Application of the helquist annulation in lycopodium alkaloid synthesis: Unified total syntheses of (-)-8-deoxyserratinine, (+)-fawcettimine, and (+)-lycoflexine

Article abstract of DOI:10.1021/jo1023188

A unified strategy for total synthesis of the Lycopodium alkaloids (-)-8-deoxyserratinine (7), (+)-fawcettimine (1), and (+)-lycoflexine (4) is detailed. The key features include a highly efficient Helquist annulation to assemble the cis-fused 6/5 bicycle

Full text of DOI:10.1021/jo1023188

ARTICLE
pubs.acs.org/joc
Application of the Helquist Annulation in Lycopodium Alkaloid
Synthesis: Unified Total Syntheses of (ꢀ)-8-Deoxyserratinine,
(
þ)-Fawcettimine, and (þ)-Lycoflexine
,†
†,§
†,‡
†,§
†
Yu-Rong Yang,* Liang Shen, Jiu-Zhong Huang, Tao Xu, and Kun Wei
†
State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of
Sciences, Kunming 650204, China
‡
Key Laboratory of Medicinal Chemistry for Natural Resource, Yunnan University, Kunming 650091, China
§
Graduate School of Chinese Academy of Sciences, Beijing 100049, China
S Supporting Information
b
ABSTRACT: A unified strategy for total synthesis of the Lycopodium alkaloids
(
ꢀ)-8-deoxyserratinine (7), (þ)-fawcettimine (1), and (þ)-lycoflexine (4) is
detailed. The key features include a highly efficient Helquist annulation to
assemble the cis-fused 6/5 bicycle, facile construction of the aza nine-membered
ring system employing double N-alkylation strategy, providing access to the
common tricyclic skeleton, asymmetric Shi epoxidation, delivering the desired
β-epoxide stereospecifically to furnish (ꢀ)-8-deoxyserratinine (7), SmI reduc-
2
tion of dihydroxylation derivative 35 to enable formation of (þ)-fawcettimine
(
1), and a rapid biomimetic transformation of (þ)-fawcettimine (1) into (þ)-
lycoflexine (4) via an intramolecular Mannich cyclization.
’
INTRODUCTION
group at C-4, which is known to be removed by reduction with
calcium in ammonia; the total synthesis of alopecuridine has not
been achieved to date.
The Lycopodium alkaloids are a diverse group of structurally
1
complex natural products. Owing both to their appealing, synthe-
tically challenging polycyclic systems with a dense stereochemical
array and wide-ranging biological activities, a wealth of total syn-
theses of Lycopodium alkaloids have been reported. Among the
known Lycopodium alkaloids, fawcettimine (1), alopecuridine
7
Serratinine (3), like fawcettimine, is another important Lycopo-
dium alkaloid. The complicated stereochemistry within its unique
skeleton renders it a synthetic challenge. The first total syntheses of
serratinine (3) and 8-deoxyserratinine (7) were accomplished in
racemic forms by the Inubushi group in 1974 and 1979, respec-
2
(
2), serratinine (3), and lycoflexine (4) are noteworthy due to
8
tively. Very recently, we have successfully achieved the first asym-
their unique tetracyclic structures containing two contiguous qua-
metric total synthesis of (ꢀ)-8-deoxyserratinine (7) via a novel
efficient route by which some of the critical issues such as synthetic
efficiency and selectivity have been improved remarkably com-
ternary stereocenters and the interesting biogenetic interconnec3-
tions within this intriguing family. Fawcettimine (1) (Figure 1),
one of the parent members of Lycopodium alkaloids, was first
isolated by Burnell in 1959. Since the first synthesis by Inubushi
in 1979, a number of other syntheses have been reported. One
of the seminal works in this regard came from the Heathcock
group. They reported a protecting-group-free synthesis and
investigated thoroughly the relationship of tautomeric ringꢀchain
equilibria between keto carbinolamines and the corresponding diketo
amines. In Toste’s recent asymmetric synthesis of fawcettimine, a
signature gold(I)-catalyzed cyclization to form the 6/5 bicycle
was utilized successfully. Alopecuridine (2), the C-4 hydroxy de-
9
10
pared to its previous route. Lycoflexine (4), an unusual 17-
carbon Lycopodium alkaloid possessing a carbonꢀnitrogen ske-
leton differing from those of other members of the fawcettimine
class, was first isolated by the Ayer group in 1973. Lycoflexine (4)
differs from fawcettimine (1) in that the nitrogen and C-4 are
linked through a methylene bridge. Ayer proposed that a transan-
nular Mannich cyclization of fawcettimine (1) could afford lyco-
flexine (4) along biogenetic lines. Of the aforementioned alkaloids,
4
5
5
a,b
5c
6
rivative of fawcettimine (1), was isolated by Ayer in 1974. However,
alopecuridine is almost identical to fawcettimine but for a hydroxy
Received: November 24, 2010
Published: April 07, 2011
r 2011 American Chemical Society
3684
dx.doi.org/10.1021/jo1023188
|J. Org. Chem. 2011, 76, 3684–3690

The Journal of Organic Chemistry
ARTICLE
Scheme 2. Representative Synthetic Strategies for
Hydrindanone of Lycopodium Alkaloids
Figure 1. Structures of some Lycopodium alkaloids.
Scheme 1. Retrosynthetic Analysis of 8-Deoxyserratinine
(7), Alopecuridine (2), Fawcettimine (1), and Lycoflexine (4)
alopecuridine (2) was initially believed to be the cornerstone of our
research program in terms of synthetic diversity. We confidently
expected that both fawcettimine (1) and lycoflexine (4) could be
achieved via facile deoxygenation at C-4 of alopecuridine (2).
Moreover, we considered that if we accomplished the synthesis of
alopecuridine (2), then the biomimetic syntheses of novel alkaloids
11
12
sieboldine A (6) and lycojapodine A (5) would be possible
because alopecuridine (2) is their logical biosynthetic precursor,
thereby allowing accessibility to a maximal diversity of Lycopodium
alkaloids. To this end, a flexible and general strategy is needed. For-
tunately, we found that our developed strategy for (ꢀ)-8-deox-
yserratinine (7) could well serve this purpose. In this full account, we
describe our efforts in this regard, which culminated in the unified
total syntheses of (ꢀ)-8-deoxyserratinine (7),(þ)-fawcettimine (1),
and (þ)-lycoflexine (4).
diversity. The key aza nine-membered ring of 8 was expected to
be constructed by double N-alkylations of the bisiodide com-
13
pound 9 with a nitrogen atom. The iodo side chains of 9 were
envisioned to arise from bisallyl substitutes. The hydrindanone
14
core of 10 would be forged through the Helquist annulation of
known 2-allyl-5-methyl-cyclohex-2-en-1-one (11).
Synthesis of Hydrindanone 6/5 Bicyclic System. Given the
importance of hydrindanone motif to the strategic planning,
various methods for its synthesis have been devised. It is necessary
to summarize them briefly prior to introduction of our work. As
4
shown in Scheme 2, in Inubushi’s pioneering synthesis, they
’
RESULTS AND DISCUSSION
recognized the possibility to control the stereochemistry of the
challenging quaternary stereocenter of 12 by using a Dielsꢀ
Alder addition between butadiene and racemic 2-allyl-5-methyl-
cyclohex-2-en-1-one (11) (29% yield). Scission of the double
bond of the newly formed six-membered ring of 12 followed by
reclosure of the five-membered ring via aldol and Wadsworthꢀ
Emmons reaction afforded substituted 6/5 bicycle 13. In Heathcock’s
approach, HosomiꢀSakurai reaction of cyano enone 14 with
bis-silane 15 provided an adduct intermediate (not shown) that
underwent oxidation and Wittig condensation to produce dieno-
ate 16. Intermediate 16 was treated with ethanolic sodium
ethoxide, leading to 6/5 bicyclic system 17 via an intramolecular
Retrosynthetic Analysis of 8-Deoxyserratinine (7), Alope-
curidine (2), Fawcettimine (1), and Lycoflexine (4). Our syn-
thetic design is outlined in Scheme 1. We proposed to acquire the
four alkaloids from a common advanced intermediate 8. The
CꢀC double bond of tricycle 8 is the pivot of our unified strategy
because it was envisioned either to be converted into an epoxide
by which a cascade epoxide-opening reaction can take place to
install the tetracyclic system of 8-deoxyserratinine (7), or to be
dihydroxylated to yield alopecuridine (2). As we discussed above,
alopecuridine (2) was an ideally versatile natural product en route to
other members of this family from the standpoint of synthetic
5a,b
3
685
dx.doi.org/10.1021/jo1023188 |J. Org. Chem. 2011, 76, 3684–3690

The Journal of Organic Chemistry
ARTICLE
Scheme 3. Our Application of the Helquist Annulation in Cis
Scheme 4. Synthesis of Common Aza Tricyclic
Intermediates 8
6
/5 Bicycle Synthesis
Michael reaction. Very recently, Jung and co-workers reported a
5e
formal synthesis of fawcettimine based on Heathcock’s results.
In the Jung route, cyano enone 14 and silyl enol ether 18 were
treated with triflimide to give ketone 19 stereospecifically. Wittig
methenylation of 19 afforded the olefin (not shown), which was
transformed into 6/5 bicyclic annulation product 20 through
cyclopropane ring-opening under Sc(OTf) conditions. Another
3
synthetic approach toward hydrindanone has been reported by
Synthesis of Common Aza Tricyclic Intermediates 8. As
outlined in Scheme 4, allyl bromide Grignard reagent attacked
the cyclopentanone 10, providing the carbinol (not shown) in
quantitative yield, which was subsequently exposed to thionyl
chloride/pyridine and was dehydrated readily with complete
regioselectivity, furnishing the endocyclic olefin 27 as the only
product (80%). Prior to closure of the critical aza nine-membered
ring, the key intermediate bisiodide 9 was prepared through iodina-
tion of the two corresponding terminal hydroxyl groups, which
were obtained via selective hydroboration of the triene 27 with
9-BBN (70%, two steps) (Scheme 4). Through treatment of this
5c
Toste et al. recently, who employed a sequence of conjugated
propargylation, acetylene iodination, and gold(I)-catalyzed cy-
clization to construct hydrindanone derivative 23 in their elegant
total synthesis of (þ)-fawcettimine.
Our solution to this hydrindanone was originally inspired by
Helquist’s work, who developed a methodology for bicyclic annula-
tion via copper-catalyzed conjugated addition of an acetal-con-
taining Grignard reagent. To our knowledge, Helquist annulation
was still an underutilized approach for the synthesis of polycyclic
systems, and prior to us there had been no one who enlisted it to
construct the framework of Lycopodium alkaloids. As outlined in
Scheme 3, our synthesis began with (ꢀ)-2-allyl-5-methyl-cyclohex-
bisiodide with TsNH
as the nucleophilic nitrogen source in the
2
presence of Bu NI and NaOH in refluxing benzene, a smooth
4
2
-en-1-one (11), which could be prepared in multigram quan-
double N-alkylative ring closure was realized in a respectable 60%
15
16
tities according to a known procedure. With this enone in hand,
we then tried the Helquist annulation and were gratified to find it
was very successful. Conjugate addition of the acetal-containing
Grignard reagent 24 derived from 2-(2- bromoethyl)-1,3-dioxo-
lane to enone 11 performed well to provide trans silyl enol ether
yield. Removal of the tosyl group of 28 was effected by the use
of sodium naphthalenide as reducing agent, providing amine 29
17
nearly quantitatively. This amine is another key compound for
our strategic flexibility. Trifluoroacetylation or Boc-protection of
this amine and subsequent cleavage of the cyclic acetal could be
carried out in one pot and afforded keto trifluoroacetyl amide 8a
and carbamate 8b, respectively, in 70% yield.
2
5 diastereoselectively (93%). This enol ether, upon treatment
with warm 2 N HCl, successfully underwent cascade desilylation,
acetal hydrolysis, and intramolecular aldol cyclization to give a
separable 2.5:1 mixture of cis-fused 6/5 alcohols (75%) having
the correct configuration of the quaternary stereocenter. Oxida-
tion of a mixture of the Helquist annulation products with PCC
followed by brief treatment with ethylene glycol via azeotropic
distillation afforded selectively protected cyclic acetal 10, in
which the carbonyl group within the cyclopentanone ring was
untouched (88%). In contrast to prior precedents, our Helquist
annulation strategy still has some merits: it is stereospecific, and
the product diketone 26 contains all the desired stereochemistry.
Further, the derivative 10 has a versatile cyclopentanone func-
tionality, which can serve a dual purpose as an efficient cross-
coupling part via enol triflation (equivalent to vinyl iodide 23) or
as a substrate for organometallic reagent 1,2-addition (i.e., allyl-
magnesium bromide) to elongate the side chain.
Completion of Synthesis of (ꢀ)-8-Deoxyserratinine (7).
With keto trifluoroacetyl amide 8a in hand, the stage was now set
for the challenging selective epoxidation. However, as demon-
strated in Inubushi’s synthesis, this manipulation posed a pro-
blem, as the standard m-CPBA procedure provided only minor
amounts of β-epoxide 31 accompanied by its R-isomer as the
18
predominant product. This stereochemical outcome can be
attributed to approach of the reagent from the convex face of the
tricyclic system. After considerable efforts, it was found that this
issue could be addressed satisfactorily through Shi asymmetric
19
epoxidation with Shi D-fructose-derived catalyst 30. A complete
stereoselective epoxidation took place to provide the desired
β-epoxide 31 as the sole product in 60% yield along with the
recovered starting material 8a (96% yield, brsm). For completion
of the 8-deoxyserratinine, the β-epoxide 31 was converted into
3
686
dx.doi.org/10.1021/jo1023188 |J. Org. Chem. 2011, 76, 3684–3690

The Journal of Organic Chemistry
ARTICLE
Scheme 5. Synthesis of (ꢀ)-8-Deoxyserratinine (7)
Scheme 6. Attempted Synthesis of Alopecuridine (2) and
4-epi-Alopecuridine
tetracyclic diketone 32 via Inubushi’s protocol, that is, tandem
cleavage of trifluoroacetamide and intramolecular epoxide ring-
opening reaction to furnish tetracyclic amino alcohol (not shown)
in a refluxing methanolic KOH solution and subsequent Jones
oxidation of the amino alcohol intermediate. At this stage, we
were pleased to find our diketone 32 was identical with the literature
2
0
compound, which was obtained from the degradation of natural
(
ꢀ)-serratinine (see Supporting Information). Selective reduc-
tion of the carbonyl functionality of the six-membered ring through
careful addition of 2 equiv of NaBH at 0 °C was realized, pro-
The resulting diketone 35, after removal of the Boc group with
TFA, would be expected to give 4-epi-alopecuridine via carbino-
lamine formation at C-13 (Scheme 6). At the outset, we did not
notice another possibility of carbinolamine formation at C-5 of
cyclopentanone and then tried the oxidative cleavage for the synthesis
of lycojapodine A but with no success. Later, with the aid of 2D NMR
analysis, the correct structure of this cyclization was reassigned as keto
carbinolamine 36. Note that Heathcock’s investigation into the
chemoselectivty of carbinoamine formation at cyclopentanone being
preferred over that at cyclohexanone for 4-epi-fawcettimine is
applicable to our 4-epi-alopecuridine (see Figure 2).
4
viding 8-deoxyserratinine (7). The synthetic 8-deoxyserratinine
(
7) exhibited a rotation of ꢀ14.2 (c 0.38, EtOH), essentially
21
identical to that of the natural substance, and its structure was
again secured by single-crystal X-ray analysis (see Supporting
Information). Therefore, the absolute configuration of 8-deoxy
serratinine (7) has been established as shown in Scheme 5.
Attempted Synthesis of Alopecuridine (2) and 4-epi-Alo-
pecuridine. For installation of the two oxygen atoms at the
C4ꢀC5 bond of alopecuridine (2), cis-dihydroxylation of the
double bond of C4ꢀC5 was expected. As shown in Scheme 6,
this routine transformation initially proved difficult in substrate
5b
Completion of the Synthesis of Fawcettimine (1) and
Lycoflexine (4). Although the pursuit of alopecuridine (2) en
route to lycojapodine A (5) stopped, as shown in Scheme 7,
22
8
b and its analogues because the double bond of C4ꢀC5 was
resistant to attack by oxidative reagents, presumably because of
strong steric shielding exerted by the aza nine-membered ring.
Eventually, use of pyridine as solvent with stoichimetric OsO4
proved effective. To our surprise, the product was not the desired
diol 33. The configuration of the C-4 was incorrect compared to
that of alopecuridine (2), and simultaneously the lactol was
cyclized in the course of dihydroxylation. Single-crystal X-ray
analysis (see the Supporting Information) confirmed our struc-
tural elucidation. To invert the selectivity of dihydroxylation and
avoid the lactol formation, exposure of the corresponding cyclic
acetal of 8b to analogous conditions provided no observable
product even at elevated temperature and prolonged reaction
time. This rigid tetracyclic lactol framework of 34 was quite stable
and inert to several oxidative reagents. Fortunately, Ley oxidation
gratifyingly, SmI reduction of the R-hydroxy ketone 35 gave
2
24
tricyclic diketone 37 efficiently as we anticipated. The resulting
configuration of C-4 of 37 was inconsequential and remained
unassigned according to Heathcock’s conclusion that the un-
natural 4R isomer, after removal of Boc group, could also be
epimerized and spontaneously cyclized into fawcettimine. The
spectral data and optical rotation for the synthetic fawcettimine
2
5
(1) were consistent with those reported in the literature.
With synthetic fawcettimine (1) in hand, and inspired by Ayer’s
hypothesis about the biogenetic pathway of lycoflexine (4), we
naturally turned our attention to the biomimetic synthesis of lyco-
26
flexine (4) based on Ayer’s proposal of Mannich cyclization. Unlike
Ayer’s original procedure (acidic aqueous HCHO, 24 h, ca. 50%),
we found a better choice for carrying out this reaction (Scheme 8).
23
(TPAP, NMO) afforded tricyclic diketone 35 in 90% yield.
3
687
dx.doi.org/10.1021/jo1023188 |J. Org. Chem. 2011, 76, 3684–3690

The Journal of Organic Chemistry
ARTICLE
sample ([R] þ57.4, c 0.42, MeOH); therefore, the absolute
D
28
configuration of lycoflexine (4) was established.
In summary, we have developed an efficient and general strategy
to Lycopodium alkaloid synthesis, furnishing (ꢀ)-8-deoxyserra-
tinine (7), (þ)-fawcettimine (1), and (þ)-lycoflexine (4), in which
Helquist annulation and double N-alkylations play the key roles
in construction of the common framework. Stereospecific Shi ep-
oxidation for 8-deoxyserratinine (7) and unexpected dihydroxyla-
tion, Ley oxidation, and SmI reduction for fawcettimine (1) were
2
utilized. Biomimetic synthesis of (þ)-lycoflexine (4) is also pre-
sented based on a rapid intramolecular Mannich cyclization.
’
EXPERIMENTAL SECTION
Figure 2. Selectivity of carbinoamine formation for 4-epi-fawcettimine
and 4-epi-alopecuridine.
General Methods. All reactions were performed under a nitro-
gen atmosphere with dry solvents under anhydrous conditions, unless
otherwise noted. Dry tetrahydrofuran (THF) and diethyl ether (Et O)
Scheme 7. Synthesis of (þ)-Fawcettimine (1)
2
were distilled over sodiumꢀpotassium alloy. Dichloromethane, toluene,
dimethylformamide (DMF), acetonitrile, and benzene were distilled
over calcium hydride. Yields refer to chromatographically and spectro-
scopically homogeneous materials, unless otherwised noted. Reagents
were used as received without further purification, unless otherwise
stated. Silica gel (200ꢀ300 mesh) was used for flash column chroma-
tography. NMR spectra were recorded on 400 or 500 MHz spectro-
meters and calibrated using undeuterated solvent as an internal reference. IR
spectra were recorded with KBr pellets. Optical rotations were deter-
mined on a polarimeter. HRMS data were recorded via positive ion
electrospray or electron impact mass spectrometry using a time-of-flight
analyzer.
Silyl Enol Ether 25. To a stirred mixture containing 1.02 g (42.5
mmol) of fresh magnesium turnings in 16.5 mL of THF were added a
small crystal of iodine and then a solution of 2-(2-bromoethyl)-1,3-
dioxolane (8.06 g, 44.5 mmol) in 45 mL of THF over 25 min; a water
cooling bath was used to maintain the reaction temperature below 25 °C.
The resulting Grignard reagent mixture continued to stir for 1 h at rt
before being transferred into a stirred precooled mixture of 4-DMAP
(5.54 g, 45.4 mmol) and CuBr Me S (0.856 g, 4.1 mmol) in 30 mL of
THF at ꢀ78 °C via a cannula. Next, the resulting reaction mixture was
stirred at ꢀ78 °C for 1 h, and then TMSCl (5.6 mL) and a solution of
enone 11 (3.07 g, 20.5 mmol) in 30 mL of THF were added dropwise,
respectively, at the same temperature. The reaction was allowed to warm
slowly to ꢀ10 °C over 6 h and finally quenched by addition of 16.8 mL of
triethylamine and diluted with 100 mL of petroleum ether and 50 mL of
water. The resulting slurry was filtered through Celite, which was washed
Scheme 8. Biomimetic Synthesis of (þ)-Lycoflexine (4)
2
3
with petroleum ether and Et O (1:1 mixture). The phases were separated,
2
and the aqueous layer was extracted with ether. The combined organic
layers were dried with Na
purified with flash column chromatography (petroleum ether/ether
00:1, 1% triethylamine) to provide 6.20 g (93%) of 25 as a clear oil:
2 4
SO and concentrated. The crude product was
1
ꢀ
1
[
R]
D
= þ97.4 (c = 0.31, CHCl
3
); IR (KBr) νmax (cm ) 2953, 2910,
1
2
5
4
1
878, 1675, 1456, 1410, 1251, 1201, 1142, 1037, 842; H NMR (CDCl ,
3
00 MHz) δ 5.71 (1H, dt, J = 17.1, 7.9 Hz), 4.94 (2H, m), 4.81 (1H, t, J =
.6 Hz), 3.96 (2H, t, J = 6.5 Hz), 3.83 (2H, t, J = 6.5 Hz), 3.09 (1H, dd, J =
4.7, 5.4 Hz), 2.50 (1H, dd, J = 14.6, 7.7 Hz), 2.06ꢀ2.01 (2H, m), 1.87
Upon treatment with paraformaldehyde in heated 3-methylbutan-
27
1
-ol (120 °C oil bath), synthetic fawcettimine (1) was com-
(
1H, m), 1.74ꢀ1.52 (5H, m), 1.27 (1H, m), 1.17 (1H, dt, J = 12.5, 5.5 Hz),
pletely transformed into lycoflexine (4) instantly (less than 5 min
by TLC monitoring) and cleanly, in 95% isolated yield. The
remarkable ease of the transannular Mannich cyclization supports
Ayer’s working hypothesis of a biomimetic process (Scheme 8).
The synthetic lycoflexine (4) showed NMR spectra identical to
those of authetic sample, generously provided by Prof. Zhao of
KIB, which was reisolated in the course of discovery of lycoja-
podine A (5). Our synthetic lycoflexine (4) exhibited a rotation
of þ55.2 (c 0.45, MeOH) and was consistent with that of Zhao's
13
0
1
2
3
.93 (3H, d, J = 6.5 Hz), 0.15 (9H, s); C NMR (CDCl , 125 MHz) δ
3
44.2, 137.2, 116.8, 114.5, 104.8, 64.8, 64.7, 38.6, 35.6, 34.5, 32.6, 32.3,
6.9, 24.8, 21.8, 0.7; HR ESI m/z calcd for C18H O
33 3
Si [M þ H]þ
25.2198, found 325.2196.
6
/5 Bicyclic Diketone 26. To a solution of silyl enol ether 25
(6.20 g, 19.14 mmol) in 130 mL of THF was added 32 mL of 2 N
aqueous HCl at rt. The reaction mixture was allowed to stir at rt for 1 h
and then was placed in a 50 °C oil bath and stirred for 1.5 h. After being
cooled to rt, the reaction mixture was extracted with ether. The combined
3
688
dx.doi.org/10.1021/jo1023188 |J. Org. Chem. 2011, 76, 3684–3690

The Journal of Organic Chemistry
ARTICLE
extracts were washed with saturated NaHCO
3
and brine, dried over
(4H, m), 2.55ꢀ1.64 (15H, m), 1.46 (9H, s), 1.37ꢀ1.31(1H, m), 1.06
1
3
Na SO , and concentrated, and the crude product was purified with flash
(3H, d, J = 6.5 Hz); C NMR (CDCl , 125 MHz) δ 220.4, 216.6, 157.4,
2
4
3
column chromatography (petroleum ether/ethyl acetate 20:1 f 5:1) to
provide a 2.97 g (75%) mixture of alcohol isomers. To a solution of
mixture of alcohol isomers (2.97 g, 14.3 mmol) in DCM (120 mL) were
added Celite (5.0 g) and PCC (5.22 g, 24.3 mmol). The reaction mixture
was stirred for 4 h at rt and then was filtered through a short pad of Celite.
The collected filtrate was further purified with flash column chromatogra-
phy (petroleum ether/ethyl acetate 50:1) to provide 26 (2.89 g, 98%) as
83.2, 79.3, 59.3, 46.9, 46.3, 44.6, 39.6, 35.0, 31.3, 28.5, 26.3, 24.5, 23.2,
þ
21.9, 18.9; HR EI m/z calcd for C21
H
33NO
5
[M] 379.2359, found
379.2352.
Tricyclic Diketone 37. To a stirred solution of 35 (10.0 mg, 0.026
mmol) in THF/t-BuOH (4:1) was added samarium diiodide (0.1 M in
THF, 0.87 mL, 0.087 mmol). The resulting solution was stirred at rt for
10 min, opened to air, quenched with addition of saturated NaHCO3,
extracted with ether (10.0 mL ꢁ 3), washed with brine (1.0 mL ꢁ 2),
ꢀ1
a colorless oil: [R]
078, 2970, 2916, 1749, 1730, 1291, 1231, 1121, 1103, 920; H NMR
CDCl , 500 MHz) δ 5.67 (1H, m), 5.09 (2H, m), 2.58 (1H, m), 2.49
2H, d, J = 7.5 Hz), 2.39 (1H, m), 2.30ꢀ2.20 (2H, m), 2.14 (2H, m),
D
= þ101.7 (c = 0.5, CHCl
3
); IR (KBr) νmax (cm )
1
3
2 4
dried over Na SO , and concentrated. The crude product was purified
(
(
1
3
with flash column chromatography (petroleum ether/ethyl acetate 10:1)
ꢀ
1
to provide 37 (7.8 mg, 81%) as a white solid: IR (KBr) νmax (cm ) 3440,
13
1
.96 (1H, m), 1.74 (3H, m), 1.03 (3H, d, J = 6.5 Hz); C NMR (CDCl
3
,
2957, 2927, 1744, 1692, 1480, 1414, 1366, 1167, 1128, 772; H NMR
1
25 MHz) δ214.2, 208.0, 132.8, 118.9, 66.8, 47.0, 41.1, 35.9, 32.9, 30.0, 24.0,
(CDCl , 500 MHz) δ 3.77ꢀ3.50 (2H, m), 2.94ꢀ2.74 (2H, m), 2.44ꢀ1.59
3
þ
13
21.5; HR EI m/z calcd for C H O [M] 206.1307, found 206.1323.
(17H, m), 1.41 (9H, s), 1.05 (3H, dd, J = 6.4, 4.0 Hz); C NMR
13 18 2
Double N-Alkylative Ring Closure Product 28. A 100 mL
three-necked flask was equipped with a reflux condenser and two 25 mL
dropping funnels. The flask was charged with benzene (30.0 mL), NaOH
3
(CDCl , 100 MHz) δ 215.2, 214.7, 156.8, 79.7, 60.37, 59.9, 50.1, 49.4,
47.3, 46.9, 46.6, 44.7, 44.5, 42.5, 42.2, 38.9, 38.3, 31.6, 30.9, 30.1, 28.5,
28.4, 25.5, 23.4, 22.3, 22.1, 22.0, 21.6, 21.0.
(
1.18 g, 29.5 mmol), Bu
4
NI (73 mg, 0.20 mmol), and H
2
O (2.8 mL).
Fawcettimine (1). To a solution of 37 (4.5 mg, 0.0123 mmol) in
The mixture was refluxed at 100 °C, and then a solution of bisiodide 9
CH
resulting solution was stirred at rt for 3 h, quenched with addition of
saturated NaHCO
, and extracted with chloroform (10.0 mL ꢁ 3), and
the combined organic phases were washed with saturated NaHCO
SO , and con-
2 2
Cl (1.0 mL) was added trifluoroacetic acid (0.050 mL) at 0 °C. The
(
75.0 mg, 0.14 mmol) in benzene (7.0 mL) and a solution of TsNH
41.0 mg, 0.24 mmol) in benzene (7.0 mL) were added dropwise
2
(
3
through two dropping funnels over 30 min, respectively. The reaction
3
mixture was refluxed for 2 h, cooled to rt, poured into H O (10.0 mL),
(2.0 mL ꢁ 2) and brine (2.0 mL ꢁ 2), dried over Na
2
4
2
extracted with ether (20 mL ꢁ 3), washed with brine (10 mL ꢁ 2), dried
centrated. The crude product was purified with flash column (chloroform/
methanol 15:1) to provide fawcettimine (2.6 mg, 80%) as a yellowish
with Na SO , concentrated in vacuo, and flash chromatographed (petroleum
2
4
ꢀ
1
ether/ethyl acetate 20:1) to provide 28 (38 mg, 60%) as a colorless oil:
foam: [R]
2
D
= þ85.4 (c = 0.67, MeOH); IR (KBr) νmax (cm ) 3425,
1
ꢀ
1
953, 2923, 2867, 2689, 2620, 1734, 1460; H NMR (CDCl , 500 MHz)
[
2
(
5
R]
948, 2919, 1599, 1456, 1339, 965, 937, 814, 714, 548; H NMR
CDCl , 400 MHz) δ 7.68 (2H, d, J = 8.4 Hz), 7.29 (2H, d, J = 8.0 Hz),
.66 (1H, m), 3.85 (4H, m), 3.45 (1H, m), 2.96 (1H, m), 2.76ꢀ2.68
D
= þ67.4 (c = 0.7, CHCl
3
); IR (KBr) νmax (cm ) 3432, 3027,
3
1
δ 4.95 (1H, s, br), 3.80ꢀ2.80 (4H, m), 2.61 (1H, dd, J = 17.5, 14.0 Hz),
1
3
2
.32ꢀ1.83 (10H, m), 1.74ꢀ1.38 (6H, m), 0.98 (3H, d, J = 6.5 Hz);
C
3
3
NMR (CDCl , 125 MHz) δ 218.1, 77.2, 59.5, 54.1, 50.2, 47.8, 43.1, 41.7,
41.4, 34.4, 31.5, 27.3, 26.1, 23.6, 21.5, 20.6.
(
1H, m), 2.60ꢀ2.53 (1H, m), 2.41 (3H, m), 2.23ꢀ1.13 (16H, m),
13
Lycoflexine (4). To a solution of synthetic fawcettimine (1) (11.0
0
1
3
.89ꢀ0.85 (3H, m); C NMR (CDCl , 125 MHz) δ 145.6, 142.9, 134.7,
3
mg, 0.041 mmol) in isoamyl alcohol (1 mL) was added paraformalde-
hyde (11.0 mg, 0.366 mmol). The mixture was placed in a 120 °C oil
bath, stirred for 5 min, cooled to rt, filtered, and concentrated. The crude
product was purified with flash column chromatography (chloroform/
methanol 50:1) to provide lycoflexine (11.0 mg, 95%) as a yellow solid:
29.4, 129.0, 127.6, 113.0, 64.6, 63.0, 58.0, 50.1, 43.8, 41.1, 38.9, 36.7,
3.7, 29.5, 25.0, 23.7, 22.6, 22.3, 22.2, 21.5; HR EI m/z calcd for
þ
C
25
H
35NO
4
S [M] 445.2287, found 445.2297.
Tetracyclic Hemiacetal 34. To a solution of 8b (18.3 mg, 0.0527
mmol) in pyridine was added OsO (4% in water, 0.402 mL, 0.0632 mmol)
4
ꢀ
1
[
2
(
2
R]
868, 1727, 1697, 1631, 1459; H NMR (CDCl
2H, m), 3.02ꢀ2.94 (1H, m), 2.90ꢀ2.77 (2H, m), 2.70ꢀ2.60 (2H, m),
.39ꢀ2.07 (8H, m), 1.99ꢀ1.71 (5H, m), 1.63ꢀ1.53 (1H, m), 1.40ꢀ1.32
D
= þ55.2 (c = 0.45, MeOH); IR (KBr) νmax (cm ) 3431, 2925,
at rt. The resulting mixture was stirred for 4 h at rt and then quenched by
1
3
, 500 MHz) δ 3.22ꢀ3.11
saturated Na S O solution. The mixture was stirred at 45 °C for 4 h,
2
2 3
extracted with ether (10.0 mL ꢁ 3), washed with brine (1.0 mL ꢁ 2),
2 4
dried over Na SO , and concentrated. The crude product was purified
1
3
(
1H, m), 1.03 (3H, d, J = 6.5 Hz); C NMR (CDCl
3
, 125 MHz) δ
with flash column chromatography (petroleum ether/ethyl acetate 10:1) to
2
17.9, 213.7, 60.5, 58.2, 56.6, 53.3, 53.2, 46.7, 40.3, 39.9, 36.0, 31.2, 29.0,
provide 34 (17.1 mg, 85%) as a colorless crystal: mp 156ꢀ157 °C; [R]
D
þ
ꢀ1
28.0, 25.7, 22.3, 19.2; HR ESI m/z calcd for C H NO [M þ H]
=
þ18.8 (c = 0.53, CHCl
3
); IR (KBr) νmax (cm ) 3285, 2966, 2931,
17 26
2
1
276.1963, found 276.1961.
1
5
(
(
(
3
688, 1470, 1413, 1172, 1151, 1121, 967, 779, 758; H NMR (CDCl ,
00 MHz) δ 5.34 (1H, s, br), 4.28 (1H, s, br), 3.73 (1H, s), 3.48ꢀ3.32
2H, m), 3.03 (1H, d, J = 14.0 Hz), 2.94 (1H, t, J = 12.5 Hz), 2.24ꢀ2.00
2H, m), 1.91ꢀ1.52 (9H, m), 1.43 (9H, s), 1.40ꢀ0.92 (5H, m), 0.86
’ ASSOCIATED CONTENT
13
S
3H, d, J = 6.5 Hz); C NMR (CDCl
3
, 100 MHz) δ 156.6, 107.7, 85.5,
b
Supporting Information. NMR spectra and CIF files for
8
4.1, 79.5, 50.5, 49.6, 48.4, 39.4, 33.2, 33.1, 31.1, 28.7, 28.5, 28.3, 26.4, 23.4,
synthetic 8-deoxyserratinine (7) and compound 34. This material is
þ
21.3; HR EI m/z calcd for C21
H
35NO
5
[M] 381.2515, found 381.2517.
available free of charge via the Internet at http://pubs.acs.org.
Tricyclic r-Hydroxyketone 35. To a stirred solution of 34 (31.0
mg, 0.081 mmol) in DCM were added tetrapropylammonium perruthe-
’ AUTHOR INFORMATION
nate (15.0 mg, 0.041 mmol) and N-methylmorpholine N-oxide (0.051 mL,
Corresponding Author
*
4
.8 N, 0.243 mmol). The resulting solution was stirred at rt for 10 min,
quenched with addition of saturated NaHSO , extracted with ether
SO , and
E-mail: yangyurong@mail.kib.ac.cn.
3
(10.0 mL ꢁ 3), washed with brine (1.0 mL ꢁ 2), dried over Na
2
4
concentrated. The crude product was purified with flash column
’
ACKNOWLEDGMENT
chromatography (petroleum ether/ethyl acetate 20:1) to provide 35
(
(
1
27.6 mg, 90%) as a clear oil: [R]
D
= þ137.0 (c = 0.45, CHCl
3
); IR
We thank National Natural Science of China (no. 20802084,
21072200), National Basic Research Program of China (973
Program, no. 2009CB522300), Programs of “One Hundred Talented
ꢀ1
KBr) νmax (cm ) 3461, 2965, 2927, 2873, 1756, 1688, 1486, 1412, 1366,
1
170, 1134; H NMR (CDCl , 500 MHz) δ 5.25 (1H, s), 3.85ꢀ2.82
3
3
689
dx.doi.org/10.1021/jo1023188 |J. Org. Chem. 2011, 76, 3684–3690

The Journal of Organic Chemistry
ARTICLE
People” and “Western Light” Joint Scholar of Chinese Academy
of Sciences, and State Key Laboratory of Phytochemistry and
Plant Resources in West China for financial support. We are
indebted to Prof. Qin-Shi Zhao of KIB for providing an authentic
sample of lycoflexine. Prof. Sheng-Hong Li of KIB is also thanked
for detailed analysis of 2D NMR for compound 36.
(15) Caine, D.; Procter, K.; Cassell, R. A. J. Org. Chem. 1984, 49,
2647–2648.
2 2
(16) Other nitrogen sources such as BnNH and PMBNH were
also tried, but TsNH was finally chosen not only because it can give the
2
best yield but also because it can be readily deprotected into the free
amine 1, leaving the other functionalities of the substrate untouched.
(17) Similar conditions used by the Heathcock group in their
fawcettimine total synthesis. See ref 5b.
(
18) The ratio of undesired R isomer to desired β is high, up to 2:1.
’
REFERENCES
(19) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y. J. Am.
Chem. Soc. 1997, 119, 11224–11235.
(
1) For recent reviews of the Lycopodium alkaloids, see (a) Hirasawa,
(
20) Katakawa, K.; Kitajima, M.; Aimi, N.; Seki, H.; Yamaguchi, K.;
Furihata, K.; Harayama, T.; Takayama, H. J. Org. Chem. 2004, 70,
58–663.
21) The reported optical rotation for natural (ꢀ)-8-deoxyserrati-
nine is [R]
ꢀ15.6 (c 1.0, EtOH). Ishii, H.; Yasui, B.; Nishino, R.;
Harayama, T.; Inubushi, Y. Chem. Pharm. Bull. 1970, 18, 1880–1888.
Y.; Kobayashi, J.; Morita, H. Heterocycles 2009, 77, 679–729. (b) Ma, X.;
Gang, D. R. Nat. Prod. Rep 2004, 21, 752–772.
(
a) Liau, B. B.; Shair, M. D. J. Am. Chem. Soc. 2010, 132, 9594–9595.
b) Chandra, A.; Pigza, J. A.; Han, J.-S.; Mutnick, D.; Johnston, J. N.
J. Am. Chem. Soc. 2009, 131, 3470–3471. (c) Nilsson, B. L.; Overman,
L. E.; Alaniz, J. R.; Rohde, J. M. J. Am. Chem. Soc. 2008, 130,
1297–11299. (d) Kozak, J. A.; Dake, G. R. Angew. Chem., Int. Ed.
008, 47, 4221–4223. (e) Bisai, A.; West, S. P.; Sarpong, R. J. Am. Chem.
Soc. 2008, 130, 7222–7223.
6
2) For selected recent total syntheses of Lycopodium alkaloids, see
(
(
(
D
(
22) DihydroxylationwithAD-mix-R, OsO /NMO,catalyticKMnO ,
4
4
or RuCl failed to give the products.
3
1
2
(
23) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D.
J. Chem. Soc., Chem. Commun. 1987, 1625–1627.
24) For a review of SmI in total synthesis, see Nicolaou, K. C.;
(
2
(
(
3) Burnell, R. H. J. Chem. Soc. 1959, 3091–3093.
4) Harayama, T.; Takatani, M.; Inubushi, Y. Tetrahedron Lett. 1979,
Ellery, S. P.; Chen, J. S. Angew. Chem., Int. Ed. 2009, 48, 7140–7165.
25) Our synthetic sample exhibited a rotation of þ85.4 (c 0.67,
MeOH). Toste reported [R]
þ89 (c 0.55, MeOH).
26) During the early stage of preparation of this manuscript, one
(
2
0, 4307–4310.
(
D
5) (a) Heathcock, C. H.; Smith, K. M.; Blumenkopf, T. A. J. Am.
(
Chem. Soc. 1986, 108, 5022–5024. (b) Heathcock, C. H.; Blumenkopf,
T. A.; Smith, K. M. J. Org. Chem. 1989, 54, 1548–1562. (c) Linghu, X.;
Kennedy-Smith, J. J.; Toste, F. D. Angew. Chem., Int. Ed. 2007,
6, 7671–7673. (d) Otsuka, Y.; Inagaki, F.; Mukai, C. J. Org. Chem.
010, 75, 3420–3426. (e) Jung, M. E.; Chang, J. J. Org. Lett. 2010, 12,
962–2965.
elegant synthesis of lycoflexine has been published by Mulzer et al.:
Ramharter, J.; Weinstabl, H.; Mulzer, J. J. Am. Chem. Soc. 2010,
32, 14338–14339. Mulzer used similar conditions as Ayer did, namely,
acidic aqueous HCHO in extended reaction time (48 h) employed to
accomplish this important biomimetic transformation. The first asymmetric
total synthesis of lycoflexine including the assignment of the absolute
configuration was reported by the Mulzer group.
1
4
2
2
(
6) Ayer, W. A.; Altenkirk, B.; Fukazawa, Y. Tetrahedron 1974, 30,
213–4214.
7) (a) Inubushi, Y.; Ishii, H.; Yasui, B.; Hashimoto, M.; Harayama,
4
(27) A similar protocol adopted by the Evans group in their total
(
synthesis of the Lycopodium alkaloid luciduline; see Scott, W. L.; Evans,
D. A. J. Am. Chem. Soc. 1972, 94, 4779–4780.
T. Chem. Pharm. Bull. 1968, 16, 82–91. (b) Inubushi, Y.; Ishii, H.; Yasui,
B.; Hashimoto, M.; Harayama, T. Chem. Pharm. Bull. 1968, 16, 92–100.
(
28) Mulzer documented [R] þ9.0 (c 1.04, DCM). The absolute
D
(
8) For related research from the Inubushi group, see (a) Inubushi,
Y.; Yasui, H.; Yasui, B.; Hashimoto, M.; Harayama, T. Tetrahedron Lett.
966, 7, 1537–1549. (b) Inubushi, Y.; Ishii, H.; Yasui, B.; Harayama, T.
configuration was obtained by comparison of the CD spectrum with that
of natural product. We remeasured the optical rotation of our samples in
DCM instead of in MeOH. The results are in agreement with those of
Mulzer. Our synthetic lycoflexine showed [R]D þ4.8 (c 0.45, DCM);
1
Tetrahedron Lett. 1966, 7, 1551–1559. (c) Ishii, H.; Yasui, B.; Harayama,
T.; Inubushi, Y. Tetrahedron Lett. 1966, 7, 6215–6219. (d) Inubushi, Y.;
Ishii, H.; Harayama, T.; Burnell, R. H.; Ayer, W. A.; Altenkirk, B.
Tetrahedron Lett. 1967, 8, 1069–1072. (e) Harayama, T.; Ohtani, M.;
Oki, M.; Inubushi, Y. J. Chem. Soc., Chem. Commun. 1974, 827–828.
authentic sample, [R]
D
þ5.0 (c 0.42, DCM).
’
NOTE ADDED AFTER ASAP PUBLICATION
(
f) Harayama, T.; Takatani, M.; Inubushi, Y. Chem. Pharm. Bull. 1980,
2
8, 2394–2402. For other synthetic efforts toward serratinine
References 13 and 26 were corrected in the version that reposted
April 19, 2011.
alkaloids, see (g) Mehta, G.; Reddy, M. S.; Radhakrishnan, R.; Manjula,
M. V.; Viswamitra, M. A. Tetrahedron Lett. 1991, 32, 6219–6222.
(
4
h) Cassayre, J.; Gagosz, F.; Zard, S. Z. Angew. Chem., Int. Ed. 2002,
1, 1783–1785.
9) Yang, Y.-R.; Lai, Z.-W.; Shen, L.; Huang, J.-Z.; Wu, X.-D.; Yin, J.-L.;
Wei, K. Org. Lett. 2010, 12, 3430–3433.
10) Ayer, W. A.; Fukazawa, Y.; Singer, P. P.; Altenkirk, B. Tetra-
hedron Lett. 1973, 14, 5045–5048.
11) (a) Hirasawa, Y.; Morita, H.; Shiro, M.; Kobayashi, J. Org. Lett.
003, 5, 3991–3993. (b) Canham, S. M.; France, D. J.; Overman, L. E.
J. Am. Chem. Soc. 2010, 132, 7876–7877.
12) (a) He, J.; Chen, X.-Q.; Li, M.-M.; Zhao, Y.; Xu, G.; Cheng, X.;
Peng, L.-Y.; Xie, M.-J.; Zheng, Y.-T.; Wang, Y.-P.; Zhao, Q.-S. Org. Lett.
009, 11, 1397–1400. (b) Yang, Y.-R.; Shen, L.; Wei, K.; Zhao, Q.-S.
J. Org. Chem. 2010, 75, 1317–1320.
13) In this regard, the Fukuyama group has made seminal con-
tributions in the 2-nitrobenzenesulfonamide (NsNH ) as an efficient
(
(
(
2
(
2
(
2
nitrogen nucleophile for synthesis of medium- or large-sized cyclic
amines.
(
14) Bal, S. A.; Marfat, A.; Helquist, P. J. Org. Chem. 1982, 47,
5045–5050.
3
690
dx.doi.org/10.1021/jo1023188 |J. Org. Chem. 2011, 76, 3684–3690