Conformationally constrained analogues of diacylglycerol. 11.1 Ultrapotent protein kinase C ligands based on a chiral 5-disubstituted tetrahydro-2-furanone template

Article abstract of DOI:10.1021/jm950277n

Conformationally constrained analogues of diacylglycerol (DAG) built on a racemic 5-[(acyloxy)-methyl]-5-(hydroxymethyl)tetrahydro-2-furanone template were shown previously to have excellent binding affinities for protein kinase C (PK-C). Since the interaction of PK-C with DAG is stereospecific, it was anticipated that PK-C would bind tightly to only one enantiomeric form of the compounds constructed with this new lactone template. Separation of enantiomers by chiral HPLC was discarded due to the ease with which acyl migration occurs in these class of compounds, and a total chiral synthesis was undertaken. Prior to chemical synthesis, the selection of the "correct" enantiomeric template was predicted by a molecular conformational analysis that compared the two enantiomers of DAG in their presumed "active" conformation with the two enantiomeric lactone templates. This presumed "active" conformation for DAG was derived from a previously developed pharmacophore model that uses the molecule of a potent phorbol diester as the ideal rigid template. The results from this analysis indicated that the "correct" lactone template corresponded to the inactive (R)-isomer of DAG. This analysis also predicted that the lactone template corresponding to the active (S)-DAG enantiomer would not fit adequately into the pharmacophore. The chiral syntheses of target compounds 2, 4, and 6, constructed on the selected, and presumably "correct" lactone template, were achieved from a common bicyclic intermediate (5R,8R,9R)-8,9-O-isopropylidene-2-keto-1,7-dioxaspiro[4.4]nonane (10) that was synthesized from commercially available 1,2:3,5-di-O-isopropylidene-α-D-threo-apiofuranose (7) by a very effective spirolactonization approach. On the basis of their ability to inhibit the binding of [³H-20]phorbol 12,13-dibutyrate (PDBU) to PK-Ca, the enantiomeric ligands 2, 4, and 6 were twice as potent as the corresponding racemates. These results confirm that binding of these lactones is stereospecific and consistent with a binding mechanism similar to that of DAG.

Full text of DOI:10.1021/jm950277n

J . Med. Chem. 1996, 39, 29-35
29
Con for m a tion a lly Con str a in ed An a logu es of Dia cylglycer ol. 11.1 Ultr a p oten t
P r otein Kin a se C Liga n d s Ba sed on a Ch ir a l 5-Disu bstitu ted
Tetr a h yd r o-2-fu r a n on e Tem p la te
J eewoo Lee,† Shaomeng Wang,† George W. A. Milne,† Rajiv Sharma,† Nancy E. Lewin,‡ Peter M. Blumberg,‡ and
Victor E. Marquez*,†
Laboratory of Medicinal Chemistry, Developmental Therapeutics Program, Division of Cancer Treatment, and Molecular
Mechanisms of Tumor Promotion Section, Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer
Institute, National Institutes of Health, Bethesda, Maryland 20892
Received April 12, 1995X
Conformationally constrained analogues of diacylglycerol (DAG) built on a racemic 5-[(acyloxy)-
methyl]-5-(hydroxymethyl)tetrahydro-2-furanone template were shown previously to have
excellent binding affinities for protein kinase C (PK-C). Since the interaction of PK-C with
DAG is stereospecific, it was anticipated that PK-C would bind tightly to only one enantiomeric
form of the compounds constructed with this new lactone template. Separation of enantiomers
by chiral HPLC was discarded due to the ease with which acyl migration occurs in these class
of compounds, and a total chiral synthesis was undertaken. Prior to chemical synthesis, the
selection of the “correct” enantiomeric template was predicted by a molecular conformational
analysis that compared the two enantiomers of DAG in their presumed “active” conformation
with the two enantiomeric lactone templates. This presumed “active” conformation for DAG
was derived from a previously developed pharmacophore model that uses the molecule of a
potent phorbol diester as the ideal rigid template. The results from this analysis indicated
that the “correct” lactone template corresponded to the inactive (R)-isomer of DAG. This
analysis also predicted that the lactone template corresponding to the active (S)-DAG
enantiomer would not fit adequately into the pharmacophore. The chiral syntheses of target
compounds 2, 4, and 6, constructed on the selected, and presumably “correct” lactone template,
were achieved from a common bicyclic intermediate (5R,8R,9R)-8,9-O-isopropylidene-2-keto-
1,7-dioxaspiro[4.4]nonane (10) that was synthesized from commercially available 1,2:3,5-di-
O-isopropylidene-R-D-threo-apiofuranose (7) by a very effective spirolactonization approach. On
the basis of their ability to inhibit the binding of [3H-20]phorbol 12,13-dibutyrate (PDBU) to
PK-CR, the enantiomeric ligands 2, 4, and 6 were twice as potent as the corresponding
racemates. These results confirm that binding of these lactones is stereospecific and consistent
with a binding mechanism similar to that of DAG.
Ch a r t 1
In tr od u ction
The high binding affinity of ligands to either catalytic
or allosteric sites in enzymes is normally stereospecific.
Protein kinase C (PK-C) is no exception, and activation
of this enzyme ensues only after the binding of the (S)-
DAG enantiomer to the regulatory site of PK-C.2-4 In
our previous paper,1 we have disclosed the properties
of a new lactone, 5-[(acyloxy)methyl]-5-(hydroxymethyl)-
tetrahydro-2-furanone (Chart 1), that serves as a tem-
plate for the construction of compounds that function
as semirigid DAG mimetics.1 The increased binding
affinities for PK-C shown by compounds constructed on
this template appear to be derived principally from a
reduction in the entropic penalty paid during the
binding of the more rigid analogues, relative to the
binding of the more flexible DAG molecule. In addition,
the lactone template provides for an excellent spatial
disposition of the DAG pharmacophores (two ester
carbonyls and a primary alcohol function) that is pos-
sibly very similar to that found in bound DAG. Since
our previous investigations with this template were
performed with racemic compounds, it was reasonable
to expect that PK-C would bind tightly to only one
enantiomeric form of the compounds built with this new
lactone template.
If the same idealized intramolecular cyclization that
led to the racemic 5-[(acyloxy)methyl]-5-(hydroxymeth-
yl)tetrahydro-2-furanone template is now executed with
either (S)-DAG or (R)-DAG (Chart 1), the resulting
templates would be chiral also. In such a process, the
relative chiralities of the sn-2 carbon in the starting
DAG and the corresponding C-5 carbon in the lactone
remain unchanged, and a first educated guess would
perhaps predict that the lactone derived from the active
(S)-DAG enantiomer should provide the “active” tem-
plate. However, when the resulting (S)- and (R)-
† Laboratory of Medicinal Chemistry, Developmental Therapeutics
Program.
‡ Molecular Mechanisms of Tumor Promotion Section, Laboratory
of Cellular Carcinogenesis and Tumor Promotion.
X Abstract published in Advance ACS Abstracts, November 1, 1995.
This article not subject to U.S. Copyright. Published 1996 by the American Chemical Society

30 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 1
Lee et al.
Ch a r t 2
lactones were modeled to fit the phorbol ester pharma-
cophore which was derived from phorbol 12,13-dibutyrate
(PDBU),5 only the (R)-lactone provided an excellent fit
to phorbol (vide infra).
Selection of th e Cor r ect En a n tiom er ic Tem -
p la te. The selection of the correct enantiomeric tem-
plate was performed before embarking on a total
enantioselective synthesis. As a first step, simple (S)-
and (R)-DAG diacetates were modeled to fit the phorbol
ester pharmacophore. The critical functional pharma-
cophores in the phorbol esters that are responsible for
PK-C recognition are thought to be the C-20 hydroxyl,
the C-3 carbonyl, and the C-9 hydroxyl (PDBU, Chart
2).5 The latter two appear to function as hydrogen bond
acceptors, while the primary alcohol at C-20 functions
as a hydrogen bond donor. Correspondingly in DAG,
the two ester carbonyls behave as hydrogen bond
acceptors and the primary alcohol as a hydrogen bond
donor.5 The conformation of (S)-DAG demanded by the
phorbol ester pharmacophore (Figure 1a) was found to
be a stable and low-energy conformation only 4 kcal/
mol above the global energy minimum. On the other
hand, the conformation of (R)-DAG demanded by the
phorbol ester pharmacophore (Figure 1b) was a high-
energy conformation 10 kcal/mol above the global energy
minimum. Although (R)-DAG could in principle inter-
act with PK-C, the resulting system would have little
or no free energy gain since it would have to pay a larger
conformational energy penalty. This would explain why
PK-C binds exclusively to (S)-DAG and why (R)-DAG
is not an effective ligand. For the lactones, which are
conceptually generated as indicated in Chart 1, the
situation is reversed. Cyclization of (S)-DAG into the
(S)-lactone changes the desired orientation of pharma-
cophores as a result of the C-O bond rotation required
to achieve ring closure (Figure 2a). Conversely, when
(R)-DAG is cyclized into the (R)-lactone, the unstable
but desired conformation of (R)-DAG becomes a stable
conformation in the (R)-lactone with the pharmacoph-
ores maintained in their correct disposition (Figure 2b).
On the basis of this analysis, the (R)-lactone was
selected as the potentially “active” template for the
chemical synthesis of chiral ligands. This conclusion
was validated by our results comparing the PK-C
binding data of compounds containing the (R)-lactone
template versus the corresponding racemates (vide
infra).
F igu r e 1. (a) Molecular superposition of a conformation of
(S)-DAG (4.07 kcal/mol above the global minimum) that
provides the best fit (rms ) 0.170 Å) to phorbol 12,13-
dibutyrate (PDBU). (b) Molecular superposition of a conforma-
tion of (R)-DAG (10.35 kcal/mol above the global minimum)
that provides the best fit (rms ) 0.330 Å) to phorbol 12,13-
dibutyrate (PDBU).
F igu r e 2. (a) Idealized cyclization of diacetyl (S)-DAG from
Figure 1a into a lactone template. The C-O bond indicated
by the arrow must rotate as shown to achieve ring closure. (b)
Idealized cyclization of diacetyl (R)-DAG from Figure 1b into
a lactone template. No bond rotations are required and the
correct orientation of pharmacophores is maintained.
The decision to proceed with the stereospecific syn-
thesis of the presumed active (R)-lactones (2, 4, and 6,
Table 1), as opposed to the alternative of separating
each of the enantiomers by chiral HPLC from the
racemates (1, 3, and 5, Table 1), was based on the
knowledge that a rapid acyl migration anticipated to
occur during chromatography would inevitably cause
racemization.

Conformationally Constrained Analogues of Diacylglycerol. 11
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 1 31
Ta ble 1. Apparent Ki (nM) Values for Ligands as Inhibitors of
PDBU Binding to PK-C
new diol intermediate 9. This diol was oxidized with
pyridinium chlorochromate (PCC) which after the ensu-
ing cyclization to the lactol was further oxidized in situ
to the spirolactone 10. Alternatively, and with the
identical stereochemical outcome, a one-pot SmI2-
catalyzed reductive coupling of 8 with methyl acrylate9
produced 10 directly but in lower yield. Since only the
convex face of 8 is accessible, the attacking reagents
added stereospecifically from the less hindered â-side
ensuring that the resulting spirolactone 10 contained
the required stereochemistry for the construction of the
target chiral template. Indeed, after removal of the
acetonide group, metaperiodate cleavage of the vicinal
diol group in 11 provided the required lactone 12.
Wittig olefination of 12 with methyltriphenylphospho-
nium bromide proceeded with the simultaneous cleav-
age of the formate ester to give 13. Protection of the
primary alcohol function as the benzyl ether was fol-
lowed by the simultaneous treatment with OsO4 and
sodium metaperiodate to cleave the cis-hydroxylated
intermediate to the aldehyde stage. Sodium borohy-
dride reduction of the aldehyde produced the chiral,
singly protected lactone 14. This indirect approach was
necessary since reduction of aldehyde 12 would have
given an achiral diol. Esterification with myristoyl
chloride to the penultimate intermediate 15 and cleav-
age of the benzyl ether with BCl3 at -78 °C was followed
by a careful, low-temperature workup that involved
quenching the reaction with a neutral buffer and
extraction of the desired chiral lactone with ethyl ether.
As anticipated, exposure of compound 2 to either
chromatographic conditions, silica or neutral alumina,
caused rapid racemization.
Sch em e 1
The more complex R-alkylidene lactones 4 and 6 were
prepared from the pivotal intermediate spirolactone 10
(Scheme 2). In this instance, only the oleoyl aldehyde
was used which after dehydration gave compound 16
as a mixture of Z- and E-isomers. Removal of the
isopropylidene function from 16 gave 17, and sodium
borohydride reduction of the resulting lactol afforded
the R-alkylidene lactone 18 with the desired chirality
at C-4 (C-5 if named as a tetrahydro-2-furanone). The
vicinal diol function on the side chain was protected as
the acetonide, while the remaining primary alcohol
function was protected as a benzyl ether to give 19.
Metaperiodic acid cleavage of 19, followed by reduction
with sodium borohydride, gave the antepenultimate
intermediate 20. After the conversion of 20 to a mixture
of acetates, compounds 21 and 22 were separated by
column chromatography. As before, cleavage of the
benzyl ether with BCl3 at -78 °C was followed by a
similar workup as described before. This approach
produced the individual target Z- and E-isomers 4 and
6.
Ch em istr y
Biologica l Resu lts
The simplest compound with the (R)-lactone template
and with the myristate acyl chain (compound 2) was
constructed as shown in Scheme 1. Commercially
available 1,2:3,5-di-O-isopropylidene-R-D-threo-apiofura-
nose (7) was hydrolyzed under mild conditions to
selectively remove the 3,5-O-isopropylidene group, and
the resulting free diol was oxidized to the known keto
intermediate 8.6 Two spirolactonization approaches
were attempted.7 The first one began with the addition
of the Grignard reagent, MgClO(CH2)3MgCl,8 to give a
The affinity of the enantiomeric ligands 2, 4, and 6
for PK-CR was expressed in terms of their ability to
displace bound [3H-20]phorbol 12,13-dibutyrate (PDBU)
and was measured in the same manner as for the
racemates described in the preceding paper.1 Comparing
the inhibition constants (Ki) obtained for each enanti-
omer with those of the corresponding racemate (Table
1), one finds, as expected, that the Ki values for the
active enantiomers are very close to half the values for
the racemates. These results not only substantiate the

32 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 1
Lee et al.
analyzed by competition with [3H]PDBU binding essentially
as described in our previous work,11 except that the PK-C
preparation used here was the single isozyme PK-CR. This
recombinant PK-CR was expressed in the baculovirus system
and was isolated as described in ref 12. The ID50 values were
determined from the competition curves, and the correspond-
ing Ki values for the ligands were calculated from the ID50
values as described before.11,12 Values represent the mean (
standard error (three determinations).
Sch em e 2
Molecu la r Mod elin g. All molecular mechanics calcula-
tions were carried out with QUANTA molecular modeling
package (version 3.1) with CHARMm 2.2 parameter set
running on a Silicon Graphics IRIS Indigo workstation. Since
the molecular mechanics calculation depended upon the atom
types, the atom type for every atom was carefully examined
to ensure it was correct before any molecular calculations were
conducted. All the conformational analysis was performed
with the conformational search module within QUANTA.
Monte Carlo random sampling algorithm was used to generate
5000 conformations for each structure. For each generated
conformation, 5000 steps of adopted-basis Newton Raphson
(ABNR) minimization was carried out or until convergence
(with a convergence criterion equal to 0.01 kcal Å-1). For both
(S)-DAG and (R)-DAG diacetates (Figure 1a), the seven sp3-
sp3 bonds between the hydroxyl oxygen and the two carbonyl
carbons, which are relevant to the pharmacophore, were
sampled in the Monte Carlo conformational analysis. Simi-
larly, for both (S)- and (R)-lactones (Figure 1b), the five sp3-
sp3 bonds between the hydroxyl oxygen and the two carbonyl
carbons were sampled. An angle increment of 60° was used
for each defined bond in the Monte Carlo conformational
sampling, and a dielectric constant of 1 was used (CDIE ) 1)
in all the calculations. For each minimized conformer, the
relevant distance information, as well as the conformational
energy, was downloaded into an ASCII file. A program was
then developed to compare automatically these conformations
to the phorbol pharmacophore template to identify the best
conformations according to considerations of both the root
mean square (rms) value and the conformational energy. The
phorbol 12,13-dibutyrate (PDBU) structure (Figure 1) was
constructed starting from the crystal structure of phorbol13
obtained from the Cambridge Structural Database,14 using the
molecular editor within QUANTA. Due to the lack of some
CHARMm parameters for PBDU, a semiempirical quantum
mechanics method, AM1 in the MOPAC 6.0 package running
on a host mainframe Convex C240, was employed to minimize
the PDBU structure which was used in this study.
stereospecificity of PK-C for these lactones but also
indirectly confirm the validity of our pharmacophore
model5 used to predict the correct enantiomeric lactone.
In addition, these results support the concept that the
new ligands are interacting at the binding site of the
enzyme with their key pharmacophores spatially ar-
ranged in a manner similar to that encountered in the
natural bound ligand DAG, but with a definitive en-
tropic advantage.1 To date, compounds 4 and 6 repre-
sent the most potent PK-C agonists possessing a dia-
cylglycerol-like structure. These results contradict an
earlier notion that the DAG molecule was considered
too simple to give rise to a high-affinity ligand.10
1,2-O-Isop r op ylid en e-r-D-glycer o-t et r ose-3-u lose (8).
This compound was prepared from di-O-isopropylidene-R-D-
apiose 7 in two steps according to the method of Carey et al.6
1,2-O-Isop r op ylid en e-3-C-(h yd r oxyp r op yl)-r-D-er yth -
r ofu r a n ose (9). A solution of 3-chloropropanol (2.84 g, 30
mmol) in THF (10 mL) was cooled to -20 °C, treated dropwise
with methylmagnesium chloride (3 M in tetrahydrofuran, 10
mL, 30 mmol), and stirred for 20 min. The reaction mixture
was warmed to room temperature, and magnesium (1.10 g,
45 mmol) was added. The suspension was refluxed for 3 h
with intermittent addition of dibromoethane (0.02 mL each
at 0, 1 and 2 h) and cooled to room temperature. A solution
of 8 (1.58 g, 10 mmol) in THF (10 mL) was added dropwise to
this mixture, and after stirring for 1 h, the reaction mixture
was cooled over an ice bath and the reaction quenched by the
slow addition of a saturated NH4Cl solution (20 mL). The
reaction mixture was filtered, and the filtrate was extracted
several times with EtOAc. The combined organic layer was
dried (Na2SO4) and concentrated. The residue was purified
by flash column chromatography over silica gel with EtOAc/
hexanes (5:1) as eluant to give the title compound 9 (1.702 g,
Exp er im en ta l Section
Gen er a l Exp er im en ta l. All chemical reagents were com-
mercially available. Melting points were determined on a Mel-
Temp II apparatus, Laboratory Devices, USA, and are uncor-
rected. Silica gel column chromatography was performed on
silica gel 60, 230-400 mesh (E. Merck). Proton and 13C NMR
spectra were recorded on a Bruker AC-250 instrument at 250
and 62.9 MHz, respectively. Spectra were referenced to the
solvent in which they were run (7.24 ppm for CDCl3). Infrared
spectra were recorded on a Perkin-Elmer 1600 Series FTIR,
and specific rotations were measured in a Perkin-Elmer Model
241 polarimeter. Elemental analyses were performed by
Atlantic Microlab, Inc., Atlanta, GA.
78%) as a white solid: mp 93-95 °C; [R]22 +28.46° (c 1.3,
D
CHCl3); IR (CHCl3) 3433 cm-1 (OH); H NMR (CDCl3) δ 1.34
1
and 1.56 (singlets, 3 H, CH3), 1.60-1.80 (m, 4 H, CH2CH2-
CH2OH), 2.48 (br s, 2 H, OH), 3.60-3.76 (m, 4 H, CH2OH,
H-4a,b), 4.14 (d, J ) 3.8 Hz, 1 H, H-2), 5.79 (d, J ) 3.8 Hz, 1 H,
H-1); 13C NMR (CDCl3) δ 26.32, 26.49, 26.53, 32.03, 62.44,
72.49, 78.20, 81.83, 105.10, 112.43. Anal. (C10H18O5) C, H.
An a lysis of In h ibition of [3H]P DBU Bin d in g by Non -
r a d ioa ctive Liga n d s. Enzyme-ligand interactions were

Conformationally Constrained Analogues of Diacylglycerol. 11
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 1 33
(5R,8R,9R)-8,9-O-Isop r op ylid en e-2-k eto-1,7-d ioxa spir o-
[4.4]n on a n e (10): Meth od A. A solution of 9 (1.702 g, 7.8
mmol) in CH2Cl2 (100 mL) was treated with pyridinium
chlorochromate (6.725 g, 31.2 mmol) and 4 Å molecular sieves
(7.8 g). After the reaction mixture was stirred for 1 h at room
temperature, the reaction was quenched with a slurry of Celite
in ether and the mixture stirred for 30 min more. The
suspension was filtered through a pad of silica gel and further
washed with EtOAc. The filtrate was concentrated, and the
residue was purified by flash column chromatography over
silica gel with EtOAc/hexanes (2:1) as eluant to give the title
compound 10 (1.654 g, 99%) as a white solid: mp 101-102
(CDCl3) δ 2.05 (m, 1 H, H-4a), 2.35-2.74 (m, 3 H, H-4b, H-3a,b),
3.55 (AB q, J ) 10.4 Hz, 2 H, CH2OH), 4.60 (AB q, J ) 12.0
Hz, 2 H, CH2OCH2Ph), 5.22 (dd, J ) 10.9, 0.5 Hz, 1 H,
CHdCHH), 5.37 (dd, J ) 17.3, 0.6 Hz, 1 H, CHdCHH), 5.88
(dd, J ) 17.3, 10.9 Hz, 1 H, CHdCH2); 13C NMR (CDCl3) δ
28.88, 29.37, 73.66, 74.50, 86.81, 115.73, 127.58, 127.80,
128.46, 136.72, 137.65, 176.82. Anal. (C14H16O3) C, H. The
above compound (0.162 g, 0.7 mmol) was dissolved in aqueous
acetone (1:1, 10 mL) and the solution was stirred with
4-methylmorpholine N-oxide (0.164 g, 1.4 mmol), sodium
metaperiodate (0.300 g, 1.4 mmol), and osmium tetroxide (2.5
wt % in tert-butyl alcohol, 0.88 mL, 0.07 mmol) for 20 h at
room temperature. The reaction mixture was then quenched
with saturated sodium thiosulfate solution (5 mL), stirred for
10 min, and extracted three times with EtOAc. The combined
organic layer was washed with water, dried (Na2SO4), and
concentrated. The residue was purified by flash column
chromatography over silica gel with EtOAc/hexane (4:1) as
eluant to give the corresponding aldehyde (0.156 g, 96%) as
an oil. This compound was immediately dissolved in MeOH
(10 mL), cooled to -10 °C, and treated with sodium borohy-
dride (0.090 g, 2.4 mmol). After stirring for 30 min, the
reaction mixture was quenched with 1 N HCl and diluted with
EtOAc. The organic layer was washed with water, dried (Na2-
SO4), and concentrated. The residue was purified by flash
column chromatography over silica gel with EtOAc/hexane
(from 3:1 to 6:1) as eluant to give the title compound 14 (0.110
°C; [R]22 +74.0° (c 1.0, CHCl3); IR (CHCl3) 1785 cm-1(CdO);
D
1H NMR (CDCl3) δ 1.34 and 1.59 (singlets, 3 H, CH3), 2.07-
2.33 (m, 2 H, H-4a,b), 2.64 (t, J ) 8.3 Hz, 2 H, H-3a,b), 3.70 (d
of AB, J ) 8.5 Hz, 1 H, H-6a), 4.16 (d of AB, J ) 8.5 Hz, 1H,
H-6b), 4.33 (d, J ) 3.5 Hz, 1 H, H-9), 5.82 (d, J ) 3.5 Hz, 1 H,
H-8); 13C NMR (CDCl3) δ 26.51, 26.68, 27.47, 30.05, 70.49,
82.31, 86.71, 104.87, 114.23, 174.81. Anal. (C10H14O5) C, H.
Meth od B. A solution of 8 (0.316 g, 2 mmol) in THF (5
mL) was cooled to 0 °C and treated with a solution of methyl
acrylate (0.36 mL, 4 mmol) in a mixture of 2-propanol (0.23
mL, 3 mmol) and hexamethylphosphoramide (2 mL). Sa-
marium iodide (0.1 M in THF, 60 mL, 6 mmol) was added
dropwise to the reaction mixture which was then allowed to
warm to room temperature. After stirring for 30 min, the
mixture was quenched with ether (50 mL) and filtered through
a pad of silica gel. The filtrate was concentrated, and the
residue was purified by flash column chromatography over
silica gel with EtOAc/hexanes (2:1) as eluant to give the
identical compound 10 (0.124 g, 30%) as a white solid.
g, 70%) as a white solid: mp 76-78 °C; [R]22 +7.73° (c 4.4,
D
CHCl3); IR (CHCl3) 3446 (OH) and 1771 cm-1 (CdO); 1H NMR
(CDCl3) δ 2.10-2.20 (m, 2 H, H-4a,b), 2.50-2.75 (m, 2 H, H-3a,b),
3.60 (AB q, J ) 10.1 Hz, 2 H, CH2OH), 3.62 (d of AB, J ) 12.1
Hz, 1 H, CHHOCH2Ph), 3.75 (d of AB, J ) 12.1 Hz, 1 H,
CHHOCH2Ph), 4.54 (br s, 2 H, CH2OCH2Ph), 7.20-7.40 (m, 5
H, Ph); 13C NMR (CDCl3) δ 25.70, 29.15, 65.53, 72.43, 73.72,
87.46, 127.62, 127.90, 128.50, 137.49, 177.24. Anal. (C13H16O4)
C, H.
(S)-5-Vin yl-5-(h yd r oxym eth yl)tetr a h yd r o-2-fu r a n on e
(13). A solution of 10 (0.857 g, 4 mmol) in THF (20 mL) was
treated with 1 N HCl solution (20 mL) and stirred for 20 h at
room temperature. The reaction mixture was neutralized with
solid NaHCO3 and diluted with EtOAc. The mixture was dried
and concentrated to give hemiacetal 11, which was used in
the next step without further purification. This compound was
dissolved in a mixture of MeOH (40 mL) and water (20 mL),
and the solution was stirred with sodium metaperiodate (1.71
g, 8 mmol) for 2 h at room temperature. The reaction mixture
was filtered, and the filtrate was concentrated to dryness. The
residue was dissolved in EtOAc, dried (Na2SO4), and concen-
trated to give aldehyde 12, which was also used without
further purification. Aldehyde 12 was dissolved in THF (10
mL), and the solution was slowly added to a suspension of
methylphosphonium ylide [this ylide was prepared from me-
thyltriphenylphosphonium bromide (2.858 g, 8 mmol) and
potassium tert-butoxide (1.0 M in THF, 8 mL, 8 mmol) in THF
(10 mL) after stirring for 30 min at room temperature]. The
reaction mixture was stirred for 1 h at room temperature and
for 1 h at 60 °C before it was cooled to 0 °C. The reaction
mixture was then quenched with AcOH (0.5 mL) and filtered,
and the filtrate was concentrated. The residue was purified
by flash column chromatography over silica gel with EtOAc/
hexanes (4:1) as eluant to give the title compound 13 (0.387
g, 68% from 10) as an oil: IR (neat) 3440 (OH), 1770 (CdO),
and 1644 cm-1 (CdC); 1H NMR (CDCl3) δ 2.05 (m, 1 H, H-4a),
2.25 (s, 1 H, OH), 2.35-2.70 (m, 3 H, H-4b, H-3a,b), 3.54 (d of
AB, J ) 12.3 Hz, 1 H, CHHOH), 3.75 (d of AB, J ) 12.3 Hz,
1 H, CHHOH), 5.26 (d, J ) 10.9 Hz, 1 H, CHdCHH), 5.37 (d,
J ) 17.2 Hz, 1 H, CHdCHH), 5.83 (dd, J ) 17.2, 10.9 Hz, 1 H,
CHdCH2); 13C NMR (CDCl3) δ 28.05, 28.79, 66.70, 88.37,
116.10, 136.46, 177.41. Anal. (C7H10O3) C, H.
(R)-5-[(Ben zyloxy)m eth yl]-5-[(tetr a d eca n oyloxy)m eth -
yl]tetr a h yd r o-2-fu r a n on e (15). A solution of 14 (0.024 g,
0.1 mmol) in CH2Cl2 (5 mL) was stirred with a drop of pyridine
and a catalytic amount of (dimethylamino)pyridine (0.002 g,
0.016 mmol) and treated with tetradecanoyl chloride (0.054
mL, 0.2 mmol) for 2 h at room temperature. The solution was
concentrated, and the residue was purified by flash column
chromatography over silica gel with hexane/EtOAc (2:1) as
eluant to give the title compound 15 (0.041 g, 92%) as an oil:
[R]22 +1.43° (c 4.2, CHCl3); IR (neat) 1783 and 1743 cm-1
D
1
(CdO); H NMR (CDCl3) δ 0.85 (distorted triplet, 3 H, CH3),
1.10-1.40 (m, 20 H, CH3(CH2)10CH2CH2CO), 1.60 (br m, 2 H,
CH3(CH2)10CH2CH2CO), 2.00-2.30 (m, 2 H, H-4a,b), 2.30 (t, J
) 7.5 Hz, 2 H, CH3(CH2)10CH2CH2CO), 2.50-2.75 (m, 2 H,
H-3a,b), 3.60 (br s, 2 H, CH2OCH2Ph), 4.20 (AB q, J ) 11.9 Hz,
2 H, CH2OCO), 4.55 (br s, 2 H, CH2OCH2Ph), 7.20-7.40 (m, 5
H, Ph); 13C NMR (CDCl3) δ 14.09, 22.65, 24.81, 26.40, 28.78,
29.07, 29.20, 29.32, 29.42, 29.57, 29.61, 29.64, 31.89, 34.03,
65.99, 72.21, 73.76, 84.94, 127.63, 127.93, 128.49, 137.33,
173.14, 176.30. Anal. (C27H42O5) C, H.
(R)-5-[(Tetr a d eca n oyloxy)m eth yl]-5-(h yd r oxym eth yl)-
tetr a h yd r o-2-fu r a n on e (2). A solution of 15 (0.036 g, 0.08
mmol) in CH2Cl2 (4 mL) was cooled to -78 °C, treated with
boron trichloride (1.0 M in dichloromethane, 0.24 mL, 0.24
mmol), and stirred at that temperature for 1.5 h. The reaction
was quenched by the slow addition of a saturated NaHCO3
solution (0.3 mL) at -78 °C, and the mixture was immediately
partitioned between ice-cold ether and a pH 7 phosphate buffer
solution. The organic layer was washed five times with the
pH 7 buffer solution, dried (Na2SO4), and concentrated to give
a white solid. This solid was washed with cold hexane several
times to give a pure sample of the title compound 2 (24 mg,
(S)-5-[(Ben zyloxy)m eth yl]-5-(h yd r oxym eth yl)tetr a h y-
d r o-2-fu r a n on e (14). A solution of 13 (0.140 g, 1.0 mmol) in
DMF (5 mL) was stirred with silver(I) oxide (0.232 g, 1.0 mmol)
and benzyl bromide (0.30 mL, 2.5 mmol) for 5 days at room
temperature. The reaction mixture was filtered, diluted with
water, and extracted three times with EtOAc. The organic
layer was washed with water, dried (Na2SO4), and concen-
trated. The residue was purified by flash column chromatog-
raphy over silica gel with hexanes/EtOAc (2:1) as eluant to
give the intermediate 5-[(benzyloxy)methyl]-5-vinyltetrahydro-
84%) as a solid: mp 65-66 °C; [R]22 +1.43° (c 4.2, CHCl3).
D
The IR, 1H NMR, and 13C NMR were identical to those
reported for the racemate in the preceding paper (compound
number 3e).
(R)-(Z)-5-[(Acetyloxy)m eth yl]-5-(h ydr oxym eth yl)-3-[(Z)-
9-octa d eca en ylid en e]tetr a h yd r o-2-fu r a n on e (4) a n d (R)-
(E)-5-[(Acetyloxy)m eth yl]-5-(h yd r oxym eth yl)-3-[(Z)-9-oc-
ta d eca en ylid en e]tetr a h yd r o-2-fu r a n on e (6). A stirred
2-furanone (0.186 g, 80%) as an oil: [R]22 -29.68° (c 3.08,
D
CHCl3); IR (neat) 1772 (CdO) and 1653 cm-1 (CdC); 1H NMR

34 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 1
Lee et al.
solution of 10 (0.428 g, 2.0 mmol) in THF (4 mL) was cooled
to -78 °C and treated slowly with lithium bis(trimethylsilyl)
amide (1.0 M in tetrahydrofuran, 2.4 mL, 2.4 mmol) for 1 h. A
mixture of oleyl aldehyde (0.640 g, 2.4 mmol) and hexameth-
ylphosphoramide (0.430 g, 5.35 mmol) was added, and stirring
was continued for 1 h at -78 °C and for 1 h at -40 °C. The
mixture was quenched with a solution of saturated ammonium
chloride and diluted with ether. The organic layer was washed
with water, dried (Na2SO4), and concentrated. The residue
was purified by flash column chromatography over silica gel
with hexanes/EtOAc (3:1) as eluant to give the intermediate
â-hydroxy lactone (0.865 g, 90%). This compound was dis-
solved in CH2Cl2 (20 mL) and cooled to 0 °C. The solution
was then stirred with triethylamine (1.12 mL, 8 mmol) and
methanesulfonyl chloride (0.31 mL, 4 mmol) for 30 min. The
reaction mixture was warmed to room temperature and stirred
for 1 h before the addition of 1,8-diazabicyclo[5.4.0]undec-7-
ene (1.5 mL, 10 mmol). After further stirring for 14 h at room
temperature, the mixture was concentrated and diluted with
ether. The ethereal solution was washed with diluted HCl,
water, and brine. The organic layer was dried (Na2SO4) and
concentrated. The residue was purified by flash column
chromatography using silica gel with hexanes/EtOAc (4:1) as
eluant to give an inseparable oily mixture of E- and Z-isomers
(16, 0.800 g, 96%) with a E/Z ratio of 2:1 according to 1H NMR.
The 0.50 ppm difference between the â-cis (δ ) 6.82 ppm,
distortet triplet) and â-trans (δ ) 6.32 ppm, distorted triplet)
protons of the R,â-enone system in 16 allows one to differenti-
ate the E-isomer from the Z-isomer. The above mixture of 16
(0.463 g, 1 mmol) was dissolved in THF (15 mL) and stirred
in the presence of 2 N HCl (15 mL) for 5 days at room
temperature. The solution was then cooled over an ice bath
and neutralized with solid NaHCO3. The reaction mixture was
filtered, and the filtrate was concentrated. The residue was
diluted with EtOAc, dried (Na2SO4), and concentrated to give
hemiacetal 17 as an oil, which was used in the next step
without further purification. Hemiacetal 17 was dissolved in
MeOH (10 mL), cooled to -10 °C, and treated with small
portions of sodium borohydride until all the starting material
was consumed. The reaction mixture was slowly acidified with
acetic acid and concentrated. The residue was diluted with
EtOAc and was washed with 1 N HCl and water. The organic
layer was dried (Na2SO4) and concentrated to give triol 18 as
an oil, which also was used for the next step without further
purification. The above triol 18 was dissolved in acetone (20
mL) and cooled to 0 °C. The solution was treated with a
catalytic amount of p-toluenesulfonic acid and stirred for 1 h.
It was then neutralized with solid NaHCO3, filtered, and
concentrated. The residue was purified by flash column
chromatography over silica gel with hexanes/EtOAc (2:1) as
eluant to give the corresponding mixture of E- and Z-ac-
etonides (0.232 g, 50%) as an oil. The above mixture of
acetonides (0.232 g, 0.5 mmol) was dissolved in THF (10 mL)
and treated with sodium hydride (60% dispersion, 40 mg, 1
mmol) and a mixture of benzyl bromide (0.12 mL, 1 mmol)
and tetrabutylammonium iodide (0.037 g, 0.1 mmol). The
reaction mixture was stirred for 12 h at room temperature and
quenched by the addition of acetic acid (0.1 mL) and ether.
The suspension was filtered through a short pad of silica gel
which washed with ether. The filtrate was concentrated, and
the residue was purified by flash column chromatography over
silica gel with hexanes/EtOAc (4:1) as eluant to give 19 (0.240
g, 86%) as oil. Despite the very similar Rf values for the E-
and Z-isomers of 19, small amounts of each isomer were
separated in some fractions, as semisolid materials, in the
above chromatographic experiment.
(m, 2 H, >CdCHCH2), 2.75 (AB q, H-4a,b), 3.50 (s, 2 H, CH2-
OCH2Ph), 4.00 (m, 2 H, OCH2CHO), 4.25 (t, 1 H, OCH2CHO),
4.55 (s, 2 H, OCH2Ph), 5.33 (m, 2 H, CH2CHdCHCH2), 6.70
(m, 1 H, >CdCH), 7.20-7.50 (m, 5 H, Ph).
The mixture of E- and Z-isomers (compound 19, 0.239 g,
0.43 mmol) was dissolved in ether (30 mL) and treated with
periodic acid (0.490 g, 2.15 mmol). The reaction mixture was
stirred for 20 h at room temperature, filtered, and concen-
trated. The residue was purified by flash column chromatog-
raphy over silica gel with hexanes:EtOAc (3:2) as eluant to
give the corresponding mixture of E- and Z-aldehydes (0.200
g, 96%) as an oil. This mixture was immediately dissolved in
THF (10 mL) and H2O (1 mL), cooled to -10 °C, and treated
with small portions of sodium borohydride (total amount:
0.030 g) until all starting material was consumed. The
reaction mixture was acidified by the slow addition of acetic
acid and concentrated. The residue was diluted with ether
and washed with 1 N HCl solution and water. The organic
layer was dried (Na2SO4) and concentrated. The residue was
purified by flash column chromatography using silica gel with
hexanes/EtOAc (3:2) as eluant to give 20 (0.145 g, 72%) as an
oil. The above alcohol 20 (0.145 g, 0.3 mmol) was dissolved in
CH2Cl2 (20 mL), cooled to -10 °C, and treated with pyridine
(0.15 mL, 1.85 mmol), acetic anhydride (0.15 mL, 1.59 mmol),
and a catalytic amount of (dimethylamino)pyridine (0.015 g,
0.13 mmol). After stirring for 30 min, the reaction mixture
was concentrated at 0 °C. The residue was purified by flash
column chromatography over silica gel with ether/hexanes (1:
1) as eluant to give the Z-isomer 21 (0.052 g, 33%) and the
E-isomer 22 (0.103 g, 65%) as semisolid materials.
Z-Isomer 21: [R]22 +1.43° (c 0.28, CHCl3); IR (neat) 1752
D
and 1670 (CdO) and 1455 cm-1
;
1H NMR (CDCl3) δ 0.86
(distorted t, 3 H, CH3), 1.00-1.50 (m, 22 H, CH2(CH2)5CH2-
CHdCHCH2(CH2)6CH3), 2.00 (m, 4 H, CH2CHdCHCH2), 2.02
(s, 3 H, CH3CO), 2.60-2.90 (m, 4 H, >CdCHCH2, H-4a,b), 3.49
(d of AB, J ) 9.9 Hz, 1 H, CHHOCH2Ph), 3.57 (d of AB, J )
9.9 Hz, CHHOCH2Ph), 4.19 (s, 2 H, CH2COCH3), 4.54 (s, 2 H,
OCH2Ph), 5.33 (m, 2 H, CH2CHdCHCH2), 6.17 (m, 1 H,
>CdCH), 7.20-7.40 (m, 5 H, Ph).
E-Isomer 22: [R]22 -2.58° (c 0.62, CHCl3); IR (neat) 1752
D
and 1684 (CdO) and 1456 cm-1
;
1H NMR (CDCl3) δ 0.86
(distorted t, 3 H, CH3), 1.10-1.50 (m, 22 H, CH2(CH2)5CH2-
CHdCHCH2(CH2)6CH3), 2.00 (m, 4 H, CH2CHdCHCH2), 2.01
(s, 3 H, CH3CO), 2.15 (m, 2 H, >CdCHCH2), 2.62 (d of AB, J
) 17.0 Hz, 1 H, H-4a), 2.82 (d of AB, J ) 17.0 Hz, 1 H, H-4b),
3.50 (d of AB, J ) 9.9 Hz, 1 H, CHHOCH2Ph), 3.58 (d of AB,
J ) 9.9 Hz, CHHOCH2Ph), 4.21 (s, 2 H, CH2COCH3), 4.55 (s,
2 H, OCH2Ph), 5.33 (m, 2 H, CH2CHdCHCH2), 6.72 (m, 1 H,
>CdCH), 7.20-7.40 (m, 5 H, Ph).
A solution of Z-isomer 21 (0.103 g, 0.196 mmol) in CH2Cl2
(15 mL) was cooled to -78 °C, treated with boron trichloride
(1.0 M in dichloromethane, 0.8 mL, 0.8 mmol), and stirred for
1.5 h. The reaction mixture was quenched by the slow addition
of a saturated solution of NaHCO3 (0.8 mL) and immediately
partitioned between ice-cold ether and a pH 7 phosphate buffer
solution. The organic layer was washed five times with the
pH 7 buffer solution, dried (Na2SO4), and concentrated to give
the title compound 4, which precipitated in cold hexane. The
solid was filtered off and washed with cold hexane several
times to give optically active 4 (0.068 g, 80%) as a semisolid
1
gum; [R]22 +3.00° (c 0.3, CHCl3). The IR, H NMR, and 13C
D
NMR were identical to the spectra reported for the racemate
in the preceding paper (compound number 10f). The corre-
sponding E-isomer 6 was also prepared by the same procedure
in 80% yield: mp 46 °C; [R]22 +12.5° (c 0.24, CHCl3). The
D
IR, 1H NMR, and 13C NMR were identical to the spectra
reported for the racemate in the preceding paper (compound
number 11f).
Z-isomer 19: 1H NMR (CDCl3) δ 0.86 (distorted t, 3 H, CH3),
1.20-1.50 (m, 28 H, isopropylidene: 2 × CH3, CH2(CH2)5CH2-
CHdCHCH2(CH2)6CH3), 2.00 (m, 4 H, CH2CHdCHCH2), 2.65
(m, 2 H, >CdCHCH2), 2.85 (AB q, H-4a,b), 3.50 (s, 2 H, CH2-
OCH2Ph), 4.00 (m, 2 H, OCH2CHO), 4.23 (t, 1 H, OCH2CHO),
4.55 (s, 2 H, OCH2Ph), 5.35 (m, 2 H, CH2CHdCHCH2), 6.15
(m, 1 H, >CdCH), 7.20-7.50 (m, 5 H, Ph).
Refer en ces
(1) Sharma, R.; Lee, J .; Lewin, N. E.; Blumberg, P. M.; Marquez,
V. E. J . Med. Chem. 1996, 39, 19-28.
(2) Rando, R. R.; Young, N. The Stereospecific Activation of Protein
Kinase C. Biochem. Biophys. Res. Commun. 1984, 122, 818-
823.
E-isomer 19: 1H NMR (CDCl3) δ 0.86 (distorted t, 3 H, CH3),
1.20-1.50 (m, 28 H, isopropylidene: 2 × CH3, CH2(CH2)5CH2-
CHdCHCH2(CH2)6CH3), 2.00 (m, 4 H, CH2CHdCHCH2), 2.15

Conformationally Constrained Analogues of Diacylglycerol. 11
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 1 35
(3) Boni, L. T.; Rando, R. R. The Nature of Protein Kinase C
Activation by Physically Defined Phospholipid Vesicles and
Diacylglycerols. J . Biol. Chem. 1985, 260, 10819-10825.
(4) Rando, R. R. Regulation of Protein Kinase C Activity by Lipids.
FASEB J , 1988, 2, 2348-2355.
(5) Wang, S.; Milne, G. W. A.; Nicklaus, M. C.; Marquez, V. E.; Lee,
J .; Blumberg, P. M. Protein Kinase C. Modeling of the Binding
Site and Prediction of Binding Constants. J . Med. Chem. 1994,
4, 1326-1338.
(6) Carey, F. A.; Ball, D. H.; Long, L., J r. 1,2:3,5-Di-O-Isopropy-
lidene-R-D-Apio-L-Furanose. Carbohydr. Res. 1966, 3, 205-213.
(7) Lee, J .; Marquez, V. E.; Lewin, N. E.; Blumberg, P. M. Synthesis
of Two Rigid Diacylglycerol Analogues Having a 1,7-Dioxaspiro-
[4.4]nonane Bis-γ-Butyrolactone Skeleton. 4. Synlett 1994, 206-
208.
(8) Cahiez, G.; Alexakis, A.; Normant, J . F. Derives Organomag-
nesiens ω-Alcoolates: Preparation et Proprietes. Tetrahedron
Lett. 1978, 19, 3013-3014.
(11) Teng, K.; Marquez, V. E.; Milne, G. W. A.; Barchi, J . J ., J r.;
Kazanietz, M. G.; Lewin, N. E.; Blumberg, P. M.; Abushanab,
E. Conformationally Constrained Analogues of Diacylglycerol.
Interaction of γ-Lactones with the Phorbol Ester Receptor of
Protein Kinase C. J . Am. Chem. Soc. 1992, 114, 1059-1070.
(12) Kazanietz, M. G.; Areres, L. B.; Bahador, A.; Mischak, H.;
Goodnight, J .; Mushinski, J . F.; Blumberg, P. M. Characteriza-
tion of Ligand and Substrate Specificity for the Calcium-
Dependent and Calcium-Independent Protein Kinase C Isozymes.
Mol. Pharmacol. 1993, 44, 298-307.
(13) Brandl, V. F.; Ro¨hrl, M.; Zechmeister, K.; Hoppe, W. Rontgen-
strukturanalysen von Neophorbol, C31H35O9Br, und Phorbol,
C
20H28O6. Acta Crystallogr. 1971, B27, 1718-1730.
(14) Allen, F. H.; Bellard, S.; Brice, M. D.; Cartwright, B. A.;
Doubleday, A.; Higgs, H.; Hummelink, T.; Hummelink-Peters,
B. G.; Kennard, O.; Motherwell, W. D. S.; Rodgers, J . R.; Watson,
D. G. The Cambridge Crystal Data Centre: Computer-Based
Search, Retrieval, Analysis, and Display of Information. Acta
Crystallogr. 1979, B35, 2331-2339.
(9) Fukuzawa, S.; Nakanishi, A.; Fujinami, T.; Sakai, S. Reductive
Coupling of Ketone or Aldehydes with Electron-deficient Alkenes
Promoted by Samarium Di-iodide. J . Chem. Soc., Chem. Com-
mum. 1986, 624-625.
(10) Gescher, A.; Dale, I. L. Protein Kinase C. A Novel Target for
Rational Anti-Cancer Drug Design? Anti-Cancer Drug Des. 1989,
4, 93-105.
J M950277N