Enantioselective total synthesis of pinnaic acid and halichlorine

Article abstract of DOI:10.1002/asia.201301248

The enantioselective total synthesis of the bioactive marine natural products pinnaic acid and halichlorine is reported in detail. Our total synthesis features the construction of the five-membered ring and C9 and C13 stereogenic centers through a palladium-catalyzed trimethylenemethane [3+2] cyclization; the installation of the nitrogen atom through a regioselective Beckmann rearrangement of a poorly reactive ketone; the stereoselective cyclization of the spiro ring through a four-step, one-pot hydrogenation- cyclization; and efficient connection of the sterically hindered lower chain through a reduced-pressure cross olefin metathesis reaction. Copyright

Full text of DOI:10.1002/asia.201301248

FULL PAPER
DOI: 10.1002/asia.201301248
Enantioselective Total Synthesis of Pinnaic Acid and Halichlorine
Shu Xu,[a] Daisuke Unabara,[a] Daisuke Uemura,[b] and Hirokazu Arimoto*[a]
Abstract: The enantioselective total
synthesis of the bioactive marine natu-
ral products pinnaic acid and halichlor-
ine is reported in detail. Our total syn-
thesis features the construction of the
five-membered ring and C9 and C13
stereogenic centers through a palladi-
[3+2] cyclization; the installation of
the nitrogen atom through a regioselec-
tive Beckmann rearrangement of
a poorly reactive ketone; the stereose-
lective cyclization of the spiro ring
through a four-step, one-pot hydroge-
nation–cyclization; and efficient con-
nection of the sterically hindered lower
chain through a reduced-pressure cross
olefin metathesis reaction.
Keywords: cyclization · metathesis ·
natural products · stereoselectivity ·
total synthesis
um-catalyzed
trimethylenemethane
Introduction
The architectural 6-aza-spiroACTHUGNETRNNU[G 4.5]decane structures of
these two molecules are even more impressive than their
bioactivities. They have attracted considerable attention in
the synthetic chemistry community and have been the topic
of specific reviews[7] and a large number of synthetic stud-
ies.[8] However, significant problems still exist and, since
Danishefskyꢀs first total synthesis of both chiral molecules,[9]
only two asymmetric total syntheses of 1 from our[10] and
Zhaoꢀs group,[11] and one asymmetric recent total synthesis
of 2 from Cliveꢀs group,[12] have been reported. Construction
of the contiguous stereogenic centers at C9, C13, and C14 is
a challenge, and thus, many reported examples, including
the elegant total synthesis by Heathcock and Christie,[13] in-
volved the preparation of only the spiro framework in race-
mic form. Another significant problem is the addition of the
side chains to the spiro core. All of the reported total syn-
theses suffered low yields for the lower-chain connection
and/or poor selectivity in the subsequent C17 stereogenic
center installation.[9–13] In fact, none of the formal total syn-
theses reported to date, attempted this chain connection.[14]
In continuation of earlier reports,[10] we herein disclose
our recent efforts, which have culminated in the enantiose-
lective total synthesis of 1 and 2. Our total synthesis features
1) construction of the five-membered ring and C9 and C13
stereogenic centers through a palladium-catalyzed trimethy-
lenemethane (TMM) [3+2] cyclization; 2) installation of the
nitrogen atom through a regioselective Beckmann rear-
rangement of a poorly reactive ketone by using a modified
reaction protocol; 3) stereoselective cyclization of the spiro
ring through a one-pot, four-step hydrogenation–cyclization;
and 4) efficient connection of the sterically hindered lower
chain through a reduced-pressure cross olefin metathesis re-
action.
In 1996, one of us reported the isolation of a novel class of
marine natural products represented by pinnaic acid (1)[1]
and halichlorine (2).[2] Compound 1 was obtained from the
Okinawan bivalve Pinna muricata. It inhibits a cytosolic
85 kDa phospholipase (cPLA2)[3] in vitro.[1] Compounds that
inhibit cPLA2 have been targeted as anti-inflammatory
agents. Compound 2 was obtained from the black marine
sponge Halichondria okadai Kadota, and shows bioactivity
for reducing the induced mRNA expression of adhesion
molecules (VCAM-1, ICAM-1, E-selectin)[4] and induced
monocyte (U937) adhesion to endothelial cells by attenuat-
ing NF-kB activity,[2,5] which may be useful for treating athe-
rosclerosis. Compound 2 also inhibits L-type Ca2+ channels
in vascular smooth muscle cells, making it promising as an
antihypertensive agent.[6]
[a] Dr. S. Xu, D. Unabara, Prof. Dr. H. Arimoto
Graduate School of Life Sciences
Tohoku University, 2-1-1 Katahira
Sendai 980-8577 (Japan)
Fax : (+81)22-217-6204
E-mail: arimoto@biochem.tohoku.ac.jp
[b] Prof. Dr. D. Uemura
Department of Chemistry, Faculty of Science
Kanagawa University, Hiratsuka 259-1293 (Japan)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/asia.201301248.
Chem. Asian J. 2014, 9, 367 – 375
367
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
Results and Discussion
9, which could be synthesized by Pd-TMM [3+2] cycliza-
tion.[16] This cyclization was anticipated to occur from the
opposite face of the adjacent methyl group in cyclopente-
none 10 to construct the three contiguous stereogenic cen-
ters diastereoselectively. The chiral compound 10 could be
prepared from commercially available (R)-(+)-pulegone.
Retrosynthetic Analysis of 1 and 2
Because of the abovementioned difficulty with the installa-
tion of the lower chain, we planned to employ olefin meta-
thesis to form the C15=C16 bond. Thus, the C17 center
could be preinstalled in the lower-chain fragment with defi-
nite stereochemistry. Because 1 and 2 are proposed to be
biosynthetically related[15] and share the same structural
backbone, the common intermediate 3 was designed and ex-
pected to lead to 1 through cross metathesis with the upper-
chain unit 4 (A=hydrogen) and chiral lower-chain unit 5
(Scheme 1). By using the same compound 5 and a modified
upper-chain unit (A=leaving group), intermediate 3 could
Preparation of Chiral Building Block 10
From the retrosynthetic analysis, (S)-2-alkyl-4-methylcyclo-
pent-2-enone (10) was our first target. When this study was
initiated, there were no reports on methods for preparation
of this type of building block in chiral form.[17] Thus, a short
route from (R)-(+)-pulegone was designed by using
Dieckmann condensation[18] and Tsujiꢀs decarboxylation–de-
hydrogenation[19] reactions.
Ozonolysis[20] of technical grade (R)-pulegone[21] cleanly
gave the dicarboxylic acid 11, which was esterified to afford
diallyl ester 12 in 89% yield (2 steps, 40 gram scale,
Scheme 2). Our synthesis of 11 is superior to other literature
procedures[22] because of the metal-free conditions and easy
workup. Dieckmann condensation[18] of 12 with potassium
metal as the base gave the mixed enolates 13a and 13b
(ꢀ2:1), which in situ reacted with the iodide 14 to afford
a mixture of 15 and 16. The combination of the potassium
salt and alkyl iodide provided the best C- (15a, b) to O-al-
kylation (16a, b) ratio (ꢀ9:1), as expected based on hard
and soft acids and bases (HSAB) theory.[23] Because the by-
products 16a and 16b interrupted the subsequent palladi-
um-catalyzed reaction and were difficult to separate from
compounds 15a and 15b, hydrolysis was employed to trans-
form them into the corresponding protonated compounds
13a and 13b. The hydrolyzed mixture was then subjected to
Tsuji decarboxylation–dehydrogenation,[19] which afforded
compound 10 in 49% yield (3 steps, 20 gram scale), together
with isolated byproduct 17 derived from 13b. The yield of
this decarboxylation–dehydrogenation was largely depen-
Scheme 1. Retrosynthetic analysis of
TBDPS=tert-butyldiphenylsilyl, PMP=p-methoxyphenyl.
1 and 2. Cbz=carboxybenzyl,
also afford tetraene 6; a precur-
sor designed to provide
2
through a one-pot double ring-
closing metathesis (RCM) pro-
cess.
The spiro core of 3 was ex-
pected to be prepared from the
unsaturated ketone 7 through
a
diastereoselective, one-pot,
four-step hydrogenation–cycli-
zation developed in our previ-
ous synthetic studies.[10c] Inter-
mediate
from compound
7
could be formed
through
8
a Horner–Wadsworth–Emmons
olefination (HWE reaction).
The nitrogen atom of 8 would
be installed through a regiose-
lective Beckmann rearrange-
Scheme 2. Synthesis of the chiral building block 10: a) O3, 08C, HOAc, H2O, EtOAc; b) AllylOH, H2SO4
(cat.), Dean–Stark, reflux, 89% (2 steps); c) K, AllylOH, PhMe, reflux; then 14, reflux; d) HCl (aq), acetone,
ment from the bicyclic ketone RT; e) PdACHTUNTRGNE(NUG OAc)2 (5 mol%), MeCN (4m), 908C, 49% (3 steps).
Chem. Asian J. 2014, 9, 367 – 375
368
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
dent on the concentration; a higher concentration favored
a higher conversion and fewer byproducts. However, neat
conditions led to a decrease in the yield owing to the diffi-
culty in achieving efficient stirring.
employed highly reactive substrates (sterically unhindered
ketones or strained cyclobutanones), and the solution of
MSH/dichloromethane was added to the solution of the
ketone substrate. However, we observed that MSH rapidly
self-decomposed in dichloromethane at room temperature
particularly at high concentrations, and sometimes the solu-
tion refluxed on its own.[26] This self-decomposition may
compete with the formation of the desired oxime when
using hindered low-reactive substrates. Thus, to avoid de-
composition of MSH, the experimental procedure was modi-
fied to involve the direct addition of solid MSH to the solu-
tion of ketone 9 (Table 1, entry 7). Gratifyingly, this ap-
proach led to clean, full, and reproducible conversion of 9
to 20 in 85% yield on a 5 gram scale.
Pd-TMM [3+2] Cyclization
With the chiral building block 10 in hand, we attempted the
key Pd-TMM [3+2] cyclization (Scheme 3).[16] As anticipat-
ed, compound 10 reacted with the TMM precursor 19 highly
selectively, and only the diastereomer 9 was detected (80%
yield).[23] The configuration of the C9 and C13 centers was
verified by NOE analysis of 9.
Preparation of 28 and Subsequent One-Pot, Four-Step
Hydrogenation–Cyclization
For the synthesis of the spiro core, cleavage of the exo-
olefin bond was required (Table 2). Ozonolysis of compound
20 smoothly afforded ketone 22. However, the next deoxy-
genation reaction proved to be troublesome. The thioacetal–
Raney nickel method (Table 2, entry 1) led to partial remov-
al of PMP (plausibly through benzene-ring hydrogenation
and hydrolysis). The NH2NHTs–NaBH4 method[27] (Table 2,
entry 2) decomposed the lactam ring. The NH2NHTs–
NaBH3CN method[28] (Table 2, entry 3) only gave a modest
yield. Although the Yamamura-modified Clemmensen re-
duction[29] (Table 2, entry 4) worked well, it was not repro-
ducible on a larger scale owing to the poor solubility of 22
in Et2O. Finally, the use of a combination of TMSCl and
H2O in THF afforded a reproducibly good result on the
gram scale (Table 2, entry 5). Notably, Yamamura et al. re-
ported that a solution of HCl in THF was not suitable for
this transformation.[29b] Based on these results, we also de-
veloped a one-pot ozonolysis–Clemmensen reduction pro-
cess[30] that provided 8 directly from 20 in 72% yield.
Protection of 8 with Cbz led to compound 24 (Scheme 4).
The next reductive ring opening was first attempted with
NaBH4,[31] which proceeded well, but was contaminated with
a small amount of solvolyzed
Scheme 3. Key Pd-TMM [3+2] cyclization of compound 10. THF=tetra-
hydrofuran, TMS=trimethylsilyl.
Beckmann Rearrangement with a Modified MSH
Manipulation
For installation of the nitrogen atom, the Schmidt reac-
tion[24] and Beckmann rearrangement[25] were tested with
substrate 9 (Table 1). Schmidt conditions (Table 1, entry 1)
led to a complex mixture. The standard two-step Beckmann
procedure with NH2OH[25a] (Table 1, entry 2) proceeded
well, but only with 3:1 regioselectivity, presumably owing to
weak stereocontrol during formation of the oxime. Thus,
a hindered nitrogen reagent, MSH,[25b,c] was tested (Table 1,
entries 3–7). As expected, only the desired isomer 20 was
detected. However, the conversion of the reaction was
never clean and sometimes not reproducible. Although
larger amounts of MSH did provide higher conversion, the
enhancement of the yield was limited. A literature screen
revealed that all reported MSH–Beckmann rearrangements
byproduct.[23] Thus, LiBH4 in
Table 1. Nitrogen atom installation in compound 9.[a]
the aprotic solvent THF was
tested, which yielded 25 quan-
titatively. Next, protection with
TBDPS, removal of PMP,
Parikh–Doering oxidation, and
the HWE reaction with 27 af-
forded spiro-cyclization precur-
sor 28 as the single trans
isomer.[32]
Entry
Conditions
Results[b]
1
2
NaN3 (5.0 equiv), MeSO3H, MeOCH2CH2OMe, RT
i) NH2OH·HCl (5 equiv), DABCO (1.1 equiv), MeOH, RT
ii) TsCl (1.2 equiv), NaOH (aq), acetone, reflux
MSH (1.0 equiv), CH2Cl2, RT; then silica gel
MSH (1.5 equiv), MS4ꢂ, alumina, CH2Cl2, RT
MSH (5.0 equiv), CH2Cl2, RT; then silica gel
MSH (10.0 equiv), CH2Cl2, RT; then silica gel
MSH (1.5 equiv, solid), CH2Cl2, RT; then silica gel
complex mixture
84% in 2 steps
(20/21=3:1)
20 (35%), 9 (50%)
20 (43%), 9 (57%)
20 (60%), 9 (1%)
20 (63%), 9 (0%)
20 (85%), 9 (0%)
3
4
5
6
7
Our one-pot hydrogenation–
cyclization[10] is
a powerful
protocol for constructing the
piperidine ring of 1 and 2 with
the desired C5 configuration
(Scheme 4). It consists of four
[a] DABCO=1,4-diazabicycloACTHNUTRGNE[NUG 2.2.2]octane, Ts=p-toluenesulfonyl, MSH=O-mesitylenesulfonylhydroxyla-
mine, MS4ꢂ=4 ꢂ molecular sieves. [b] Yield of isolated product.
Chem. Asian J. 2014, 9, 367 – 375
369
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
Table 2. Cleavage of the exo-olefin bond of 20.
basic amino group of product
31 seemed to be the reason for
the high catalyst loading; it may
deter the Cbz hydrogenolysis
step.[34] Thus, when excess
HOAc was used, only 20 mol%
of the palladium catalyst was
sufficient to cleanly and repro-
ducibly afford the desired dia-
stereomer 31. Next, TFA pro-
tection of the amino group, se-
lective removal of TBS,[35] and
Grieco elimination[36] led to ter-
minal olefin 33, which exhibited
NOE correlations that con-
firmed the desired stereochem-
istry at C5.
Entry
1
Conditions
Results[a] ([%])
i) HSCH2CH2SH, TiCl4, À158C, CH2Cl2;
ii) Raney nickel, EtOH, reflux
8 (35), 23 (17)
2
3
4
5
i) NH2NHTs, MeOH, reflux; ii) NaBH4, MeOH, reflux
i) NH2NHTs, MeOH, reflux; ii) NaBH3CN, TsOH, THF, reflux
activated Zn, HCl (sat. in Et2O), 08C
decomposed
8 (60)
8 (80), low reproducibility
8 (86), reproducible on gram scale
Zn, TMSCl-H2O (5:2), THF, 58C
[a] Yield of isolated product. For two-step procedures, the yield is the overall yield.
Synthesis of the Lower-Chain
Fragment
As mentioned in the Introduc-
tion, the stereoselective con-
struction of the C17 center is
difficult during or after installa-
tion of the lower chain. There-
fore, we decided to construct
the C17 stereogenic center in
the
lower-chain
fragment
before the coupling reaction.
After failed attempts with
Sharpless
desymmetrization
and Brown asymmetric allyla-
tion,[23] the Corey–Bakshi–Shi-
bata (CBS) method[37] became
our
method
of
choice
(Scheme 5). Prechiral ketone 36
was readily prepared from bu-
tynol 34 through TBDPS pro-
tection, addition to acrolein,[38]
and IBX oxidation.[39] For the
asymmetric reduction, the com-
mercially available (R)-Me-
CBS chiral catalyst was used.
Scheme 4. Preparation of the spiro core: a) NaH, CbzCl, THF, reflux, 87%; b) NaBH4, LiBr, THF, 508C,
quant.; c) TBDPSCl, DMAP, Et3N, CH2Cl2, RT; d) CAN, MeCN, H2O, 0 8C, 79% (2 steps); e) SO3·py, Et3N,
DMSO, RT; f) 27, Et3N, LiCl, THF, 308C, quant. (2 steps); g) H2, Pd(OH)2/C (20 mol%), HOAc (2 equiv),
EtOH, 258C; h) TFAA, iPr2NEt, CH2Cl2, 08C, 81% (2 steps); i) PPTS, EtOH, 258C; j) o-NO2PhSeCN, nBu3P,
THF, RT; then mCPBA, RT, 88% (2 steps). CAN=(NH4)2CeACHTNUTRGNEUNG(NO3)6, DMAP=4-dimethylaminopyridine, py=
pyridine, DMSO=dimethylsulfoxide, TBS=tert-butyldimethylsilyl, TFAA=trifluoroacetic anhydride, TFA=
trifluroacetyl, PPTS=pyridium p-toluenesulfonate, mCPBA=m-chloroperbenzoic acid.
consecutive transformations: 1) saturation of the olefin,
2) removal of the Cbz group, 3) intramolecular cyclic imine
and/or enamine formation, and 4) stereoselective reduction
of the imine/enamine intermediate. The stereoconformation
of 30 clearly demonstrated that attack from the right side
(which is also the b face of 30) would be disfavored because
of steric hindrance of the C22 methyl group. Similar reduc-
tive cyclizations have since then also been conducted by
others.[11,13] In our earlier studies,[10b–d] when this reaction
was carried out under neutral or acidic conditions with a cat-
alytic amount of HOAc,[33] typically 50 mol% of palladium
catalyst was required to obtain reproducible results. Careful
examination of the reaction conditions revealed that the
After initially suffering from low reproducibility, a procedure
involving the slow addition of 0.6 equivalents of Me2S·BH3
in THF to a premixed solution of 36 and 0.1 equivalents of
Me–CBS in THF ensured an acceptable yield with 90% ee.
Notably, Noyori asymmetric transfer hydrogenation[40] gave
no reaction with substrate 36. To the best of our knowledge,
our result was the first example of the catalytic asymmetric
reduction of an alkenyl–alkynyl–ketone substrate. Next, in-
stallation of chlorine by using the Red-Al/NCS method pro-
À
vided the chiral C15 C21 fragment 37a in a total of just
five steps from inexpensive, commercially available starting
materials. The enantiomeric excess (ee) of 37a was con-
firmed again to ensure that no racemization occurred during
Chem. Asian J. 2014, 9, 367 – 375
370
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
that in lower-chain unit 37. The group of Grubbs summar-
ized experience-based rules for the cross olefin metathe-
sis.[43] They categorized olefin substrates into four types with
different reactivity and predicted the products and reaction
difficulty according to the “types” of the substrates. Thus,
according to Grubbsꢀ theory, the terminal olefin in com-
pound 44 should be a “Type I substrate” (the most reactive
category, fast homodimerization), and the secondary allylic
alcohols 37 (protected and unprotected) should be “Type II
substrates” (less reactive, slow homodimerization). Howev-
er, under all of our tested conditions, the dimer of 44 was
never detected and the dimer of TBS-protected 37b always
appeared in considerable quantities (the free alcohol 37a
decomposed under the reaction condition, entry 1). Thus,
substrate 44 is less reactive than 37b for cross metathesis
and should belong to Grubbsꢀ “Type III substrate” (least re-
active, no homodimerization). The low reactivity of 44 is
thought to be due to high steric hindrance from the nearby
spiro core. It is well known that for cross metathesis be-
tween Type II and Type III olefins, a large excess of the
Type III olefin is required to suppress the competing dimeri-
zation of Type II olefins.[43] However, this approach was not
feasible for our substrates because Type III olefin 44 is
much more precious than the readily prepared lower-chain
unit 37. Therefore, as shown in Table 3, entries 2–4, even
Scheme 5. Preparation of the lower-chain fragment through CBS asym-
metric reduction: a) TBDPSCl, Et3N, DMAP, CH2Cl2, RT, 99%;
b) nBuLi, LiBr, THF, RT; then acrolein, À788C, 80% (95% brsm);
c) IBX, EtOAc, reflux, quant.; d) (R)-Me-CBS (0.1 equiv), BH3·Me2S
(0.6 equiv), THF, À408C, 66% (90% ee); e) Red-Al, THF, RT; then
NCS, À788C, 65% (90% ee); f) TBSOTf, Et3N, CH2Cl2, RT, 84%;
g) TBDPSCl, Et3N, DMAP, CH2Cl2, RT, quant. CBS=Corey–Bakshi–
Shibata catalyst, brsm=based on recovered starting material, IBX=2-io-
doxybenzoic acid, Red-Al=sodium bis(2-methoxyethoxy) aluminum hy-
dride, NCS=N-chlorosuccinimide, Tf=trifluoromethanesulfonyl.
the chlorination step. Protection of 37a with TBS and
TBDPS afforded 37b and 37c, respectively.
Cross Olefin Metathesis for the Total Synthesis of 1
For the synthesis of 1, the upper chain was connected to
compound 33 through a cross olefin metathesis reaction
(Scheme 6). With the less reac-
tive counterpart, 38, as the sol-
vent and 8 mol% Hoveyda–
Grubbs II catalyst,[41] 81%
yield of olefin 41 was obtained
exclusively in the desired E
configuration, together with
5% yield of homodimer 42.
With the Grubbs II catalyst,[42]
phenyl-substituted byproduct
43 was also detected. When
isolated 42 and 43 were sub-
jected again to the above
metathesis conditions, they
reached an equilibrium with
41, but we did not pursue this
approach for preparative pur-
poses. Removal of TBDPS and
Grieco elimination generated
the terminal olefin 44.
Scheme 6. Cross metathesis installation of the upper chain and preparation of 44: a) 40 (8 mol%), 38/toluene
(1:1), 1008C, 81%; b) HF·(py)x, py, 458C; c) o-NO2PhSeCN, nBu3P, THF, RT; then mCPBA, RT, 90%
(2 steps). Mes=mesityl, Cy=cyclohexyl.
Table 3. Cross metathesis installation of the lower chain of 1.
The cross metathesis of 44
was tested with the lower-
chain units 37a–c (Table 3).
Four carbon–carbon double
bonds exist in the substrates.
The two tri-substituted double
bonds were expected[43] and
proved to be inert under the
tested conditions. However,
the monosubstituted olefin in
compound 44 showed unex-
pected lower reactivity than
Entry
37 ([equiv])
Cat. ([mol%])
Conditions
Results ([%])[a]
1
2
3
4
5
6
7
37a (6.2)
37b (6.8)
37b (4.0)
37b (5.5)
37c (5.0)
37c (5.0)
37c (6.0)
39 (20)
40 (20)
39 (50)
39 (80)
39 (44)
39 (29)
39 (50)
PhMe, 908C, 6 h
neat, 908C, 12 h
neat, 908C, 6 h
PhMe, 908C, 24 h
45a (0), 44 (90)
45b (7), 44 (80)
45b (20), 44 (70)
45b (40), 44 (42)
45c (5), 44 (90)
45c (100), 44 (0)
45c (69), 44 (0)
neat, 908C, 10 h
neat, 908C, 20 mmHg, 6.5 h
C6H3Cl3,[b] 908C, 20 mmHg, 22 h
[a] Yield of isolated product. [b] 1,2,4-Triclorobenzene, which is a high-boiling-point solvent (2158C, 1 atm).
Chem. Asian J. 2014, 9, 367 – 375
371
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
when excess 37b was used, the conversion of 44 was low
and 80 mol% of catalyst had to be employed to afford
a moderate yield, as reported in our preliminary communi-
cation (Table 3, entry 4).[10a]
To avoid undesired homodimerization, the cross metathe-
sis using TBDPS-protected 37c was attempted (Table 3, en-
tries 5–7). Compound 37c is bulkier around the olefin than
TBS-protected 37b. As expected, it showed lower metathe-
sis reactivity and only tiny amounts of its dimer could be de-
tected. However, owing to the low reactivity of both 37c
and 44, the conversion of the reaction remained low
(Table 3, entry 5). Thus, forcing conditions were applied
with reduced pressure for removal of the in situ generated
ethylene and styrene (Table 3, entries 6 and 7). To our de-
light, under these conditions, full conversion was achieved,
and the fully protected pinnaic acid 45c was obtained in
quantitative yield under neat conditions (Table 3, entry 6).
The use of high vacuum was reported originally by the
group of Grubbs for the dimerization of unhindered termi-
nal olefins[44a] and recently by the groups of Schrock[44b] and
Grubbs[44c] to achieve cis-selective metathesis. It should be
noted that under all of the conditions tested for our sub-
strates, only the desired trans isomer (C15=C16 bond) of 45
was formed.
Scheme 8. Preparation of compound 48 and its RCM: a) NH4F, MeOH,
608C; b) o-NO2PhSeCN, nBu3P, THF, RT; then mCPBA, RT, quant.
(2 steps); c) LiBH4, THF, 508C; d) 47, K2CO3, MeCN, 808C, 76% (2
steps).
unexpected cyclization is thought to occur because of the
spatially neighboring C3 and C15 centers protruding from
the spiro core. All attempts to transform 48 or 50 into the
desired 49 by using a variety of metathesis conditions failed.
Therefore, we changed our idea from the one-pot RCM
strategy to a stepwise RCM strategy.
Deprotection of 45c successively with HF·py (2ꢃ
TBDPS), NaBH4 (TFA amide), and basic hydrolysis (ethyl
ester), afforded chiral 1 in its sodium salt form (66%,
2 steps, Scheme 7), which was identical to the synthetic race-
mic sodium salt of 1[13] based on a comparison of the
1H NMR spectra.
The stepwise RCM strategy commenced with compound
33 (Scheme 9), which smoothly afforded compound 53 after
removal of TFA, alkylation with bromide 51, and six-mem-
bered RCM. Next, removal of TBDPS, Grieco elimination,
acidic hydrolysis of the tert-butyl ester, and condensation
with lower-chain unit 54 by using the Shiina reagent[45] pro-
vided the macrocyclic RCM precursor 55. However, precur-
sor 55 proved to be inert to metathesis conditions,[23] and
surprisingly, contrary to our experience in the study of the
cross metathesis of 1, the TBS-protected lower chain was
less reactive than the C15 olefin. It is well known that free
amino groups usually inhibit metathesis through coordina-
tion to the ruthenium atom in the catalyst.[46] However, with
substrates 48 and 52, metathesis did occur in the presence of
tertiary amines. Thus, it was reasoned that, for substrate 55,
Scheme 7. Synthesis of 1 by deprotection of 45c: a) HF·(py)x, py, 258C;
b) NaBH4, EtOH, 258C; then NaOH (aq), 258C to 408C, 66% (2 steps).
Macrocyclic RCM toward 2
In Scheme 1, tetraene 6 was designed as the precursor for 2
through two RCMs: a six-membered RCM and a macrocyclic
RCM. Because the lower-chain metathesis showed lower re-
activity than the upper-chain reaction in the total synthesis
of 1, it was expected that the two RCMs could be achieved
in one pot. Thus, the six-membered RCM in the presence of
C15=C16 terminal olefin was first explored by using model
compound 48 (Scheme 8), which was readily prepared from
32 through double silyl deprotection, double Grieco elimina-
tion, removal of TFA, and alkylation with the upper-chain
unit 47. The RCM of 48 was tested with Hoveyda–Grubbs II
catalyst 40 (10 mol%, 608C in toluene, similar conditions to
those used by the Kibayashi group[14a]). However, the de-
sired product 49 was not detected at all and compound 48
was fully converted into the C3=C15 RCM product 50. This
Scheme 9. Preparation of compound 55 and its RCM: a) NaBH4, EtOH,
RT; b) 51, K2CO3, MeCN, 908C; c) NH4F, MeOH, 608C; d) o-
NO2PhSeCN, nBu3P, THF, RT; then H2O2, RT, 85% (2 steps);
e) HCO2H, RT; f) 54, MNBA, DMAP, RT, CH2Cl2, 80% (2 steps).
MNBA=2-methyl-6-nitrobenzoic anhydride.
Chem. Asian J. 2014, 9, 367 – 375
372
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
coordination of both the decalin nitrogen and C15 olefin to
the ruthenium atom might form a stable complex and deac-
tivate the catalyst.
Compound 59 was prepared from the synthetic intermedi-
ate 26 (Scheme 11) through a sequence that including
a Parikh–Doering oxidation, the HWE reaction with 60
(quantitative, 2 steps), our key one-pot hydrogenation–cycli-
zation, TFA protection (80%, 2 steps), TBDPS deprotection
(quantitative), Grieco elimination (99%), and MOM depro-
tection (86%). The cross metathesis of 59 then proceeded
smoothly with the lower-chain unit 37c under static vacuum
conditions[44] to afford compound 58 in 72% yield. Next,
Grieco elimination (84%), deprotection of TFA, and alkyla-
tion with bromo-substituted upper-chain unit 47 (84%,
2 steps) gave compound 57. Six-membered RCM of 57
under the conditions reported by Kibayashi et al.[14a] afford-
ed the tricyclic compound 63
Revised Strategy for the Total Synthesis of 2
Given the above results, we speculated that for the synthesis
of 2 the lower-chain metathesis must be carried out with
protection of the amino group, and therefore, before the
upper-chain RCM reaction. Thus, the retrosynthesis of 2 was
revised to macrolactonization, RCM, and alkylation, which
led to the key intermediate 58, possibly prepared by cross
metathesis of compounds 37c and 59 (Scheme 10).
(quantitative). Deprotection of
the two TBDPS groups (93%),
basic hydrolysis of the methyl
ester, and Shiina macrolactoni-
zation[45b] (56%, 2 steps) com-
pleted the total synthesis of 2.
Conclusion
Scheme 10. Revised retrosynthetic analysis of 2.
We described the enantioselec-
tive total synthesis of 1 and 2
in detail. From a new chiral
building block, 10, the azaspiro core was constructed stereo-
selectively by using Pd-TMM [3+2] cyclization, a modified
MSH–Beckmann rearrangement, and a one-pot four-step
hydrogenation–cyclization. The C17 center was preinstalled
to the lower-chain fragment through catalytic CBS reduc-
tion. The lower-chain connection was achieved effectively
through reduced-pressure cross metathesis of sterically hin-
dered olefins. In addition, during our synthetic journey,
a TMSCl-modified Yamamura–Clemmensen reduction and
its use in a one-pot procedure to cleave olefins to alkanes
were also developed.
Experimental Section
The synthetic procedures and characterization of the compounds studied
herein can be found in the Supporting Information.
Acknowledgements
Scheme 11. Total synthesis of 2: a) SO3·py, Et3N, DMSO, RT; b) 60, Et3N,
LiCl, THF, 308C, quant. (2 steps); c) H2, Pd(OH)2/C (20 mol%), HOAc
(2 equiv), EtOH, RT; d) TFAA, iPr2NEt, CH2Cl2, 08C, 80% (2 steps);
e) TBAF, THF, 08C, quant.; f) o-NO2PhSeCN, nBu3P, THF, RT; then
mCPBA, RT, 99%; g) PPTS, tBuOH, 908C, 86%; h) 37c (4.3 equiv), 40
(28 mol%), PhMe, static vacuum, 608C, 72%; i) o-NO2PhSeCN, nBu3P,
THF, RT; then H2O2, RT, 84%; j) NaBH4, EtOH, 308C; k) 47, K2CO3,
MeCN, 608C, 84% (2 steps); l) 39 (10 mol%), CH2Cl2, reflux, quant.;
m) HF·(py)x/py (1:3), RT, 93%; n) NaOH, THF, MeOH, H2O, 50 8C;
o) MNBA, DMAP, THF, RT, 56% (2 steps). MOM=methoxymethyl,
TBAF=tetra-n-butylammonium fluoride.
We are grateful for the financial support of Grants-in-Aid for Scientific
Research (16GS0206 and 16310150) from the JSPS.
[1] T. Chou, M. Kuramoto, Y. Otani, M. Shikano, K. Yazawa, D.
Uemura, Tetrahedron Lett. 1996, 37, 3871–3874.
[2] a) M. Kuramoto, T. Chou, K. Yamada, T. Chiba, Y. Hayashi, D.
Uemura, Tetrahedron Lett. 1996, 37, 3867–3870; b) H. Arimoto, I.
Hayakawa, M, Kuramoto, D. Uemura, Tetrahedron Lett. 1998, 39,
861–862.
Chem. Asian J. 2014, 9, 367 – 375
373
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
[3] G. Y. Sun, P. B. Shelat, M. B. Jensen, Y. He, A. Y. Sun, A. Simonyi,
NeuroMol. Med. 2010, 12, 133–148.
[4] J. M. Cook-Mills, M. E. Marchese, H. Abdala-Valencia, Antioxid.
Redox Signaling 2011, 15, 1607–1638.
[5] Y. Tsubosaka, T. Murata, K. Yamada, D. Uemura, M. Hori, H.
Ozaki, J. Pharmacol. Sci. 2010, 113, 208–213.
[6] Y. Tsubosaka, T. Murata, K. Kinoshita, K. Yamada, D. Uemura, M.
Hori, H. Ozaki, Eur. J. Pharmacol. 2010, 628, 128–131.
[7] a) D. L. J. Clive, M. Yu, J. Wang, V. S. C. Yeh, S. Kang, Chem. Rev.
2005, 105, 4483–4514; b) G. Dake, Tetrahedron 2006, 62, 3467–
3492; c) C. Kibayashi, S. Aoyagi, J. Synth. Org. Chem. Jpn. 2011, 69,
1006–1019.
[11] a) H. Wu, H. Zhang, G. Zhao, Tetrahedron 2007, 63, 6454–6461;
b) H. L. Zhang, G. Zhao, Y. Ding, B. Wu, J. Org. Chem. 2005, 70,
4954–4961.
[12] a) D. Z. Liu, H. P. Acharya, M. L. Yu, J. Wang, V. S. C. Yeh, S. Z.
Kang, C. Chiruta, S. M. Jachak, D. L. J. Clive, J. Org. Chem. 2009,
74, 7417–7428; b) D. L. J. Clive, V. S. C. Yeh, Tetrahedron Lett. 1999,
40, 8503–8507; c) D. L. J. Clive, J. Wang, M. Yu, Tetrahedron Lett.
2005, 46, 2853–2855; d) M. Yu, D. L. J. Clive, V. S. C. Yeh, S. Kang,
J. Wang, Tetrahedron Lett. 2004, 45, 2879–2881; e) D. L. J. Clive, M.
Yu, Z. Li, Chem. Commun. 2005, 906–908.
[13] H. Christie, C. H. Heathcock, Proc. Natl. Acad. Sci. USA 2004, 101,
12079–12084.
[8] a) S. P. Keen, S. M. Weinreb, J. Org. Chem. 1998, 63, 6739–6741;
b) M. J. Martꢄn-Lꢅpez, F. Bermejo, Tetrahedron 1998, 54, 12379–
12388; c) S. Lee, Z. Zhao, Org. Lett. 1999, 1, 681–683; d) S. Lee, Z.
Zhao, Tetrahedron Lett. 1999, 40, 7921–7924; e) J. L. Koviach, C. J.
Forsyth, Tetrahedron Lett. 1999, 40, 8529–8532; f) M. Shindo, Y.
Fukuda, K. Shishido, Tetrahedron Lett. 2000, 41, 929–932; g) D. L.
Wright, J. P. Schulte II, M. A. Page, Org. Lett. 2000, 2, 1847–1850;
h) J. D. White, P. R. Blakemore, E. A. Korf, A. F. T. Yokochi, Org.
Lett. 2001, 3, 413–415; i) W. Yokota, M. Shindo, K. Shishido, Het-
erocycles 2001, 54, 871–885; j) T. Ito, N. Yamazaki, C. Kibayashi,
Synlett 2001, 10, 1506–1510; k) S. Ciblat, J.-L. Canet, Y. Troin, Tetra-
hedron Lett. 2001, 42, 4815–4817; l) K. Takasu, H. Ohsato, M. Ihara,
Org. Lett. 2003, 5, 3017–3020; m) Y. Matsumura, S. Aoyagi, C. Ki-
bayashi, Org. Lett. 2003, 5, 3249–3252; n) P. B. Hurley, G. R. Dake,
Synlett 2003, 14, 2131–2134; o) G. R. Dake, M. D. B. Fenster, P. B.
Hurley, B. O. Patrick, J. Org. Chem. 2004, 69, 5668–5675; p) L. G.
Arini, P. Szeto, D. L. Hughes, R. A. Stockman, Tetrahedron Lett.
2004, 45, 8371–8374; q) T. Huxford, N. S. Simpkins, Synlett 2004, 13,
2295–2298; r) K. S. Feldman, A. L. Perkins, K. M. Masters, J. Org.
Chem. 2004, 69, 7928–7932; s) A. L. de Sousa, R. A. Pilli, Org. Lett.
2005, 7, 1617–1619; t) E. Roulland, A. Chiaroni, H.-P. Husson, Tet-
rahedron Lett. 2005, 46, 4065–4068; u) D. G. Hilmey, L. A. Paquette,
Org. Lett. 2005, 7, 2067–2069; v) D. G. Hilmey, J. C. Gallucci, L. A.
Paquette, Tetrahedron 2005, 61, 11000–11009; w) H. Kim, J. H. Seo,
K. J. Shin, D. J. Kim, D. Kim, Heterocycles 2006, 70, 143–146;
x) S. H. Yang, V. Caprio, Synlett 2007, 1219–1222; y) P. B. Hurley,
G. R. Dake, J. Org. Chem. 2008, 73, 4131–4138; z) R. J. Bahde, S. D.
Rychnovsky, Org. Lett. 2008, 10, 4017–4020; aa) G. E. Keck, S. A.
Heumann, Org. Lett. 2008, 10, 4783–4786; ab) S. H. Yang, G. R.
Clark, V. Caprio, Org. Biomol. Chem. 2009, 7, 2981–2990; ac) I. D.
Jenkins, F. Lacrampe, J. Ripper, L. Alcaraz, P. Van Le, G. Nikolako-
poulos, P. A. de Leone, R. H. White, R. J. Quinn, J. Org. Chem.
2009, 74, 1304–1313; ad) S. M. Amorde, I. T. Jewett, S. F. Martin,
Tetrahedron 2009, 65, 3222–3231; ae) A. C. Flick, M. J. A. Caballero,
H. I. Lee, A. Padwa, J. Org. Chem. 2010, 75, 1992–1996; af) J. J.
Sahn, D. L. Comins, J. Org. Chem. 2010, 75, 6728–6731; ag) J. C.
Killen, J. Leonard, V. K. Aggarwal, Synlett 2010, 4, 579–582; ah) B.
Stevenson, W. Lewis, J. Dowden, Synlett 2010, 672–674; ai) F. D.
Ferrari, A. J. Ledgard, R. Marquez, Tetrahedron 2011, 67, 4988–
4994; aj) Y. Kurosaki, K. Shirokane, T. Oishi, T. Sato, N. Chida, Org.
Lett. 2012, 14, 2098–2101; ak) A. C. Flick, A. Padwa, J. Sulfur
Chem. 2013, 34, 88–103.
[14] a) Y. Matsumura, S. Aoyagi, C. Kibayashi, Org. Lett. 2004, 6, 965–
968; b) R. B. Andrade, S. F. Martin, Org. Lett. 2005, 7, 5733–5735;
c) Y. Yamamoto, Y. Yasuda, H. Nasu, K. Tomioka, Org. Lett. 2009,
11, 2007–2009; d) C. Gignoux, A. F. Newton, A. Barthelme, W.
Lewis, M. L. Alcarazb, R. A. Stockman, Org. Biomol. Chem. 2012,
10, 67–69; e) R. Liu, O. Gutierrez, D. J. Tantillo, J. Aubꢆ, J. Am.
Chem. Soc. 2012, 134, 6528–6531.
[15] a) M. Kuramoto, H. Arimoto, D. Uemura, Mar. Drugs 2004, 2, 39–
54; b) S. Xu, H. Yoshimura, N. Maru, O. Ohno, H. Arimoto, D.
Uemura, J. Nat. Prod. 2011, 74, 1323–1326.
[16] a) S. Yamago, E. Nakamura, Org. React. 2002, 61, 1–217; b) B. M.
Trost, Pure Appl. Chem. 1988, 60, 1615–1626; c) R. Baker, R. B.
Keen, J. Organomet. Chem. 1985, 285, 419–427.
[17] a) M. Schelwies, P. Dꢇbon, G. Helmchen, Angew. Chem. 2006, 118,
2526–2529; Angew. Chem. Int. Ed. 2006, 45, 2466–2469; b) M.
Schelwies, P. Dꢇbon, G. Helmchen, Chem. Eur. J. 2008, 14, 6722–
6733.
[18] B. T. Becicka, F. L. Koerwitz, G. J. Drtina, N. C. Baenziger, D. F.
Wiemer, J. Org. Chem. 1990, 55, 5613–5619.
[19] a) J. Tsuji, Pure Appl. Chem. 1986, 58, 869–878; b) J. Tsuji, M.
Nisar, I. Shimizu, I. Minami, Synthesis 1984, 1009.
[20] W. G. Dauben, H. G. Wight, G. A. Boswell, J. Org. Chem. 1958, 23,
1787–1788.
[21] (R)-Pulegone was purchased from Aldrich (>85% purity). Its exact
purity may depend on the lots received. However, the impurities are
not (S)-pulegone, but other chiral or achiral substrates; see also
ref. [32].
[22] a) D. Yang, C. Zhang, J. Org. Chem. 2001, 66, 4814–4818; b) L. M.
Jackman, R. L. Webb, H. C. Yick, J. Org. Chem. 1982, 47, 1824–
1831.
[23] See the Supporting Information for details.
[24] N. Gꢈlvez, M. Moreno-MaÇas, R. M. Sebastiꢈn, A. Vallribera, Tetra-
hedron 1996, 52, 1609–1616.
[25] a) R. T. Conley, M. C. Annis, J. Org. Chem. 1962, 27, 1961–1964;
b) Y. Tamura, J. Minamikawa, M. Ikeda, Synthesis 1977, 1–17; c) Y.
Tamura, H. Fujiwara, K. Sumoto, M. Ikeda, Y. Kita, Synthesis 1973,
215–216.
[26] The plausible self-decomposition mechanism was presumed to be as
follows:
[9] a) D. Trauner, J. B. Schwarz, S. J. Danishefsky, Angew. Chem. 1999,
111, 3756–3758; Angew. Chem. Int. Ed. 1999, 38, 3542–3545;
b) M. W. Carson, G. Kim, M. F. Hentemann, D. Trauner, S. J. Dani-
shefsky, Angew. Chem. 2001, 113, 4582–4584; Angew. Chem. Int.
Ed. 2001, 40, 4450–4452; c) M. W. Carson, G. Kim, S. J. Danishefsky,
Angew. Chem. 2001, 113, 4585–4588; Angew. Chem. Int. Ed. 2001,
40, 4453–4456; d) D. Trauner, S. J. Danishefsky, Tetrahedron Lett.
1999, 40, 6513–6516; e) D. Trauner, D. G. Churchill, S. J. Danishef-
sky, Helv. Chim. Acta 2000, 83, 2344–2351.
[10] a) S. Xu, H. Arimoto, D. Uemura, Angew. Chem. 2007, 119, 5848–
5851; Angew. Chem. Int. Ed. 2007, 46, 5746–5749; b) I. Hayakawa,
H. Arimoto, D. Uemura, Heterocycles 2003, 59, 441–444; c) H. Ari-
moto, S. Asano, D. Uemura, Tetrahedron Lett. 1999, 40, 3583–3586;
d) I. Hayakawa, H. Arimoto, D. Uemura, Chem. Commun. 2004,
1222–1223.
[27] L. Caglioti, Organic Syntheses, Coll. Vol. 6, 62–67.
[28] G. R. Krow, D. A. Shaw, B. Lynch, W. Lester, S. W. Szczepanski, R.
Raghavachari, A. E. Derome, J. Org. Chem. 1988, 53, 2258–2262.
[29] a) S. Yamamura, M. Toda, Y. Hirata, Organic Syntheses, Coll. Vol. 6,
289–292; b) M. Toda, M. Hayashi, Y. Hirata, S. Yamamura, Bull.
Chem. Soc. Jpn. 1972, 45, 264–266.
[30] S. Xu, T. Toyama, J. Nakamura, H. Arimoto, Tetrahedron Lett. 2010,
51, 4534–4537.
[31] T. Fukuyama, J. J. Nunes, J. Am. Chem. Soc. 1988, 110, 5196–5198.
[32] The high optical purity of our synthetic route was confirmed by
measuring the ee of the derivative of compound 25,[10a] which proved
that the original (R)-pulegone was optically pure and no racemiza-
tion occurred in our synthetic route.
Chem. Asian J. 2014, 9, 367 – 375
374
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Hirokazu Arimoto et al.
[33] With excess HOAc, the one-pot hydrogenation of the substrate in
our previous study[10b] sometimes led to byproduct A through the
following plausible mechanism:
[40] K. Matsumura, S. Hashiguchi, T. Ikariya, R. Noyori, J. Am. Chem.
Soc. 1997, 119, 8738–8739.
[41] S. B. Garber, J. S. Kingsbury, B. L. Gray, A. H. Hoveyda, J. Am.
Chem. Soc. 2000, 122, 8168–8179.
[42] M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953–
956.
[43] A. K. Chatterjee, T. L. Choi, D. P. Sanders, R. H. Grubbs, J. Am.
Chem. Soc. 2003, 125, 11360–11370.
[44] a) H. E. Blackwell, D. J. OꢀLeary, A. K. Chatterjee, R. A. Washen-
felder, D. A. Bussmann, R. H. Grubbs, J. Am. Chem. Soc. 2000, 122,
58–71; b) S. J. Meek, R. V. OꢀBrien, J. Llaveria, R. R. Schrock,
A. H. Hoveyda, Nature 2011, 471, 461–466; c) B. K. Keitz, K. Endo,
M. B. Herbert, R. H. Grubbs, J. Am. Chem. Soc. 2011, 133, 9686–
9688.
[45] a) I. Shiina, R. Ibuka, M. Kubota, Chem. Lett. 2002, 286–287; b) I.
Shiina, M. Kubota, R. Ibuka, Tetrahedron Lett. 2002, 43, 7535–7539.
[46] G. C. Fu, S. T. Nguyen, R. H. Grubbs, J. Am. Chem. Soc. 1993, 115,
9856–9857.
[34] B. P. Czech, R. A. Bartsch, J. Org. Chem. 1984, 49, 4076–4078.
[35] C. Prakash, S. Saleh, I. A. Blair, Tetrahedron Lett. 1989, 30, 19–22.
[36] P. Grieco, Y. Masaki, D. Boxler, J. Am. Chem. Soc. 1975, 97, 1597–
1599.
[37] E. J. Corey, C. J. Helal, Angew. Chem. 1998, 110, 2092–2118;
Angew. Chem. Int. Ed. 1998, 37, 1986–2012.
[38] J. Mulzer, M. Berger, J. Org. Chem. 2004, 69, 891–898.
[39] J. D. More, N. S. Finney, Org. Lett. 2002, 4, 3001–3003.
Received: September 12, 2013
Published online: October 17, 2013
Chem. Asian J. 2014, 9, 367 – 375
375
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim