Competent Route to Unsymmetric Dimer Architectures: Total Syntheses of (?)-Lycodine and (?)-Complanadines A and B, and Evaluation of Their Neurite Outgrowth Activities

Article abstract of DOI:10.1002/chem.201604647

Valuable synthetic routes to the Lycopodium alkaloid lycodine (1) and its unsymmetric dimers, complanadines A (4) and B (5), have been developed. Regioselective construction of the bicyclo[3.3.1]nonane core structure of lycodine was achieved by a remote functionality-controlled Diels–Alder reaction and subsequent intramolecular Mizoroki–Heck reaction. A key coupling reaction of the lycodine units, pyridine N-oxide (66) and aryl bromide (65), through C?H arylation at the C1 position of 66 provided the unsymmetric dimer structure at a late stage of the synthesis. This strategy greatly simplified the construction of the dimeric architecture and functionalization. Complanadines A (4) and B (5) were synthesized by adjusting the oxidation level of the bipyridine mono-N-oxide (67). The diverse utility of this common intermediate (67) suggests a possible biosynthetic pathway of complanadines in Nature. Both enantiomers of lycodine (1) and complanadines A (4) and B (5) were prepared in sufficient quantities for biological evaluation. The effect on neuron differentiation of PC-12 cells upon treatment with culture medium, in which human astrocytoma cells had been cultured in the presence of 1, 4, or 5 was evaluated.

Full text of DOI:10.1002/chem.201604647

A Journal of
Accepted Article
Title: Competent route to the unsymmetric dimer architectures: Total
syntheses of (−)-lycodine, (−)-complanadines A and B, and
evaluation of their neurite outgrowth activities
Authors: Le Zhao, Chihiro Tsukano, Eunsang Kwon, Hisashi
Shirakawa, Shuji Kaneko, Yoshiji Takemoto, and Masahira
Hirama
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201604647
Link to VoR: http://dx.doi.org/10.1002/chem.201604647
Supported by

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
Competent route to the unsymmetric dimer architectures:
Total syntheses of (−)-lycodine, (−)-complanadines A and B, and
evaluation of their neurite outgrowth activities
Le Zhao,†[b] Chihiro Tsukano,*,‡[a] Eunsang Kwon,†[b] Hisashi Shirakawa,‡[c] Shuji Kaneko,‡[c] Yoshiji
Takemoto,‡[a] and Masahiro Hirama†[b]
Abstract: Valuable synthetic routes to the Lycopodium alkaloid
lycodine (1), and its unsymmetric dimers, complanadines A (4) and B
(5) were developed. Regioselective construction of the
bicyclo[3.3.1]nonane core structure of lycodine was achieved by a
remote functionality-controlled Diels-Alder reaction and the
subsequent intramolecular Mizoroki-Heck reaction. A key coupling
reaction of the lycodine units, pyridine N-oxide (66) and aryl bromide
(65), through C-H arylation at the C1 position of 66 provided the
unsymmetric dimer structure in the late stage of the synthesis. This
strategy greatly simplified the construction of the dimeric architecture
which was determined by X-ray crystallography.[4a,b]
Complanadine A (4), isolated from Lycopodium complanatum by
Kobayashi and co-workers in 2000 is the first dimeric alkaloid of
the lycodine family.[6a] The compound (4) featuring an
unsymmetric C1-C2‘ linkage[7] between the two lycodine units
was reported to induce secretion of neurotropic factors from
human astrocytoma (1321N1) cells, which promoted neuronal
differentiation of PC-12 cells.[6b] These phenomena are
promising for the treatment of Alzheimer‘s disease. Furthermore,
4 showed cytotoxicity against murine leukemia L1210 cells in
vitro.[6a] While its congeners, complanadines B (5), D (6) and E
(7) were also isolated,[6b,c,d] the scarcity of 5–7 has hampered
extensive biological studies on their neurite outgrowth
activity.[6b,c,d]
and functionalization. Complanadines
A (4) and B (5) were
synthesized by adjusting the oxidation level of the bipyridine mono-
N-oxide (67). The diverse utility of this common intermediate (67)
suggests a possible biosynthetic pathway of complanadines in
nature. Both enantiomers of lycodine (1) and complanadines A (4)
and B (5) were prepared in sufficient quantities for biological
evaluation. The effect on neuron differentiation of PC-12 cells upon
treatment with culture medium in which human astrocytoma cells
has been cultured in the presence of 1, 4, or 5, was evaluated.
Introduction
The Lycopodium alkaloids with a complex C16 tri- or tetracyclic
skeleton were originally identified in the family of club moss
Lycopodium[1] and more than 250 congeners have been reported.
They often show intriguing biological activities.[2] Lycodine (1),
one of the oldest member of the Lycopodium alkaloids, was
isolated from L. annotinum by Anet and Eves (Figure 1).[3a] The
structure of 1 was determined using spectroscopic methods (1H
NMR and IR) and derivatization from the related Lycopodium
alkaloid
-obscurine
(2).[3b]
Lycodine
contains
a
bicyclo[3.3.1]nonane skeleton attached by pyridine and
piperidine rings. The absolute stereochemistry was assumed to
be identical with that of the structurally related lycopodine (3),
Figure 1. Lycodine, complanadines and structurally related Lycopodium
alkaloids.
[a]
Dr. C. Tsukano*, Prof. Dr. Y. Takemoto
Department of Organic Chemistry
Graduate School of Pharmaceutical Sciences, Kyoto University
Yoshida, Sakyo-ku, Kyoto 606-8501, Japan
E-mail: tsukano@pharm.kyoto-u.ac.jp
Dr. L. Zhao, Dr. E. Kwon, Prof. Dr. M. Hirama
Department of Chemistry
Graduate School of Science, Tohoku University
Sendai 980-8578, Japan
Dr. H. Shirakawa, Prof. Dr. S. Kaneko
The Heathcock[5a] and Takayama groups[5b] reported the
synthesis of racemic 1 and an enantioselective synthesis of 1,
respectively. Total synthesis of 4 was recently reported by the
Sarpong[8a] and Siegel[9] groups. Sarpong and co-workers also
reported a total synthesis of 5,[8b] whereas the Lewis group
reported a synthetic study toward complanadines.[10] We were
also interested in the syntheses of 1 and its unsymmetric dimers
4 and 5, because dimeric alkaloids often show distinct and/or
improved biological activities when compared to those of the
[b]
[c]
Department of Molecular Pharmacology
Graduate School of Pharmaceutical Sciences, Kyoto University
Yoshida, Sakyo-ku, Kyoto 606-8501, Japan
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
monomer.[11] Recently, we accomplished a total synthesis of Diels-Alder reaction for construction of
1,[12a] and moreover, total syntheses of 4 and 5 based on a phlegmarine skeleton
strategy inspired by the biosynthetic hypothesis.[12b] In this article,
Since the Diels-Alder reaction of diene 21[18] with
we describe the full details of our synthetic studies of these
molecules and their enantiomers, which enabled evaluation of
their neurite outgrowth activities for the first time.
,unsaturated ketone 22 only led to undesired regioisomer 23
as anticipated (Scheme 2), an electron withdrawing ester group
was added at C8 as dienophile 26a to reverse the
regioselectivity.[19] Additionally, to facilitate NMR analysis of the
Diels-Alder adducts, diene 25 was employed for the reaction
instead of 21.
Key features of the biosynthetic hypothesis of 1 are i)
dimerization of a pelletierine unit and ii) bond formation between
C4 and C13 of the phlegmarine intermediate 11 (Scheme 1a).[13]
Furthermore, complanadines (4 and 5) are derived from two
units of 1.[6b,14] Therefore, we envisaged that the Diels-Alder
reaction of diene 13 and dienophile 14 would provide
phlegmarine skeleton 15 (Scheme 1b). The intramolecular
Mizoroki-Heck
reaction[15]
of
15
should
give
a
bicyclo[3.3.1]nonane 16.[16] Installation of the C15 methyl group
would lead to 17 and then to 1. Coupling of m-bromopyridine 18
and pyridine N-oxide 19 via C1 C–H activation[17] should afford
bipyridyl mono-N-oxide 20 designed as a common intermediate
for the syntheses of complanadines.
Scheme 2. Diels-Alder reaction of 21 and 22.
The Diels-Alder reaction of 25[18] and 26a[12a] proceeded
smoothly in toluene at 110 °C, while thermodynamically stable
tetrasubstituted olefins 27aE and 27aF were obtained through
double bond isomerization (Table, entry 1). For avoiding this
undesired double bond isomerization, several dienophiles were
screened. While introduction of a TfO group at the C4 position
was not effective, the presence of methoxy (at C1) or hydroxyl
(at C4) groups suppressed this isomerization (entries 2–5). For
the reaction of 26c, no regioselectivity was observed, while a
dienophile with a bromine atom at C4 further favored the
undesired regioisomer (entries 3, 4). These results indicate that
the electron density of the pyridine ring did not perturb the olefin
of the dienophile. On the other hand, introduction of a hydroxyl
group improved the selectivity (entry 5). To discern the origins of
this selectivity, we conducted a DFT calculation for the most
stable conformation of 26e.[20] It was found that conformer Y,
which was stabilized by a hydrogen bond, was the most stable
(Scheme 3). Considering the resonance of conformer Y (i.e., Z),
introduction of the hydroxyl group at C4 reduces the electron
withdrawing property of the C6 carbonyl carbon and improves
selectivity. The obtained four diastereomers 27eA–D (R1 = OH,
R2
= H) were carefully separated by silica gel column
chromatography and their relative stereochemistries were
determined by extensive NMR analysis, including NOESY
experiments.
Scheme 1. Proposed biosynthesis of lycodine, and synthetic plan of lycodine
and complanadine A and B.
Scheme 3. Resonance conformers of 26e.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
Table 1. Diels-Alder reaction of 25 and 26a-e and optimized transition states for 25 and 26e at the B3LYP/6-31G(d) level.
Entry
Dienophile
R1
R2
Conditions
Yield (%)
27xB 27xC
27aE: 40 27aF: 31
27bE: 34 27bF: 33
Regioselectivity
(27xA+27xB):(27xC+27xD)
27xA
27xD
Toluene, 110 °C
1.3:1
1.0:1
1.1:1
0.7:1
2.2:1
2.2:1
2.0:1
2.1:1
2.6:1
0.5:1
2.6:1
2.2:1
2.4:1
1
2
26a
26b
26c
26d
26e
26e
26e
26e
26e
26e
26e
26e
26e
H
H
H
Toluene, 110 °C
OTf
H
Toluene, 110 °C
22
14
45
32
44
51
52
7
12
19
15
15
11
11
7
13
19
6
19
29
21
13
17
19
14
12
3
3
OMe
OMe
H
Toluene, 110 °C
4
Br
Toluene, 110 °C
5
OH
OH
OH
OH
OH
OH
OH
OH
OH
Pyridine, 80 °C
8
6
H
Ethanol, 80 °C
10
10
9
7
H
CH3CN, 80 °C
8
H
CH3CN, RT
9
H
BF3･OEt2[b], CH2Cl2, -78 °C
K2CO3[b], CH3CN, RT
Et3N[c], CH3CN, 0 °C
NaH2PO4･2H2O[b], CH3CN, RT
1
4
10
11
12
13
H
12
22
56
1
2
H
4
5
7
H
6
9
17
H
[a] Isolated yield. [b] 1 equiv. [c] 0.05 equiv.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
We then investigated solvents, temperatures and additives
using diene 25 and dienophile 26e. Using pyridine, ethanol and
acetonitrile at 80 °C resulted in similar selectivity (entries 6–8).
These results were consistent with DFT calculations, which
indicated that conformer Y is the most stable species in these
solvents. This Diels-Alder reaction in acetonitrile proceeded at
room temperature with better selectivity (entry 9). Addition of a
Lewis acid, such as BF3 ･ OEt2, resulted in low yield and
selectivity, probably because the carbopyridyl moiety is a better
Lewis acid acceptor than the ester (entry 10). Furthermore,
decomposition of 26e was also observed under these
conditions. Various bases and inorganic salts were also
examined.[21] Use of K2CO3 or Et3N did not improve the yield or
selectivity due to the low stability of dienophile 26e under basic
conditions (entries 11, 12). Addition of a weak Brønsted acid,
NaH2PO4 ･ 2H2O, was found to suppress effectively the
decomposition of 26e and resulted in an improved yield of the
desired product 27eA (entry 13). The optimized conditions
(NaH2PO4･2H2O, CH3CN, rt) were applied successfully to a
large scale synthesis of the desired adduct 27eA.
under these conditions. It is rationalized that the 6-exo-trig
cyclization via B would be preferable than the 7-endo-trig
cyclization via C due to electronic factors in the hetero-
substituted olefin and difficulty of -hydride elimination from
C.[22] To suppress the production of by-product 31, a few
catalysts and solvents were screened. The use of PdCl2(PPh3)2
in dimethylacetamide (DMA) were found to be effective.
Additionally, use of PdCl2(PPh3)2 reduced an amount of
phosphine oxide as by-product, thus facilitating purification.
However, these optimized conditions were not reproducible
over the concentration of 0.1 M. For example, the 0.1 M
reaction in DMA resulted in quick precipitation of Pd black and
recovery of a significant amount of starting material (Table 2,
entry 1). These results indicated that the palladium catalyst was
deactivated before oxidative addition of triflate 30 because two
pyridyl ketone moieties can coordinate to the palladium and
accelerate precipitation of the palladium black. Thus, the
reaction was carried out under high dilution conditions (entries
2–4). Consequently, the reaction at 0.005 M of DMA provided
33 in 73% yield with good reproducibility.
To rationalize this selectivity, the activation energy of this
reaction was estimated by DFT calculations at the B3LYP/6-
31G(d) level. The activation energies for 27eA–D were ∆E‡ =
14.4, 15.8, 16.7, 15.5 kcal/mol, respectively (Table 1). These
results are consist with the ratio of the products (i.e., entry 5)
and indicated that the transition state for the desired product
27eA was the most favorable among the isomers. Additionally,
it was found that the distance between C7 and C12, C8 and
C15 were 2.686 and 1.955 Å, respectively. These results
suggested that this Diels-Alder reaction proceeds through a
stepwise process rather than a concerted mechanism. It is
noted that endo/exo selectivity was explained by secondary
interactions of the -orbital of the pyridyl ketone with the diene,
which was larger than that of an ester’s LUMO, according to
DFT calculations of the transition state for 27eA (Table 1).
Therefore, the transition state, in which the pyridyl ketone and
ester are in the endo and exo positions, respectively, was
favored to produce the desired adduct 27eA as the major
product.
Construction of the Tetracyclic Core and
Total Synthesis of Lycodine
After triflation of 27eA, we investigated the Mizoroki-Heck
reaction of compound 30 for construction of the tetracyclic core
(Scheme 4). In our initial attempts, the reaction using a catalytic
amount of tetrakis(triphenylphosphine)palladium (Pd(PPh3)4)
and triethylamine in dimethylformamide (DMF) did not proceed
at 100 °C. We assumed that the oxidative addition of compound
30 would require higher temperature, thus 30 was treated with
Pd(PPh3)4 in DMF at 120 °C. Predictably, the desired 6-exo-trig
cyclization proceeded to produce tetracyclic compound 32,
which was immediately isomerized to the conjugated
unsaturated ester 33 in 32% yield with a significant amount of
reduced compound 31 (30%). The regioselectivity was
exclusive, and the 7-endo-trig product 34 was never observed
Scheme 4. Construction of tetracyclic core through intramolecular Mizoroki-
Heck reaction of 30.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
which was treated with diphenylphosphoryl azide (DPPA) and
then water for 30 min. While 38 was obtained in a 30% yield, a
prolonged reaction time (6 h) resulted in retro-Claisen[29]
fragmentation of the 1,3-diketone moiety of 38 to give thermally
stable 39 in 87% yield. To prevent this undesired side-reaction,
the deoxygenated compound 36 was used and converted to
ketone 40 in 98% yield over two steps using the same
sequence.
Table 2. Concentration effect of intramolecular Mizoroki-Heck reaction of 30.
entry
Reaction time
(h)
Concentration of 30
Yield of 33[a]
(M)
(%)
1
2
3
4
48
22
24
24
0.1
0.05
0.01
0.005
18
42
61
73
Introduction of the C15 methyl group was then investigated
using enolate chemistry. Treatment of 40 with LiHMDS, then
followed by iodomethane gave a trace amount of the desired
product with recovery of 40 (Scheme 5, Table 3, entry 1).
Warming the reaction temperature did not improve the yield of
methylated product 41. This was presumably due to the
formation of the undesired aggregate of the lithium enolate,
which was not reactive. Thus, we examined a lithium-free
enolate derived from a silyl enol ether, prepared from ketone 40
under standard conditions (LiHMDS, Et3N, TMSCl, THF, –
78 °C). Following Noyori’s report,[30a] the silyl enol ether was
treated with iodomethane, TASF and 4 Å molecular sieves, the
reaction proceeded to give the desired methyl ketone 41 in 37%
yield along with 40 (47%) and a small amount of dimethylated
product (entry 2). After several screenings of fluoride sources,
we found that benzyltrimethylammonium fluoride (BTAF) was
the best fluoride source for this transformation (entries 3, 4).[30b]
In our technique, removing water from BTAF would be easier
than from the other fluoride sources. The stereochemistry of 41
was confirmed by extensive NMR spectroscopic analysis,
including NOESY experiments (Figure 2). In the course of this
reaction, no epimers at C15 were observed, because
iodomethane approaches from the desired face to avoid the
steric hindrance of the pyridine ring.
[a] Isolated yield.
With the tetracyclic core 33 in hand, we focused on (i)
introduction of the C15 methyl group, (ii) reduction of the ketone
to a methylene, and (iii) removal of the ethyl ester moiety for the
total synthesis of racemic lycodine. We initially attempted 1,4-
addition of cuprate reagents into the unsaturated ester of 33 to
introduce the C15 methyl group. However, the reaction with
various organocopper reagents (MeCu+TMSI, MeCu+TMSCl,
Me2CuLi+TMSCl, Me(Th)CuLi+TMSCl)[23] did not furnish the
1,4-addition product 35, and the starting material 33 was
recovered (Scheme 5).[24] The pyridyl ketone moiety might be
problematic due to coordination with the cuprate reagent.
Therefore, the ketone of 33 was reduced by a three-step
transformation, including Luche reduction[25] and Barton-
McCombie deoxygenation.[26] 36 was treated with methyl
cuprate reagents again to access compound 37; however,
these attempts only resulted in recovery of the starting material.
We concluded that this is presumably due to the low reactivity
of the ,-unsaturated ester rather than coordination of pyridine
to the cuprate reagent.[27] Thus, we switched tasks to the
removal of the ethyl ester moiety by hydrolysis and
decarboxylation by Curtius rearrangement.[28] Compound 33
was hydrolyzed under basic conditions to give a carboxylic acid,
Scheme 5. Total synthesis of racemic lycodine (1).
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
(16%), and the other diastereomers were obtained in less than
10% yields. Although the diastereoselectivity was moderate,
enough material was obtained to continue the synthesis. Thus,
(+)-43a was converted to tetracyclic (−)-45 using the same
sequence developed in the racemic synthesis. When (−)-45
was treated under basic conditions such as LiOH in THF-
MeOH-H2O, and NaOEt in EtOH, it resisted solvolysis
presumably due to steric hindrance. Thus, the chiral auxiliary of
(−)-45 was removed by reduction of the unsaturated ester using
DIBAL-H. The absolute stereochemistry was determined at this
stage. After oxidation of the obtained allyl alcohol (−)-46,
Curtius rearrangement and diastereoselective reduction, the
Kusumi-Mosher’s method[33] was applied to the resultant
secondary alcohol (not shown, see Supporting Information).
The results indicated that the absolute stereochemistry derived
from the Diels-Alder reaction was the unnatural type.
Table 3. Introduction of the C15 methyl group.
Yield[a] (%)
entry
Conditions
41
5
40
1
2
LiHMDS, MeI, THF, –78 C
52
47
1. LiHMDS, Et3N, TMSCl, THF, –78 C
37
2. MeI, TASF, MS 4 Å, THF, RT
3
4
1. LiHMDS, Et3N, TMSCl, THF, –78 C
2. MeI, TBAF, MS 4 Å, THF, RT
47
64
18
14
1. LiHMDS, Et3N, TMSCl, THF, –78 C
2. MeI, BTAF, MS 4 Å, THF, RT
[a] Isolated yield. TASF = Tris(dimethylamino)sulfonium difluorotrimethylsili-
cate, BTAF = benzyltrimethylammonium fluoride
Figure 2. Stereochemistry of the C15 methyl group of 41.
Finally, thioacetalization of methyl ketone 41 and subsequent
one-pot reduction and deprotection of the Cbz group using
Raney Ni (W-2) gave racemic lycodine (1). Spectroscopic (1H
and 13C NMR, UV, IR)[31] and high-resolution mass
spectrometric data of the synthetic sample were identical to
those of the natural product. The established synthetic route,
which is 15 steps from methyl 3-hydroxypicolinate, is robust
and convenient for accessing the racemic lycodine and a
protected monomer for synthesis of the complanadines.
Synthesis of Optically Pure Lycodine and
Determination of Absolute Stereochemistry
Next, we turned our attention to the synthesis of the
enantiomerically pure form of 41, which is essential for the total
synthesis of complanadines via dimerization. We employed a
chiral auxiliary, which was installed to the ester moiety, because
the hydroxyl group on pyridine is important for regioselectivity.
The Diels-Alder reaction of diene 25 and dienophile (+)-42
bearing the chiral auxiliary prepared from the natural (+)-
pulegone,[32] proceeded smoothly in acetonitrile to provide (+)-
43a as the major product in moderate yield (40%), which was
separated from the other diastereomers by silica gel column
chromatography (Scheme 6). The major adduct had the desired
relative and unnatural absolute stereochemistry, which was
determined later. The second major diastereomer was (+)-43b
Scheme 6. Diels-Alder reaction of 25 and (+)-42 with the chiral auxiliary and
alternative routes to access optically active (−)-41.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
With the enantiopure tetracyclic core (−)-46 in hand, we
synthesized the monomer 41 (unnatural). To avoid a lengthy
route through re-oxidation of allyl alcohol (−)-46 and
decarboxylation, the C15 methyl group was introduced
viaselective allylic alkylation of the allyl phosphate. Allyl
alcohol (−)-46 was converted to the corresponding allyl
phosphate (−)-47 in 78% yield. Treatment of (−)-47 with low
order methyl cuprate gave compound 48 as a 2:1 epimer at
C15, which was inconsequential, while the regioselectivity is
about : = 3:1.[34] After ozonolysis of the mixture, epimerization
using DBU proceeded smoothly to give the thermally stable
methyl ketone (−)-41 in 71% yield over three steps, because of
the disfavored 1,3-diaxal interaction in C15-epi 41.
As an alternative method, racemic 41 could be separated into
its optically active forms using chiral high-performance liquid
chromatography. Use of an amylose chiral column (Daicel
Chiral-Pak AS) with a mixture of ethanol/hexane/diethylamine
(5.8%/94%/0.2%) as the eluent provided (+)- and (−)-41 in 48
and 47% yields, respectively. This also provided sufficient
quantities of material for further studies.
Total Synthesis of Complanadines A and B
After completion of the total synthesis of lycodine (1), we
investigated
the
challenging
dimerization
process.
Biosynthetically, there are two possibilities: the intermediate is
(i) the natural product itself, and (ii) a compound that is not
isolated. Considering the unsymmetric structure of the bipyridyl
linkage and the oxidation level state of the pyridine ring, we
assumed that complanadines are biosynthetically produced
through the latter possibility. Inspired by this idea, we chose
bipyridine mono-N-oxide 20 as the more reasonable compound,
which could readily be accessed from bromopyridine 18 and its
pyridine N-oxide 19 using C-H activation chemistry (Scheme 1).
Due to the unprecedented complexity of both the coupling
partners 18 and 19, tetrahydroquinoline 51 was employed as a
model compound to establish feasible conditions for the
functionalization of the pyridine ring and direct arylation via C-H
activation of pyridine N-oxide. Pyridine N-oxide 52 was readily
prepared by mCPBA oxidation of 51 (Scheme 8). Treatment of
51 with dibromoisocyanuric acid in oleum gave bromopyridine
53 in 39% yield as a single product.[36] Because the Cbz
protecting group in monomer 41 would not be tolerated under
these harsh conditions, we also employed Miyaura-Hartwig
borylation[37] of 51 by using the procedure of Sarpong and co-
workers.[8a] Subsequent bromination of 54 with CuBr2 to give 53
in 84% improved the yield over two steps.[38]
Thioacetalization, reduction and deprotection of the Cbz
group as before resulted in completion of the total synthesis of
both natural and unnatural lycodine ((−)-1, (+)-1) (Scheme 7).
Spectroscopic (1H NMR, 13C NMR, UV, IR, optical rotation) and
high-resolution mass spectrometric data of the synthetic sample
were identical to those of the natural product.
Although the absolute stereochemistry had already been
determined using the modified Kusumi-Mosher’s method, the
stereochemistry of synthetic natural type lycodine was further
confirmed by X-ray crystallographic analysis. The synthetic (−)-
lycodine 1 was treated with p-bromobenzoyl chloride and
recrystallized from THF-hexane to give its p-bromobenzoate
derived as a single crystal (+)-50. X-ray crystallographic
analysis indicated that the synthetic sample was the natural
type enantiomer. [4, 35]
With both coupling partners in hand, we examined the key C-
H arylation reaction. Based on Fagnou’s report,[17] 52 and 53
were treated with a catalytic amount of Pd(OAc)2, tri-tert-
butylphosphine tetrafluoroborate (tBu3P·HBF4) as ligand and
potassium carbonate in mesitylene at 120 C to give bipyridine
mono-N-oxide 55 in 33% yield. After screening electron rich
phosphine
ligands,
such
as
tricyclohexylphosphine
tetrafluoroborate (Cy3P·HBF4) and 2-di-tert-butyl-phosphino-2'-
(N,N-dimethylamino)biphenyl (tBuDavePhos), the latter gave a
better yield. Addition of pivalic acid (PivOH) was effective.[39]
Finally, employment of cesium carbonate improved the product
yield significantly.
The coupling product 55 was reduced with palladium
hydroxide on carbon and ammonium formate as
a
homogeneous hydrogen source to give the complanadine A
model (56). By heating 55 in acetic anhydride, a sigmatropic
rearrangement occurred to introduce an oxygen functionality at
the C6 picolinic position.[40] The obtained 57 was further
converted to the complanadine B model 58 by hydrolysis and
oxidation.[41]
Scheme 7. Synthesis of (−)-lycodine (1) and confirmation of its absolute
stereochemistry by X-ray crystallography. Thermal ellipsoids are shown at the
50% probability level.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
unclear; although, we suspect the presence of the C15
enolizable proton.
Scheme 9. Unsuccessful attempt of the bipyridyl mono-N-oxide framework
construction using methyl ketone 41.
Thus, the ketone was reduced to methylene through radical
reduction[43] of thioacetal (+)-49 and then the established
sequence was applied to Cbz-protected lycodine (+)-64
(Scheme 10). Miyaura-Hartwig borylation and bromination
afforded bromopyridine (+)-65 in 60% yield, with a small
amount of by-product derived from bromination of the Cbz
group. Pyridine N-oxide (+)-66 was obtained in 92% yield by
mCPBA oxidation of (+)-64. Treatment of (+)-65 and (+)-66 with
10 mol % of Pd(OAc)2, 20 mol % of tBuDavePhos, cesium
carbonate (3 equiv.) and PivOH (0.3 equiv.) in mesitylene at
130 C furnished bipyridine mono-N-oxide (+)-67 in 62% yield.
This was one of the most difficult examples of the C-H coupling
of a pyridine N-oxide in terms of molecular complexity. These
results indicate that the C15 enolizable proton of 59 or 60 would
have hampered the C(sp2)-H activation process, and the
benzylic proton at C6 did not affect the coupling of (+)-65 and
(+)-66.[17c] Following our model studies, coupling product (+)-67
was reduced with palladium hydroxide on carbon and
Scheme 8. Bipyridine mono-N-oxide framework construction using model
compounds.
Table 4. Optimization of C-H arylation of pyridine N-oxide 52.
Entry
Ligand
Additive
None
Base
K2CO3
K2CO3
K2CO3
K2CO3
Cs2CO3
Yield[a] (%)
1
2
3
4
5
tBu3P·HBF4
Cy3P·HBF4
tBuDavePhos
tBuDavePhos
tBuDavePhos
33
37
41
46
56
None
None
PivOH
PivOH
[a] Isolated yield.
We initially applied this sequence to methyl ketone 41.
mCPBA oxidation gave pyridine N-oxide 59 in good yield, but
the Miyaura-Hartwig borylation was problematic. Two-step
bromination afforded m-bromopyridine 60 in only 33% yield
along with alcohols 61, 62 and a small amount of by-product
derived from bromination of the Cbz group (Scheme 9).[42]
Coupling between 59 and 60 gave no desired product 63 under
the conditions using tBu3P· HBF4; although, it was not the
optimized ligand. The de-brominated compound 41 and
unaffected pyridine N-oxide 59 were obtained. We further
ammonium formate to give (−)-complanadine
A
(4).
Rearrangement, basic hydrolysis, oxidation and deprotection
completed the total synthesis of (−)-complanadine B (5). The 1H,
13C NMR data of the synthetic samples were identical to those
of the natural products. For biological studies, unnatural
enantiomers (+)-4 and (+)-5 were also synthesized from the
common intermediate (−)-41 in the same manner as their
natural enantiomer.
It is noteworthy that our synthesis requires only one pair of
monomers for the synthesis of two unsymmetric dimers. By
designing an appropriate intermediate based on the C-H
arylation reaction, we no longer need to prepare several
monomers for each unsymmetric natural product. This fact
suggests an intermediacy of a mono-N-oxide, such as 67, in the
biosynthesis of dimeric alkaloids.
considered
the
mechanism
of
these
problematic
transformations. In the Miyaura-Hartwig borylation reaction,
bis(pinacolato)diboron is converted to pinacolborane after C-H
activation of pyridine, then the pinacolborane reduces the
ketone to an alcohol.[37] In the case of palladium-catalyzed
direct arylation, the reason why the reaction did not proceed is
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
Scheme 10. Total synthesis of complanadines A (4) and B (5).
Evaluation of Biological Activity
Figure 3. Glial cell-mediated morphological changes to PC-12 cells by (−)-
complanadine A (4). (a) Control (48 h + 48 h), (b) complanadine A (5 M, 48 h
+ 48 h), (c) ionomycin (1 M; 48 h + 48 h, positive control).
Because several compounds including antipodes of natural
products were obtained, we examined the effect of these
compounds on the neuronal differentiation of rat
pheochromocytoma (PC12) cells. Human astrocytoma
Conclusions
(1321N1) cells were incubated for 48
h with lycodine,
complanadines A and B and their enantiomers ((−)-1, (−)-4, (−)-
5, (+)-1, (+)-4, (+)-5), and then PC-12 cells were cultivated for
48 h in the above 1321N1 culture medium, respectively. In the
case of (−)-4 (5 M), neurite extension on PC-12 cells was
observed as described by the isolation group (Figure 3), while
cytotoxicity of (−)-4 hampered evaluation over a concentration
of 10 M. In sharp contrast, the other synthetic compounds
((−)-1, (−)-5, (+)-1, (+)-4, (+)-5) did not show any change at a
concentration of 10 M. These results suggest that not only the
dimeric structure but chirality and oxidation levels were
important for biological activity of (−)-complanadine A (4).
We accomplished the total syntheses of both enantiomers of
lycodine ((−)-1, (+)-1) and complanadines A and B ((−)-4, (+)-4,
(−)-5, (+)-5). Key transformations included (i) the remote
functional group controlled diastereoselective Diels-Alder
reaction, (ii) the 6-exo-trig intramolecular Mizoroki-Heck
reaction to construct the tetracyclic lycodine core, and (iii) the
direct arylation to pyridine N-oxide to yield the unsymmetric
dimer at the final stage. The bipyridyl mono-N-oxide
functionality of 67 provided successful access to the inherently
different oxidation levels of the left domain in both 4 and 5, and
enabled their biological evaluation. Preliminary biological
studies indicated that not only the dimeric structure but also
chirality and oxidation levels are important for the biological
activity of complanadine A (−)-4.
(a)
(b)
Acknowledgments
We gratefully thank Profs. J. Kobayashi, H. Morita and T.
(c)
1
Kubota for providing the natural products and H and 13C NMR
spectra. This work was financially supported by a Grant-in-Aid
for Young Scientists (Start-up) from the Japan Society for
Promotion of Science, and Scientific Research on Innovation
Area “Molecular Activation Directed toward Straightforward
Synthesis” from the Ministry of Education, Culture, Sports,
Science and Technology (Japan) (C.T.).
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
Schipper, K. Fagnou, J. Am. Chem. Soc. 2008, 130, 3266–3267; d) R.
Martin, S. L. Buchwald, Acc. Chem. Res. 2008, 41, 1461–1473.
[18] J. D. Ha, C. H. Kang, K. A. Belmore, J. K. Cha, J. Org. Chem. 1998,
63, 3810–3811.
Keywords: total synthesis • complanadine • lycodine • C-H
arylation • palladium
[1]
For reviews of Lycopodium alkaloids, see: a) Y. Hirasawa, J.
Kobayashi, H. Morita, Heterocycles 2009, 77, 679–729; b) X. Ma, D. R.
Gang, Nat. Prod. Rep. 2004, 21, 752–772; c) W. A. Ayer, Nat. Prod.
Rep. 1991, 8, 455–463.
[19] R. C. B. Bugle, D. M. S. Wheeler, M. M. Wheeler, J. R. Hohman, J.
Org. Chem. 1981, 46, 3915–3917.
[20] See Supporting Information for more details.
[21] All attempted bases and salts are shown in the Supporting Information.
[22] W. Cabri, I. Candiani, Acc. Chem. Res. 1995, 28, 2–7.
[23] a) N. Asao, S. Lee, Y. Yamamoto, Tetrahedron Lett. 2003, 44, 4265–
4266; b) M. Bergdhal, E. L. Lindstedt, M. Nilsson, T. Olsson,
Tetrahedron 1988, 44, 2055-–2062; c) E. L. Lindstedt, M. Nilsson, T.
Olsson, J. Organmet. Chem. 1987, 334, 255–261; d) B. H. Lipshutz, M.
Koerner, D. A. Parker, Tetrahedron Lett. 1987, 28, 945–948.
[24] Because the C6 carbonyl group appeared in complanadine B (3), we
attempted to synthesize a key intermediate with a ketone moiety at C6
for dimerization.
[2]
[3]
[4]
W. A. Ayer, L. S. Trifonov, Lycopodium Alkaloids, Academic Press, San
Diego, 1994.
a) F. A. L. Anet, C. R. Eves, Can. J. Chem. 1958, 36, 902–909; b) W. A.
Ayer, G. G. Iverach, Can. J. Chem. 1960, 38, 1823–1826.
a) Structurally defined natural lycopodine was chemically transformed
to lycodine in 1960. However, it was identified only by mp, mixed mp,
UV and IR spectra at that time, see: F. A. L. Anet, M. V. Rao,
Tetrahedron Lett. 1960, 1, 9–12; b) The absolute configuration of
lycopodine was determined afterwards, see: M. U. Hague, D. J.
Rogers, Chem. Soc. Perkin Trans II, 1975, 93–98.
[25] Luche reduction proceeded in
a high diastereoselective manner,
[5]
[6]
a) C. H. Heathcock, E. F. Kleinman, E. S. Binkley, J. Am. Chem. Soc.
1982, 104, 1054–1068; b) M. Azuma, T. Yoshikawa, N. Kogure, M.
Kitajima, H. Takayama, J. Am. Chem. Soc. 2014, 136, 11618–11621.
For isolation of complanadines A, B, D and E, see: a) J. Kobayashi, Y.
Hirasawa, N. Yoshida, H. Morita, Tetrahedron Lett. 2000, 41, 9069–
9073; b) H. Morita, K. Ishiuchi, A. Haganuma, T. Hoshino, Y. Obara, N.
Nakahata, J. Kobayashi, Tetrahedron 2005, 61, 1955–1960; c) K.
Ishiuchi, T. Kubota, Y. Mikami, Y. Obara, N. Nakahata, J. Kobayashi,
Bioorg. Med. Chem. 2007, 15, 413–417; d) K. Ishiuchi, T. Kubota, H.
Ishiyama, S. Hayashi, T. Shibata, K. Mori, Y. Obara, N. Nakahata, J.
Kobayashi, Bioorg. Med. Chem. 2011, 19, 749–753.
probably due to chelation. M. Kurosu, L. R. Marcin, T. J. Grinsteiner, Y.
Kishi, J. Am. Chem. Soc. 1998, 120, 6627-–6628.
[26] D. H. R. Barton, S. W. McCombie, J. Chem. Soc., Perkin Trans. 1
1975, 1574–1585.
[27] A. S. Vellekoop, R. A. J. Smith, J. Am. Chem. Soc. 1994, 116, 2902–
2913.
[28] T. Shioiri, K. Nishiyama, S. Yamada, J. Am. Chem. Soc. 1972, 94,
6203–6205.
[29] For related study, see: W. C. Han, K. Takahashi, J. M. Cook, U. Weiss,
J. V. Silverton, J. Am. Chem. Soc. 1982, 104, 318–321.
[30] a) R. Noyori, I. Nishida, J. Sakata, M. Nishizawa, J. Am. Chem. Soc.
1980, 102, 1223–1225; b) I. Kuwajima, E. Nakamura, M. Shimizu, J.
Am. Chem. Soc. 1982, 104, 1025–1030.
[7]
[8]
The numbering of carbons is according to those of complanadine A.
a) D. F. Fischer, R. Sarpong, J. Am. Chem. Soc. 2010, 132, 5926–
5927; b) J. N. Newton, D. F. Fischer, R. Sarpong, Angew. Chem. Int.
Ed. 2013, 52, 1726–1730.
[31] T. T. Nakanishi, P. P. Singer, L. M. Browne, W. A. Ayer, Can. J. Chem.
1975, 53, 1936–1942.
[9]
a) C. X. Yuan, C. T. Chang, A. Axelrod, D. Siegel, J. Am. Chem. Soc.
2010, 132, 5924–5925; b) C. X. Yuan, C. T. Chang, D. Siegel, J. Org.
Chem. 2013, 78, 5647–5668; c) T. Johnson, D. Siegel, Bioorg. Med.
Chem. Lett. 2014, 24, 3512–3515.
[32] a) E. J. Corey, H. E. Ensley, J. Am. Chem. Soc., 1975, 97, 6908–6909;
b) H. E. Ensley, C. A. Parnell, E. J. Corey, J. Org. Chem., 1978, 43,
1610–1611.
[33] I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J. Am. Chem. Soc.
1991, 113, 4092–4096.
[10] M. Uosis-Martin, G. D. Pantos, M. F. Mahon, S. E. Lewis, J. Org.
Chem. 2013, 78, 6253–6263.
[34] a) A. Yanagisawa, N. Nomura, H. Yamamoto, Tetrahedron 1994, 50,
6017–6028; b) N. Yoshikai, S. L. Zhang, E. Nakamura, J. Am. Chem.
Soc. 2008, 130, 12862–12863.
[11] For reviews on synthesis of the dimeric natural product, see: a) M.
Vrettou, A. A. Gray, A. R. E. Brewer, A. G. M. Barrett, Tetrahedron
2007, 63, 1487–1536; b) S. A. Snyder, Nature 2010, 465, 560–561.
[12] a) C. Tsukano, L. Zhao, Y. Takemoto, M. Hirama, Eur. J. Org. Chem.
2010, 22, 4198–4200; b) L. Zhao, C. Tsukano, E. Kwon, Y. Takemoto,
M. Hirama, Angew. Chem. Int. Ed. 2013, 52, 1722–1725.
[35] See the Supporting Information for the details of the X-ray
crystallographic analysis of (+)-34.
[36] W. Gottardi, Monatsch. Chem. 1968, 99, 815–822.
[37] T. M. Boller, J. M. Murphy, M. Hapke, T. Ishiyama, N. Miyaura, J. F.
Hartwig, J. Am. Chem. Soc. 2005, 127, 14263–14278.
[38] J. M. Murphy, X. Liao, J. F. Hartwig, J. Am. Chem. Soc. 2007, 129,
15434–15435.
[13] a) R. N. Gupta, M. Castillo, D. B. MacLean, I. D. Spenser, J. T. Wrobel,
J. Am. Chem. Soc. 1968, 90, 1360–1361; b) D. B. MacLean in The
alkaloids, (ed. A. Brossi), Academic Press, New York, 1985.
[14] a) E. Leete, S. A. Slattery, J. Am. Chem. Soc. 1976, 98, 6326–6330; b)
L. Nyembo, A. Goffin, C. Hootele, J. C. Braekman, Can. J. Chem.
1978, 56, 851–856.
[39] M. Lafrance, S. I. Gorelsky, K. Fagnou, J. Am. Chem. Soc. 2007, 129,
14570–14571.
[40] G. Kobayashi, S. Furukawa, Pharm. Bull. Japan. 1953, 1, 347–349.
[41] M. Shibuya, M. Tomizawa, I. Suzuki, Y. Iwabuchi, J. Am. Chem. Soc.
2006, 128, 8412–8413.
[15] For selected reviews of the intramolecular Mizoroki–Heck reaction,
see: a) I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100,
3009–3066; b) J. T. Link, Org. React. 2002, 60, 157–534; c) A. B.
Dounay, L. E. Overman, Chem. Rev. 2003, 103, 2945–2963.
[42] a) T. M. Boller, J. M. Murphy, M. Hapke, T. Ishiyama, N. Miyaura, J. F.
Hartwig, J. Am. Chem. Soc. 2002, 124, 390-391. b) J. Takagi, K. Sato,
J. F. Hartwig, T. Ishiyama, N. Miyaura, Tetrahedron Lett. 2002, 43,
5649–5651.
[16] A related strategy was reported to construct spiro compounds, see: S.
Lindström, L. Ripa, A. Hallberg, Org. Lett. 2000, 2, 2291–2293.
[17] a) L. C. Campeau, D. R. Stuart, J. P. Leclerc, M. Bertrand-Laperle, E.
Villemure, H. Y. Sun, S. Lasserre, N. Guimond, M. Lecavallier, K.
Fagnou. J. Am. Chem. Soc. 2009, 131, 3291–3306; b) S. Duric, C. C.
Tzschucke, Org. Lett. 2011, 13, 2310–2313; c) L. C. Campeau, D. J.
[43] C. G. Gutierrez, R. A. Stringham, T. Nitasaka, K. G. Glasscock, J. Org.
Chem. 1980, 45, 3393–3395.
This article is protected by copyright. All rights reserved.

10.1002/chem.201604647
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
Layout 1:
FULL PAPER
Unsymmetric dimers from the
protected monomer: Total syntheses
of the Lycopodium alkaloid lycodine,
and its unsymmetric dimers,
L. Zhao, C. Tsukano*, E. Kwon, H.
Shirakawa, S. Kaneko, Y. Takemoto,
and M. Hirama
Page No. – Page No.
complanadines A and B are described.
Unsymmetric dimerization was
achieved through regioselective
functionalization of the protected
monomer and direct C–H arylation.
Examination of neuronal differentiation
revealed the importance of the dimeric
structure, chirality and oxidation levels
for the biological activity of natural
complanadine A.
Competent route to the unsymmetric
dimer architectures: Total syntheses
of (−)-lycodine, (−)-complanadines A
and B, and evaluation of their neurite
outgrowth activities
This article is protected by copyright. All rights reserved.