3-methylbutylidene)-2-

[(phenylmethoxy)methyl]-1,3-dioxolan-4-one (XIV, R ) CH₂CH-

(CH₃)₂). A solution of XII (1.30 g, 3.8 mmol) in THF (4 mL) was

added dropwise to a solution of LDA (18 mL, 2 M in THF/hexanes,

8:1) at -78 °C. After 30 min, isovaleraldehyde (0.61 mL, 5.6 mmol)

was added and the mixture was stirred at -78 °C for 2 h. The

reaction was then quenched with aqueous saturated NH₄Cl, and

the mixture was extracted with Et₂O (2 × 100 mL). The combined

organic layer was dried over MgSO₄and concentrated in vacuo to

a yellow oil. The residual oil was then taken up in CH₂Cl₂(30

mL) and treated sequentially with Et₃N (2.2 mL, 15.8 mmol) and

MsCl (0.6 mL, 7.9 mmol) at 0 °C. After being stirred at 0 °C for

30 min, the mixture warmed to room temperature and stirred for 3

h. The mixture was then recooled to 0 °C, treated with dropwise

addition of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU, 3 mL, 19.8

mmol), and allowed to warm to room temperature overnight. The

mixture was then concentrated in vacuo. The residue was filtered

through a short silica gel pad, washing with 50% EtOAc in hexanes,

and the filtrate was concentrated in vacuo. Purification by silica

gel chromatography (hexanes/EtOAc, 7:1) gave XIV (R ) CH₂-

CH(CH₃)₂, 238 mg, 15%) as an equal mixture of Z- and E-isomers.

¹H NMR (CDCl₃) δ 7.28-7.37 (m, 5H, PhCH₂OCH₂), 6.79-6.85

(m, 4 H, MeOC₄H₄OCH₂), 5.67 (t, J ) 7.9 Hz, 0.5 H, CdCHCH₂-

CH(CH₃)₂), 5.63 (t, J ) 8.4 Hz, 0.5 H, CdCHCH₂CH(CH₃)₂), 4.64

(s, 2 H, PhCH₂OCH₂), 4.19 (m, 2 H, MeOC₄H₄OCH₂), 3.80-

3.81 (m, 2 H, PhCH₂OCH₂), 3.77 (s, 3 H, MeOC₄H₄OCH₂), 2.40-

2.53 (m, 1 H, CdCHCH₂CH(CH₃)₂), 2.09 (dd, J ) 7.9, 6.9 Hz, 1

H, CdCHCH₂CH(CH₃)₂), 1.65-1.78 (m, 1 H, CdCHCH₂CH-

(CH₃)₂), 0.91-0.94 (m, 6 H, CdCHCH₂CH(CH₃)₂). FAB-MS (m/

z, relative intensity): 412 (M^•+, 100). Anal. (C₂₄H₂₈O₆) C, H.

2-[(4-Methoxyphenoxy)methyl]-5-[5-methyl-3-(2-methylprop-

yl)hexylidene]-2-[(phenylmethoxy)methyl]-1,3-dioxolan-4-one

(XIV, R ) CH₂CH(CH₂CH(CH₃)₂)₂). A solution of XII (1.84 g,

5.34 mmol) in THF (5 mL) was added dropwise to a solution of

LDA (24 mL, 2 M in THF/hexanes) at -78 °C. After 30 min, a

Z-[2-(Hydroxymethyl)-4-(3-methylbutylidene)-5-oxo-2-(2,3-di-

hydrofuryl)]methyl 5-Methyl-3-(2-methylpropyl)hexanoate (Z-

3). According to the general procedure, Z-VIII (R ) CH₂CH(CH₃)₂,

R′ ) CH₂CH(CH₂CH(CH₃)₂)₂, 127 mg, 0.27 mol) was combined

with BCl₃(0.81 mL, 0.81 mmol) to give Z-3 (72 mg, 67%) as a

1

colorless oil. H NMR (CDCl₃) δ 6.26 (tt, J ) 7.8, 2.3 Hz, 1 H,

CdCHCH₂CH(CH₃)₂), 4.20 (AB q, J ) 11.9 Hz, 2 H, ((CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 3.66 (AB q, J ) 12.2 Hz, 2H, HOCH₂),

2.90 (dq, J ≈ 16.4, 4.5, 2.2 Hz, 1 H, H-3_a), 2.72 (dq, J ≈ 16.4,

4.1, 2.0 Hz, 1 H, H-3_b), 2.59-2.64 (m, 2 H, CdCHCH₂CH(CH₃)₂),

2.23-2.25 (m, 2 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 1.89-2.02

(m, 1 H, CdCHCH₂CH(CH₃)₂), 1.68-1.78, (m, 1 H, ((CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 1.56-1.68 (m, 2 H, ((CH₃)₂CHCH₂)₂-

CHCH₂CO₂CH₂), 1.02-1.20 (m, 4 H, ((CH₃)₂CHCH₂)₂CHCH₂-

CO₂CH₂), 0.93-0.95 (d, J ) 6.7 Hz, 6 H, CdCHCH₂CH(CH₃)₂),

0.86-0.89 (m, 12 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂). FAB-MS

(m/z, relative intensity): 383 (MH⁺, 100). Anal. (C₂₂H₃₈O₅) C, H.

E-[2-(Hydroxymethyl)-4-(3-methylbutylidene)-5-oxo-2-(2,3-di-

hydrofuryl)]methyl 5-Methyl-3-(2-methylpropyl)hexanoate (E-

3). According to the general procedure, E-VIII (R ) CH₂CH(CH₃)₂,

R′ ) CH₂CH(CH₂CH(CH₃)₂)₂, 260 mg, 0.56 mol) was combined

with BCl₃(1.7 mL, 1.7 mmol) to give E-3 (130 mg, 60%) as a

pale-yellow oil. ¹H NMR (CDCl₃) δ 6.78 (tt, J ) 7.7, 2.9 Hz, 1 H,

CdCHCH₂CH(CH₃)₂), 4.21 (AB q, J ) 11.9 Hz, 2 H, ((CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 3.68 (AB q, J ) 12.1 Hz, 2 H, HOCH₂),

2.82 (dm, J ) 17.1 Hz, 1 H, H-3_a), 2.65 (dm, J ) 17.1 Hz, 1 H,

H-3_b), 2.39 (br s, 1 H, HOCH₂), 2.22-2.24 (m, 2 H, ((CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 2.04-2.09 (m, 2 H, CdCHCH₂CH-

(CH₃)₂), 1.93 (app quint, 1 H, CdCHCH₂CH(CH₃)₂), 1.82, (sept,

1 H, J ) 6.6 Hz, 1 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 1.60 (app

sept, 2 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 1.03-1.18 (m, 4 H,

((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 0.95 (d, J ) 6.7 Hz, 6 H, Cd

CHCH₂CH(CH₃)₂), 0.85-0.89 (m, 12 H, ((CH₃)₂CHCH₂)₂CHCH₂-

CO₂CH₂). FAB-MS (m/z, relative intensity): 383 (MH⁺, 100). Anal.

(C₂₂H₃₈O₅) C, H.

1-({4-(Chloromethyl)-2-[(phenylmethoxy)methyl](1,3-dioxolan-

2-yl)}methoxy)-4-methoxybenzene (X). 3-Chloro-1,2-propanediol

(0.84 mL, 10.00 mmol) was added to a solution of II (1.43 g, 5.00

mmol) and p-TsOH‚H₂O (200 mg, 1.05 mmol) in benzene (100

mL). The mixture was refluxed for 16 h with azeotropic removal

of produced water. After cooling to room temperature, the mixture

was diluted with EtOAc and washed sequentially with saturated

NaHCO₃and brine. The organic layer was dried over MgSO₄and

concentrated in vacuo. Purification by silica gel chromatography

(hexanes/EtOAc, 15:1) gave X (1.29 g, 68%) as a pale-yellow oil.

3478 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

solution of 5-methyl-3-(2-methylpropyl)hexanal (1.82 g, 10.7 mmol)

was added and the mixture was stirred at -78 °C. After the mixture

was stirred at -78 °C for 2 h, the reaction was quenched with

aqueous saturated NH₄Cl and extracted with Et₂O (2 × 100 mL).

The combined organic layer was dried over MgSO₄and concen-

trated in vacuo to a yellow oil. The residual oil was then taken up

in CH₂Cl₂(30 mL) and treated sequentially with Et₃N (3 mL, 21.4

mmol) and MsCl (1.2 mL, 16.0 mmol) at 0 °C. After being stirred

at 0 °C for 30 min, the mixture was warmed to room temperature

and stirred for 3 h. The resultant solution was then recooled to 0

°C, treated with dropwise addition of DBU (8 mL, 53.4 mmol),

and allowed to warm to room temperature overnight. The mixture

was then concentrated in vacuo. The residue was filtered through

a short silica gel pad, washing with 50% EtOAc in hexanes, and

the filtrate was concentrated in vacuo. Purification by silica gel

chromatography (hexanes/EtOAc, 7:1) gave XIV (R ) CH₂CH-

(CH₂CH(CH₃)₂)₂, 981 mg, 37%) as a mixture of E- and Z-isomers.

¹H NMR (CDCl₃) δ 7.28-7.37 (m, 5H, PhCH₂OCH₂), 6.79-6.85

(m, 4 H, MeOC₄H₄OCH₂), 5.60-5.68 (m, 1 H, CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂), 4.64 (s, 2 H, PhCH₂OCH₂), 4.19 (d, J ) 6.3

Hz, 2 H, MeOC₄H₄OCH₂), 3.79-3.81 (m, 2 H, PhCH₂OCH₂), 3.76

(s, 3 H, MeOC₄H₄OCH₂), 2.46-2.61 (m, 1 H, CdCHCH₂CH(CH₂-

CH(CH₃)₂)₂), 2.16 (dd, J ) 7.9, 5.7 Hz, 1 H, CdCHCH₂CH(CH₂-

CH(CH₃)₂)₂), 1.56-1.71 (m, 3 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂),

1.06-1.11 (m, 4 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 0.82-0.87

(m, 12 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂). FAB-MS (m/z, relative

intensity): 496 (M^•+, 56). Anal. (C₃₀H₄₀O₆) C, H.

aqueous saturated NaHCO₃(1×), dried over MgSO₄, and concen-

trated in vacuo. Purification by silica gel chromatography (hexanes/

EtOAc, 20:1) gave an equal mixture of geometric isomers XVI (R

) CH₂CH(CH₃)₂, R′ ) CH₂CH(CH₂CH(CH₃)₂)₂, 114 mg, 46%)

1

as a yellow oil. H NMR (CDCl₃) δ 7.28-7.37 (m, 5H, PhCH₂-

OCH₂), 5.66 (t, J ) 8.0 Hz, 0.5 H, CdCHCH₂CH(CH₃)₂), 5.62 (t,

J ) 8.5 Hz, 0.5 H, CdCHCH₂CH(CH₃)₂), 4.61 (s, 2 H, PhCH₂-

OCH₂), 4.29-4.39 (m, 2 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂),

3.65-3.72 (m, 2 H, PhCH₂OCH₂), 2.37-2.51 (m, ∼1 H, Cd

CHCH₂CH(CH₃)₂), 2.20-2.26 (m, ∼1 H, ((CH₃)₂CHCH₂)₂CHCH₂-

CO₂CH₂), 2.08-2.11 (app t, 1 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂-

CH₂), 1.89-1.99 (m, 1 H, CdCHCH₂CH(CH₃)₂), 1.54-1.79 (m,

3 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 1.02-1.17 (m, 4 H,

((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 0.85-0.93 (18 H, CdCHCH₂-

CH(CH₃)₂and ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂). FAB-MS (m/z,

relative intensity): 475 (MH⁺, 2), (91, 100). Anal. (C₂₈H₄₂O₆) C,H.

{5-[5-Methyl-3-(2-methylpropyl)hexylidene]-4-oxo-2-[(phenyl-

methoxy)methyl]-1,3-dioxolan-2-yl}methyl 2,2-Dimethylpro-

panoate (XVI, R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) C(CH₃)₃).

DMAP (2 mg, 0.013 mmol), Et₃N (0.04 mL, 0.26 mmol), and

pivaloyl chloride (0.02 mL, 0.16 mmol) were added sequentially

to a 0 °C solution of XV (R ) CH₂CH(CH₂CH(CH₃)₂)₂, 51 mg,

0.13 mmol) in CH₂Cl₂(5 mL). After being stirred for 2 h, the

mixture was washed with aqueous saturated NaHCO₃(1×) and

brine (1×), dried over MgSO₄, and concentrated in vacuo. Purifica-

tion by silica gel chromatography (hexanes/EtOAc 20:1) gave XVI

(R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) C(CH₃)₃, 51 mg, 82%) as a

colorless oil that was used without further purification in the next

step. ¹H NMR (CDCl₃) δ 7.28-7.37 (m, 5 H, PhCH₂OCH₂), 5.65

(t, J ) 7.8 Hz, 0.5 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 5.62 (t, J

) 8.3 Hz, 0.5 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 4.62 (s, 2 H,

PhCH₂OCH₂), 4.37 (AB q, J ) 12.3 Hz, 1 H, CH₃C(O)OCH₂),

4.34 (s, 1 H, CH₃C(O)OCH₂), 3.67-3.69 (m, 2 H, PhCH₂OCH₂),

2.49-2.53 (m, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 2.15-2.19

(m, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 1.54-1.71 (m, 3 H, Cd

CHCH₂CH(CH₂CH(CH₃)₂)₂), 1.17 and 1.18 (s, 9 H, CH₃C(O)-

OCH₂), 1.08 (td, J ) 7.0, 2.2 Hz, 4 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂), 0.84-0.86 (m, 12 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂).

FAB-MS (m/z, relative intensity): 513 (MH⁺, 8), 475 (MH⁺, 2),

91 (100).

{5-[5-Methyl-3-(2-methylpropyl)hexylidene]-4-oxo-2-[(phenyl-

methoxy)methyl]-1,3-dioxolan-2-yl}methyl 3-Methylbutanoate

(XVI, R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂).

Isovaleryl chloride (0.05 mL, 0.41 mmol) was added dropwise to

a 0 °C stirring solution of DMAP (5 mg, 0.034 mmol), pyridine

(0.05 mL, 0.68 mmol), and XV (R ) CH₂CH(CH₂CH(CH₃)₂)₂) in

CH₂Cl₂(10 mL). The ice bath was removed and the reaction mixture

warmed to room temperature. After being stirred at room temper-

ature for 24 h, the mixture was washed with aqueous saturated

NaHCO₃(1×) and brine (1×), dried over MgSO₄, and concentrated

in vacuo. Purification by silica gel chromatography gave XVI (R

) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂, 135 mg, 84%)

as a pale-yellow oil consisting of predominantly a single isomer.

¹H NMR (CDCl₃) δ 7.29-7.38 (m, 5H, PhCH₂OCH₂), 5.65 (t, J

) 7.9 Hz, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂, 4.61 (s, 2 H,

PhCH₂OCH₂), 4.30 (AB q, J ) 12.3 Hz, 2 H, PhCH₂OCH₂), PhCH₂-

OCH₂), 3.65-3.71 (m, 2 H, (CH₃)₂CHCH₂CO₂CH₂), 2.15-2.21

(m, 4 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂) and (CH₃)₂CHCH₂CO₂-

CH₂), 2.00-2.10 (m, 1 H, (CH₃)₂CHCH₂CO₂CH₂), 1.57-1.71 (m,

3 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 1.08 (td, J ) 7.1, 1.4 Hz, 1

H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 0.93 and 0.94 (s, 6 H, (CH₃)₂-

CHCH₂CO₂CH₂), 0.84 and 0.86 (s, 12 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂). FAB-MS (m/z, relative intensity): 475 (MH⁺, 2), (91,

100). Anal. (C₂₈H₄₂O₆) C, H

2-(Hydroxymethyl)-5-(3-methylbutylidene)-2-[(phenylmethoxy)-

methyl]-1,3-dioxolan-4-one (XV, R ) CH₂CH(CH₃)₂). Am-

monium cerium(IV) nitrate (921 mg, 1.7 mmol) was added to a

solution of XIV (R ) CH₂CH(CH₃)₂, 231 mg, 0.56 mmol) in CH₃-

CN (16 mL) and H₂O (4 mL) at 0 °C, and the mixture was stirred

for 30 min. The reaction mixture was then diluted with CH₂Cl₂

and washed with H₂O. The organic layer was dried over MgSO₄

and concentrated in vacuo. Purification by silica gel chromatography

(hexanes/EtOAc, 6:1) gave XV (R ) (CH₂CH(CH₃)₂, 156 mg, 91%)

1

as a yellow oil. H NMR (CDCl₃) δ 7.28-7.37 (m, 5 H, PhCH₂-

OCH₂), 5.65 (app q, J ≈ 8.3 Hz, 1 H, CdCHCH₂CH(CH₃)₂)), 4.61

(s, 2 H, PhCH₂OCH₂), 3.82-3.84 (m, 2 H, HOCH₂), 3.67-3.74

(m, 2 H, PhCH₂OCH₂), 2.38-2.51 (m, ∼1 H, CdCHCH₂CH-

(CH₃)₂), 2.10-2.14 (m, ca. 1 H, CdCHCH₂CH(CH₃)₂), 1.99-2.02

(br m,1 H, HOCH₂), 1.63-1.80 (m, 1 H, CdCHCH₂CH(CH₃)₂),

0.90-0.94 (m, 6 H, CdCHCH₂CH(CH₃)₂). FAB-MS (m/z, relative

intensity): 306 (M^•+, 9.7). Anal. (C₁₇H₂₂O₅) C, H.

2-(Hydroxymethyl)-5-[5-methyl-3-(2-methylpropyl)hexylidene]-

2-[(phenylmethoxy)methyl]-1,3-dioxolan-4-one (XV, R ) CH₂CH-

(CH₂CH(CH₃)₂)₂). Ammonium cerium(IV) nitrate (375 mg, 0.684

mmol) was added to a solution of XIV (R ) CH₂CH(CH₂CH-

(CH₃)₂)₂, 113 mg, 0.228 mmol) in CH₃CN (8 mL) and H₂O (2 mL)

at 0 °C, and the mixture was stirred for 30 min. The reaction mixture

was then diluted with CH₂Cl₂and washed with H₂O. The organic

layer was dried over MgSO₄and concentrated in vacuo. Purification

by silica gel chromatography (hexanes/EtOAc, 6:1) gave XV (R

) CH₂CH(CH₂CH(CH₃)₂)₂, 47 mg, 54%) as a pale-yellow oil that

was not further purified and was used immediately in the following

1

step. H NMR (CDCl₃) δ 7.28-7.37 (m, 5H, PhCH₂OCH₂), 5.63

(app q, J ≈ 8.2 Hz, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 4.61 (2

s, 2 H, PhCH₂OCH₂), 3.83-3.85 (m, 2 H, HOCH₂), 3.69-3.71

(m, 2 H, PhCH₂OCH₂), 2.51 (m, ∼1 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂), 2.17-2.20 (m, ca. 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂),

1.95 (br s, 1 H, HOCH₂), 1.55-1.57 (m, 3 H, CdCHCH₂CH-

(CH₂CH(CH₃)₂)₂), 1.08 (app t, 2 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂), 0.84-0.86 (m, 12 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂).

{5-(3-Methylbutylidene)-4-oxo-2-[(phenylmethoxy)methyl]-

1,3-dioxolan-2-yl}methyl 5-Methyl-3-(2-methylpropyl)hexanoate

(XVI, R ) CH₂CH(CH₃)₂, R′ ) CH₂CH(CH₂CH(CH₃)₂)₂). Et₃N

(0.14 mL, 1.02 mmol) and 5-methyl-3-(2-methylpropyl)hexanoyl

chloride (0.02 mL, 0.16 mmol) were added sequentially to a 0 °C

stirring solution of XV (R ) CH₂CH(CH₃)₂) (156 mg, 0.51 mmol)

in CH₂Cl₂(20 mL). After the mixture was allowed to warm to room

temperature and was stirred for 4 h, the mixture was washed with

{2-(Hydroxymethyl)-5-[5-methyl-3-(2-methylpropyl)hexylidene]-

4-oxo-1,3-dioxolan-2-yl}methyl 2,2-Dimethylpropanoate (Z,E-

4). BCl₃(0.3 mL, 0.32 mmol, 1 M in CH₂Cl₂) was added to a -78

°C solution of XVI (R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) C(CH₃)₃,

51 mg, 0.11 mmol) in CH₂Cl₂(5 mL), and the mixture was stirred

for 2 h. The reaction was quenched with aqueous saturated

NaHCO₃, extracted with CH₂Cl₂(2×), dried over MgSO₄, and

DAG-dioxolanones

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3479

concentrated in vacuo. Purification by silica gel chromatography

Supporting Information Available: ¹H NMR spectra and

combustion analysis results. This material is available free of charge

via the Internet at http://pubs.acs.org.

(hexanes/EtOAc, 6:1) gave Z,E-4 (35.2 mg, 86% as a 1:1 mixture

1

of geometric isomers) as a pale-yellow oil. H NMR (CDCl₃) δ

5.67 (t, J ) 7.9 Hz, 0.5 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 5.64

(t, J ) 8.3 Hz, 0.5 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 4.29-4.41

(m, 2 H, CH₃C(O)OCH₂), 3.80 (br t, 2 H, HOCH₂), 2.50 (dd, J )

8.3, 6.0 Hz, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 2.17 (dd, J )

7.8, 5.8 Hz, 1 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂), 2.14 (v br s, 1

H, HOCH₂), 1.55-1.70 (m, 3 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂),

1.18 and 1.19 (s, 9 H, OC(O)CH₃), 1.06-1.10 (m, 4 H, CdCHCH₂-

CH(CH₂CH(CH₃)₂)₂), 0.84-0.87 (m, 12 H, CdCHCH₂CH(CH₂-

CH(CH₃)₂)₂). FAB-MS (m/z, relative intensity): 385 (MH⁺, 23),

(57, 100). Anal. (C₂₁H₃₆O₆) C, H.

References

(1) Nishizuka, Y. Intracellular signaling by hydrolysis of phospholipids

and activation of protein-kinase-C. Science 1992, 258, 607-614.

(2) Hofmann, J. Protein kinase C isozymes as potential targets for

anticancer therapy. Curr. Cancer Drug. Targets 2004, 4, 125-146.

(3) Garcia-Bermejo, M. L.; Leskow, F. C.; Fujii, T.; Wang, Q.; Blumberg,

P. M.; Ohba, M.; Kuroki, T.; Han, K. C.; Lee, J.; Marquez, V. E.;

Kazanietz, M. G. Diacylglycerol (DAG)-lactones, a new class of

protein kinase C (PKC) agonists, induce apoptosis in LNCaP prostate

cancer cells by selective activation of PKCalpha. J. Biol. Chem. 2002,

277, 645-655.

(4) Mochly-Rosen, D.; Kauvar, L. M. Pharmacological regulation of

network kinetics by protein kinase C localization. Semin. Immunol.

2000, 12, 55-61.

(5) Newton, A. C. Protein kinase C: structural and spatial regulation by

phosphorylation, cofactors, and macromolecular interactions. Chem.

ReV. 2001, 101, 2353-2364.

(6) Parker, P. J.; Murray-Rust, J. PKC at a glance. J. Cell Sci. 2004,

117, 131-132.

(7) Ono, Y.; Fujii, T.; Igarashi, K.; Kuno, T.; Tanaka, C.; Kikkawa, U.;

Nishizuka, Y. Phorbol ester binding to protein kinase-C requires a

cysteine-rich zinc-finger-like sequence. Proc. Natl. Acad. Sci. U.S.A.

1989, 86, 4868-4871.

(8) Ron, D.; Kazanietz, M. G. New insights into the regulation of protein

kinase C and novel phorbol ester receptors. FASEB J. 1999, 13,

1658-1676.

(9) Rozengurt, E.; Sinnett-Smith, J.; Zugaza, J. Protein kinase D: a novel

target for diacylglycerol and phorbol esters. Biochem. Soc. Trans.

1997, 25, 565-571.

(10) Ahmed, S.; Lee, J.; Kozma, R.; Best, A.; Monfries, C.; Lim, L. A

novel functional target for tumor-promoting phorbol esters and

lysophosphatidic acid. The P21rac-Gtpase activating protein N-

chimaerin. J. Biol. Chem. 1993, 268, 10709-10712.

{2-(Hydroxymethyl)-5-[5-methyl-3-(2-methylpropyl)hexylidene]-

4-oxo-1,3-dioxolan-2-yl}methyl 3-Methylbutanoate (Z,E-5). BCl₃

(1.4 mL, 1.4 mmol, 1 M in CH₂Cl₂) was added to a -78 °C solution

of XVI (R ) CH₂CH(CH₂CH(CH₃)₂)₂, R′ ) CH₂CH(CH₃)₂, 135

mg, 0.28 mmol) in CH₂Cl₂(10 mL), and the mixture was stirred

for 2 h. The reaction was quenched with aqueous saturated

NaHCO₃, extracted with CH₂Cl₂(2×), dried over MgSO₄, and

concentrated in vacuo. Purification by silica gel chromatography

(hexanes/EtOAc, 5:1) gave Z,E-5 (87 mg, 80% as a single geometric

isomer) as a pale-yellow oil that solidified upon standing: mp 50-

1

51 °C; H NMR (CDCl₃) δ 5.68 (t, J ) 7.9 Hz, 1 H, CdCHCH₂-

CH(CH₃)₂), 4.36 (AB q, J ) 12.4 Hz, 2 H, ((CH₃)₂CHCH₂CO₂CH₂),

3.82 (br AB d, 2 H, HOCH₂), 2.21-2.23 (m, 2 H, (CH₃)₂CHCH₂-

CO₂CH₂), 2.17 (dd, J ) 7.9, 5.7 Hz, 2 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂), 2.07 (p, J ) 6.7, Hz, 1 H, (CH₃)₂CHCH₂CO₂CH₂), 1.98

(v br t, 1 H, HOCH₂), 1.57-1.72 (m, 3 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂), 1.09 (br t, J ) 7.0 Hz, 4 H, CdCHCH₂CH(CH₂CH-

(CH₃)₂)₂), 0.95 (d, J ) 6.7 Hz, 6 H, (CH₃)₂CHCH₂CO₂CH₂), 0.85-

0.87 (m, 12 H, CdCHCH₂CH(CH₂CH(CH₃)₂)₂). FAB-MS (m/z,

relative intensity): 385 (MH⁺, 18), 85 (100). Anal. (C₂₁H₃₆O₆) C,

H.

(11) Brose, N.; Rosenmund, C.; Rettig, J. Regulation of transmitter release

by Unc-13 and its homologues. Curr. Opin. Neurobiol. 2000, 10,

303-311.

(12) Ebinu, J. O.; Bottorff, D. A.; Chan, E. Y. W.; Stang, S. L.; Dunn, R.

J.; Stone, J. C. RasGRP, a Ras guanyl nucleotide-releasing protein

with calcium- and diacylglycerol-binding motifs. Science 1998, 280,

1082-1086.

(13) Kawasaki, H.; Sprihtgett, G. M.; Toki, S.; Canales, J. J.; Harlan, P.;

Blumenstiel, J. P.; Chen, E. J.; Bany, I. A.; Mochizuki, N.; Ashbacher,

A.; Matsuda, M.; Housman, D. E.; Graybiel, A. M. A Rap guanine

nucleotide exchange factor enriched highly in the basal ganglia. Proc.

Natl. Acad. Sci. U.S.A. 1998, 95, 13278-13283.

(14) Luo, B.; Regier, D. S.; Prescott, S. M.; Topham, M. K. Diacylglycerol

kinases. Cell. Signalling 2004, 16, 983-989.

(15) van Blitterswijk, W. J.; Houssa, B. Properties and functions of

diacylglycerol kinases. Cell. Signalling 2000, 12, 595-605.

(16) Zhang, G. G.; Kazanietz, M. G.; Blumberg, P. M.; Hurley, J. H.

Crystal-structure of the Cys2 activator-binding domain of protein-

kinase C-Delta in complex with phorbol ester. Cell 1995, 81, 917-

924.

(17) Marquez, V. E.; Blumberg, P. M. Synthetic diacylglycerols (DAG)

and DAG-lactones as activators of protein kinase C (PK-C). Acc.

Chem. Res. 2003, 36, 434-443.

(18) Hamer, D. H.; Bocklandt, S.; McHugh, L.; Chun, T. W.; Blumberg,

P. M.; Sigano, D. M.; Marquez, V. E. Rational design of drugs that

induce human immunodeficiency virus replication. J. Virol. 2003,

77, 10227-10236.

[2-(Hydroxymethyl)-5-(3-methylbutylidene)-4-oxo-1,3-dioxolan-

2-yl]methyl 5-Methyl-3-(2-methylpropyl)hexanoate (Z,E-6). BCl₃

(1.2 mL, 1.1 mmol, 1 M in CH₂Cl₂) was added to a -78 °C solution

of XVI (R ) CH₂CH(CH₃)₂), R′ ) CH₂CH(CH₂CH(CH₃)₂)₂, 107

mg, 0.23 mmol) in CH₂Cl₂(10 mL), and the mixture was stirred

for 2 h. The reaction was quenched with aqueous saturated

NaHCO₃, extracted with CH₂Cl₂(2×), dried over MgSO₄, and

concentrated in vacuo. Purification by silica gel chromatography

(hexanes/EtOAc, 6:1) gave Z,E-6 (52 mg, 60% as a 3:2 mixture of

geometric isomers) as a pale-yellow oil. ¹H NMR (CDCl₃) δ 5.69

(t, J ) 8.0 Hz, 0.6 H, CdCHCH₂CH(CH₃)₂), 5.66 (t, J ) 8.6 Hz,

0.4 H, CdCHCH₂CH(CH₃)₂), 4.34 (s, 0.6 H, ((CH₃)₂CHCH₂)₂-

CHCH₂CO₂CH₂), 4.32 (AB q, J ) 12.4 Hz, 0.4 H, ((CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 3.82 (AB q, J ) 12.8 Hz, 0.6 H,

HOCH₂), 3.80 (s, 0.4 Hz, 0.4 H, HOCH₂), 2.45 (dd, J ) 8.6, 6.8

Hz, 1 H, CdCHCHHCH(CH₃)₂), 2.23-2.26 (m, 2 H, ((CH₃)₂-

CHCH₂)₂CHCH₂CO₂CH₂), 2.11 (dd, J ) 7.9, 6.9 Hz, 1 H, Cd

CHCHHCH(CH₃)₂), 1.90-1.99 (m, 1 H, CdCHCH₂CH(CH₃)₂),

1.53-1.81 (m, 3 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 1.03-1.18

(m, 4 H, ((CH₃)₂CHCH₂)₂CHCH₂CO₂CH₂), 0.93 (d, J ) 6.6 Hz, 6

H, CdCHCH₂CH(CH₃)₂), 0.86-0.89 (m, 12 H, ((CH₃)₂CHCH₂)₂-

CHCH₂CO₂CH₂). FAB-MS (m/z, relative intensity): 385 (MH⁺,

100), Anal. (C₂₁H₃₆O₆) C, H.

(19) Pu, Y.; Perry, N. A.; Yang, D.; Lewin, N. E.; Kedei, N.; Braun, D.

C.; Choi, S. H.; Blumberg, P. M.; Garfield, S. H.; Stone, J. C.; Duan,

D.; Marquez, V. E. A novel diacylglycerol-lactone shows marked

selectivity in vitro among C1 domains of protein kinase C (PKC)

isoforms alpha and delta as well as selectivity for RasGRP compared

with PKCalpha. J. Biol. Chem. 2005, 280, 27329-27338.

(20) Lee, J. Design and synthesis of bioisosteres of ultrapotent protein

kinase C (PKC) ligand, 5-acetoxymethyl-5-hydroxymethyl-3-alkyl-

idene tetrahydro-2-furanone. Arch. Pharmacal Res. 1998, 21, 452-

457.

(21) Choi, Y.; Kang, J. H.; Lewin, N. E.; Blumberg, P. M.; Lee, J.;

Marquez, V. E. Conformationally constrained analogues of diacyl-

glycerol. 19. Synthesis and protein kinase C binding affinity of

diacylglycerol lactones bearing an N-hydroxylamide side chain. J.

Med. Chem. 2003, 46, 2790-2793.

Acknowledgment. This research was supported by the

Intramural Research Program of the NIH, National Cancer

Institute, Center for Cancer Research. The authors thank Dr.

James A. Kelley from the Laboratory of Medicinal Chemistry

for mass spectral analysis and interpretation. This research has

been funded in part with federal funds from the National Cancer

Institute, National Institutes of Health, under Contract N01-CO-

12400. The content of this publication does not necessarily

reflect the views or policies of the Department of Health and

Human Services nor does mention of trade names, commercial

products, or organizations imply endorsement by the U.S.

Government.

3480 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15

Choi et al.

(22) Kang, J. H.; Benzaria, S.; Sigano, D. M.; Lewin, N. E.; Pu, Y. M.;

Peach, M. L.; Blumberg, P. M.; Marquez, V. E. Conformationally

constrained analogues of diacylglycerol. 26. Exploring the chemical

space surrounding the C1 domain of protein kinase C with DAG-

lactones containing aryl groups at the sn-1 and sn-2 positions. J. Med.

Chem. 2006, 49, 3185-3203.

(23) Wildman, S. A.; Crippen, G. M. Prediction of physicochemical

parameters by atomic contributions. J. Chem. Inf. Comput. Sci. 1999,

39, 868-873.

(24) Lee, J.; Han, K. C.; Kang, J. H.; Pearce, L. L.; Lewin, N. E.; Yan,

S.; Benzaria, S.; Nicklaus, M. C.; Blumberg, P. M.; Marquez, V. E.

Conformationally constrained analogues of diacylglycerol. 18. The

incorporation of a hydroxamate moiety into diacylglycerol-lactones

reduces lipophilicity and helps discriminate between sn-1 and sn-2

binding modes to protein kinase C (PK-C). Implications for isozyme

specificity. J. Med. Chem. 2001, 44, 4309-4312. Erratum: J. Med.

Chem. 2003, 46 (13), 2794.

(41) Hritz, J.; Ulicny, J.; Laaksonen, A.; Jancura, D.; Miskovsky, P.

Molecular interaction model for the C1B domain of protein kinase

C-gamma in the complex with its activator phorbol-12-myristate-

13-acetate in water solution and lipid bilayer. J. Med. Chem. 2004,

47, 6547-6555.

(42) Burns, D. J.; Bell, R. M. Protein-kinase-C contains 2 phorbol ester

binding domains. J. Biol. Chem. 1991, 266, 18330-18338.

(43) Zhu, J. W.; Hansen, H.; Su, L. H.; Shieh, H. L.; Riedel, H. Ligand

regulation of bovine protein-kinase-C-alpha response via either

cysteine-rich repeat of conserved region C1. J. Biochem. 1994, 115,

1000-1009.

(44) Medkova, M.; Cho, W. H. Interplay of C1 and C2 domains of protein

kinase C-alpha in its membrane binding and activation. J. Biol. Chem.

1999, 274, 19852-19861.

(45) Giorgione, J.; Hysell, M.; Harvey, D. F.; Newton, A. C. Contribution

of the C1A and C1B domains to the membrane interaction of protein

kinase C. Biochemistry 2003, 42, 11194-11202.

(25) Kang, J. H.; Peach, M. L.; Pu, Y. M.; Lewin, N. E.; Nicklaus, M.

C.; Blumberg, P. M.; Marquez, V. E. Conformationally constrained

analogues of diacylglycerol (DAG). 25. Exploration of the sn-1 and

sn-2 carbonyl functionality reveals the essential role of the sn-1

carbonyl at the lipid interface in the binding of DAG-lactones to

protein kinase C. J. Med. Chem. 2005, 48, 5738-5748.

(26) Sigano, D. M.; Peach, M. L.; Nacro, K.; Choi, Y.; Lewin, N. E.;

Nicklaus, M. C.; Blumberg, P. M.; Marquez, V. E. Differential

binding modes of diacylglycerol (DAG) and DAG lactones to protein

kinase C (PK-C). J. Med. Chem. 2003, 46, 1571-1579.

(27) Kazanietz, M. G.; Wang, S.; Milne, G. W.; Lewin, N. E.; Liu, H.

L.; Blumberg, P. M. Residues in the second cysteine-rich region of

protein kinase C delta relevant to phorbol ester binding as revealed

by site-directed mutagenesis. J. Biol. Chem. 1995, 270, 21852-

21859.

(28) Irie, K.; Nakahara, A.; Nakagawa, Y.; Ohigashi, H.; Shindo, M.;

Fukuda, H.; Konishi, H.; Kikkawa, U.; Kashiwagi, K.; Saito, N.

Establishment of a binding assay for protein kinase C isozymes using

synthetic C1 peptides and development of new medicinal leads with

protein kinase C isozyme and C1 domain selectivity. Pharmacol.

Ther. 2002, 93, 271-281.

(29) Im, W.; Feig, M.; Brooks, C. L., III. An implicit membrane

generalized Born theory for the study of structure, stability, and

interactions of membrane proteins. Biophys. J. 2003, 85, 2900-2918.

(30) Bogi, K.; Lorenzo, P. S.; Szallasi, Z.; Acs, P.; Wagner, G. S.;

Blumberg, P. M. Differential selectivity of ligands for the C1A and

C1b phorbol ester binding domains of protein kinase C delta: Possible

correlation with tumor-promoting activity. Cancer Res. 1998, 58,

1423-1428.

(31) Yang, C. F.; Kazanietz, M. G. Divergence and complexities in DAG

signaling: looking beyond PKC. Trends Pharmacol. Sci. 2003, 24,

602-608.

(32) Goekjian, P. G.; Jirousek, M. R. Protein kinase C in the treatment of

disease: signal transduction pathways, inhibitors, and agents in

development. Curr. Med. Chem. 1999, 6, 877-903.

(33) Meier, M.; King, G. L. Protein kinase C activation and its

pharmacological inhibition in vascular disease. Vasc. Med. 2000, 5,

173-185.

(34) Tortora, G.; Ciardiello, F. Antisense strategies targeting protein kinase

C: preclinical and clinical development. Semin. Oncol. 2003, 30,

26-31.

(35) Szallasi, Z.; Denning, M. F.; Smith, C. B.; Dlugosz, A. A.; Yuspa,

S. H.; Pettit, G. R.; Blumberg, P. M. Bryostatin 1 protects protein

kinase C-delta from down-regulation in mouse keratinocytes in

parallel with its inhibition of phorbol ester-induced differentiation.

Mol. Pharmacol. 1994, 46, 840-850.

(36) Szallasi, Z.; Smith, C. B.; Pettit, G. R.; Blumberg, P. M. Differential

regulation of protein kinase C isozymes by bryostatin 1 and phorbol

12-myristate 13-acetate in NIH 3T3 fibroblasts. J. Biol. Chem. 1994,

269, 2118-2124.

(37) Szallasi, Z.; Krsmanovic, L.; Blumberg, P. M. Nonpromoting 12-

deoxyphorbol 13-esters inhibit phorbol 12-myristate 13-acetate

induced tumor promotion in Cd-1 mouse skin. Cancer Res. 1993,

53, 2507-2512.

(38) Bocklandt, S.; Blumberg, P. M.; Hamer, D. H. Activation of latent

HIV-1 expression by the potent anti-tumor promoter 12-deoxyphorbol

13-phenylacetate. AntiViral Res. 2003, 59, 89-98.

(39) Kedei, N.; Lundberg, D. J.; Toth, A.; Welburn, P.; Garfield, S. H.;

Blumberg, P. M. Characterization of the interaction of ingenol

3-angelate with protein kinase C. Cancer Res. 2004, 64, 3243-3255.

(40) Hommel, U.; Zurini, M.; Luyten, M. Solution structure of a cysteine-

rich domain of rat protein-kinase-C. Nat. Struct. Biol. 1994, 1, 383-

387.

(46) Irie, K.; Nakagawa, Y.; Ohigashi, H. Indolactam and benzolactam

compounds as new medicinal leads with binding selectivity for C1

domains of protein kinase C isozymes. Curr. Pharm. Des. 2004, 10,

1371-1385.

(47) Irie, K.; Nakagawa, Y.; Ohigashi, H. Toward the development of

new medicinal leads with selectivity for protein kinase C isozymes.

Chem. Rec. 2005, 5, 185-195.

(48) Lorenzo, P. S.; Bogi, K.; Hughes, K. M.; Beheshti, M.; Bhattacharyya,

D.; Garfield, S. H.; Pettit, G. R.; Blumberg, P. M. Differential roles

of the tandem C1 domains of protein kinase C delta in the biphasic

down-regulation induced by bryostatin 1. Cancer Res. 1999, 59,

6137-6144.

(49) Szallasi, Z.; Bogi, K.; Gohari, S.; Biro, T.; Acs, P.; Blumberg, P. M.

Non-equivalent roles for the first and second zinc fingers of protein

kinase Cdelta. Effect of their mutation on phorbol ester-induced

translocation in NIH 3T3 cells. J. Biol. Chem. 1996, 271, 18299-

18301.

(50) Stahelin, R. V.; Digman, M. A.; Medkova, M.; Ananthanarayanan,

B.; Rafter, J. D.; Melowic, H. R.; Cho, W. H. Mechanism of

diacylglycerol-induced membrane targeting and activation of protein

kinase C delta. J. Biol. Chem. 2004, 279, 29501-29512.

(51) Kazanietz, M. G.; Areces, L. B.; Bahador, A.; Mischak, H.;

Goodnight, J.; Mushinski, J. F.; Blumberg, P. M. Characterization

of ligand and substrate specificity for the calcium-dependent and

calcium-independent protein kinase C isozymes. Mol. Pharmacol.

1993, 44, 298-307.

(52) Lewin, N. E.; Blumberg, P. M. [³H]Phorbol 12,13-dibutyrate binding

assay for protein kinase C and related proteins. Methods Mol. Biol.

2003, 233, 129-156.

(53) Sharma, R.; Lee, J.; Wang, S.; Milne, G. W.; Lewin, N. E.; Blumberg,

P. M.; Marquez, V. E. Conformationally constrained analogues of

diacylglycerol. 10. Ultrapotent protein kinase C ligands based on a

racemic 5-disubstituted tetrahydro-2-furanone template. J. Med.

Chem. 1996, 39, 19-28.

(54) Lee, J.; Wang, S.; Milne, G. W.; Sharma, R.; Lewin, N. E.; Blumberg,

P. M.; Marquez, V. E. Conformationally constrained analogues of

diacylglycerol. 11. Ultrapotent protein kinase C ligands based on a

chiral 5-disubstituted tetrahydro-2-furanone template. J. Med. Chem.

1996, 39, 29-35.

(55) Lee, J.; Sharma, R.; Wang, S.; Milne, G. W.; Lewin, N. E.; Szallasi,

Z.; Blumberg, P. M.; George, C.; Marquez, V. E. Conformationally

constrained analogues of diacylglycerol. 12. Ultrapotent protein kinase

C ligands based on a chiral 4,4-disubstituted heptono-1,4-lactone

template. J. Med. Chem. 1996, 39, 36-45.

(56) Teng, K.; Marquez, V. E.; Milne, G. W. A.; Barchi, J. J.; Kazanietz,

M. G.; Lewin, N. E.; Blumberg, P. M.; Abushanab, E. Conforma-

tionally constrained analogs of diacylglycerol. Interaction of gamma-

lactones with the phorbol ester receptor of protein-kinase-C. J. Am.

Chem. Soc. 1992, 114, 1059-1070.

(57) Tripos Inc., 1699 South Hanley Road, St. Louis, MO 63144

(www.tripos.com).

(58) Halgren, T. A. Merck molecular force field. 1. Basis, form, scope,

parameterization, and performance of MMFF94. J. Comput. Chem.

1996, 17, 490-519.

(59) Rarey, M.; Kramer, B.; Lengauer, T.; Klebe, G. A fast flexible

docking method using an incremental construction algorithm. J. Mol.

Biol. 1996, 261, 470-489.

(60) Bower, M. J.; Cohen, F. E.; Dunbrack, R. L. Prediction of protein

side-chain rotamers from a backbone-dependent rotamer library: a

new homology modeling tool. J. Mol. Biol. 1997, 267, 1268-1282.

(61) Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.;

Swaminathan, S.; Karplus, M. Charmm, a program for macromo-

DAG-dioxolanones

lecular energy, minimization, and dynamics calculations. J. Comput.

Journal of Medicinal Chemistry, 2007, Vol. 50, No. 15 3481

(63) Mackerell, A. D.; Feig, M.; Brooks, C. L. Extending the treatment

of backbone energetics in protein force fields: limitations of gas-

phase quantum mechanics in reproducing protein conformational

distributions in molecular dynamics simulations. J. Comput. Chem.

2004, 25, 1400-1415.

Chem. 1983, 4, 187-217.

(62) MacKerell, A. D.; Bashford, D.; Bellott, M.; Dunbrack, R. L.;

Evanseck, J. D.; Field, M. J.; Fischer, S.; Gao, J.; Guo, H.; Ha, S.;

Joseph-McCarthy, D.; Kuchnir, L.; Kuczera, K.; Lau, F. T. K.;

Mattos, C.; Michnick, S.; Ngo, T.; Nguyen, D. T.; Prodhom, B.;

Reiher, W. E.; Roux, B.; Schlenkrich, M.; Smith, J. C.; Stote, R.;

Straub, J.; Watanabe, M.; Wiorkiewicz-Kuczera, J.; Yin, D.; Karplus,

M. All-atom empirical potential for molecular modeling and dynamics

studies of proteins. J. Phys. Chem. B 1998, 102, 3586-3616.

(64) Nina, M.; Beglov, D.; Roux, B. Atomic radii for continuum

electrostatics calculations based on molecular dynamics free energy

simulations. J. Phys. Chem. B 1997, 101, 5239-5248.

JM0702579

Article Doi

Conformationally constrained analogues of diacylglycerol (DAG). 28. DAG-dioxolanones reveal a new additional interaction site in the C1b domain of PKCδ

DOI: 10.1021/jm0702579

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm0702579

Full text of DOI:10.1021/jm0702579

Products guided by the article

R&D Labs maybe for 946570-44-1

Relevant to this article

Hot Product