Synthetic studies towards total synthesis of bielschowskysin 2

1

Chapter 1
Bielschowskysin: A Structurally and Biologically Interesting
Class of Diterpene Natural Product.
1.1 Introduction and Background to Bielschowskysin.
1.1.1 Isolation and Structural Characterisation.
In 2004, the Rodríguez group at the University of Puerto Rico reported the isolation and
structural elucidation of a new cembrane diterpene, as a colorless crystalline solid from 1.07
Kg of specimen of the sea plume, Pseudopterogorgia kallos, during an expedition to the
South Western Caribbean Sea near the Old Province Island of Columbia.



Figure 1.1 Structure of Bielschowskysin
This new diterpene was named bielschowskysin (1, Figure 1.1)
1a
after Bielschowsky, the
1918 discoverer of Pseudopterogorgia kallos. About 1 Kg of selected specimens of the sea
plume Pseudopterogorgia kallos was partially air dried, frozen, lyophilized, cut into small
pieces and homogenised using methanol and dichloromethane (1:1). Subsequent
concentration, standard solvent partitioning, silica gel chromatography and HPLC
purification yielded bielschowskysin 1 (39.6 mg, 0.024% yield). The structure of 1 was
identified as a highly oxygenated hexacyclic diterpene, as fully characterised by NMR and
single-crystal X-ray analysis, to possess an unprecedented cembrane carbon skeleton.
1

O
O
CH
2

OH
OH
Me
H
H
O
O
Me
HO
OAc
1
2

The molecular formula C
22
H
26
O
9
of bielschowskysin and m/z 374.1368 [M-AcOH] implied
ten-degrees of unsaturation. The
13
C NMR spectrum showed four olefinic resonances and
two carbonyl resonances accounted for four sites of unsaturation and the remaining six sites
of unsaturation must be accounted by the rings. HMBC, HMQC, NOESY, COSY,
1
H,
13
C
and

13
C-DEPT NMR experiments provided the key connectivities, relative stereochemistry of
the 10 chiral centres, and X-ray crystal analysis proved the [9.3.3.0] tetradecane core of 1.
1

The absolute stereochemistry is currently unproven, although may be predicted as shown in
Figure 1.1 by comparison with other cembrane family members.
1.1.2 Biological Activity
The interesting ring-structure of bielschowskysin (1) is matched by its ability to exhibit
significant antiplasmodial activity against three chloroquine-resistant strains of Plasmodium
falciparum at an IC
50
of 10 µg/mL
1a
. As such, the terpene represents a [5-4-9] novel
structural archetype to fight against malaria with a yet undefined mechanism with which to
harness and understand. Unfortunately, the characterization studies of 1 have exhausted all
supplies of bielschowskysin for more in-depth and diverse biological evaluations. Due to its
unique structural properties, biosynthesis and biological activity, we embarked on the total
synthesis of bielschowskysin (1).
1.1.2.1 Antimalarial Agents and Continuing Needs.
According to the WHO report, around 243 million cases of malaria are clinically registered
every year causing over 880,000 deaths
1b,2
. This disease likely originated in Africa and has
widely spread in tropical and subtropical regions over the last several decades.
3

Figure 1.2 Structure of Antimalarial Drugs
4

There are four main parasites responsible for malaria. These are Plasmodium falciparum, P.
vivax, P. ovale, and P. malariae, of which P. falciparum and P. vivax accounts for 95% of the
world cases; one death every 30 seconds. Malaria is re-emerging as the biggest infectious
killer and ranked third among the major disease in causing deaths after pneumococcal acute
respiratory infections (pneumonia) and tuberculosis (TB). Natural sources have been the
choice for a long period to cure the disease. Chloroquinine (3) is largely used in most parts
of the world because it is cheap and effective (Figure 1.2). Artemisinin (2) is a natural
product and has been known for 2000 years
3
. Its current usage is FDA restricted solely for
complicated cerebral malaria cases. In the past 30 years, only one new synthetic material,
mefloquine, has been successfully developed. Plasmodium falciparum is rapidly increasing

its resistance to many of the drugs. Chloroquine is particularly ineffective and the search for
new antimalarial leads has gained more importance. While a number of targets have been
suggested for antimalarial design, there is a continued need to expand our understanding of
Plasmodium infection. In recent years, researchers have developed techniques such as
fluorescent-labelling or GFP fusion analysis to study the host-parasite. Due to drug
resistance, there have been no new classes of antimalarial drugs in the clinical trial since
1996. Recently, the screening of 2 million compounds from GlaxoSmithKline’s library for
the inhibitors of P. falciparum, highlighted around 13000 compounds to show parasite
inhibition by at least 80%, and 8000 compounds against multidrug resistant strains.
1.1.3 Proposed Biosynthesis
The geranylgeranyl diphosphate (GGPP) biosynthetic pathway is a well known source of the
cembrane family of natural products (Schem 1.1). Initial C
1
→C
14
cyclisation of GGPP 17
affords the 14 membered macrocycle of cembrane 18. Ring contraction or cyclisation of
C
7
→C
11
then furnishes the verrillane family
4b
21 and C
6
→C
12
cyclisation would generate the
5

bielschowskyane skeleton 22, a novel class of diterpene. The intricarene skeleton 19 can be
derived via a C
6
→C
11
and C
2
→C
12
cyclisation of cembrane (18) while the providenciane 20
could have originated via a C
2
→C
17
cyclisation
1a, 4a, 5
.






Scheme 1.1 Proposed biosynthesis of bielschowskyane skeleton and related cembranes
1.1.4 Related Diterpene Natural Products
Rodríguez and his co-workers have investigated extensively the Pseudopterogorgia kallos
6

sea plume. The octocoral fauna of West Indies is unique in its profusion of gorgonian corals.
These corals are rich in producing acetogenins, prostanoids, sesquiterpenoids, diterpenoids

and steroids, which are largely unknown from terrestrial sources. Typically the natural
products isolated from these corals display significant antimalarial, anti-inflammatory,
analgesic, and anticancer activities. The common structural features among the genus
Pseudopterogorgia include substituted furans, butenolides, epoxy butenolides, macrocyclic,
ring architectures and substituted isopropenyl groups (Figure 1.3).
6












Figure 1.3 Related Diterpene Natural Products from Pseudopterogorgia species.
In recent years, there have been many reports of isolation and structural characterisation from
the sea plumes Pseudopterogorgia, for example, intricarene (23), kallolide A (24), lophotoxin
(25) rubifolide (31), bipinnatin J (26), ciereszkolide (27), providencin (28), verrillin (29),
kallosin A (30), rubifolide (31) and other natural products from the same source.

7

Scheme 1.2 Rodríguez’s chemical isomerisation of bipinnatin J to kallolides and pinnatins
The diterpenes 23 to 31 all show rearranged carbon skeletons that depend on how a
carbocation precursor is trapped. The Rodríguez group has shown, for example, that
irradiation of bipinnatin J 26 in acetonitrile produces kallolide A 24, pinnatin A 32 and
pinnatin C 33 in a 120:1:6 ratio
10
(Scheme 1.2). Similarly the biosynthesis of plumarellide,
bielschowskysin and verrillin could have originated from bipinnatin J or rubifolide
9
(Scheme
1.3). Oxidation of rubifolide or bipinnatin J may lead to the potential intermediates 35, 38,
40, and [4+2] cycloaddition of diene 35 may produce plumarellide 36 or transannular [2+2]
cycloaddition of 38 for bielschowskysin 1 or intramolecular hetero Michael addition of 40
could be explained for verrillin
11
.
Trauner and Pattenden independently unveiled the synthesis of intricarene through
biosynthetic pathway
12
. Retrosynthetic analysis of intricarene shows that it could be prepared
via bipinnatin J followed by 1,3-dipolar cycloaddition of possible biosynthetic relevance.
8

Scheme 1.3 Biosynthetic conjecture to plumarellide, bielschowskysin and verrillin.
Treating bipinnatin J 26 with m-CPBA followed by acetylation produced the pyranone
acetate 41. Heating the pyranone acetate 41 in DMSO with tetramethyl piperidine yielded
intricarene 23

(Scheme 1.4).
12




Scheme 1.4 Trauner’s biomimetic evidence to the biosynthesis of intricarene.
O
O
O
OH
bipinnatin J (26)
O
O
OH
OH
O
O
H

HO
OH
O
O
OH
OH
H
H
O
O
H
HO
OAc
Bielschowskysin (
1
)
O
O
O
OH
O
O
H
HO
OH
O
O
O
OH
O

O
O
OH
H
H
H
O
OH
O
O
H
HO
HO
O
O
OH
HO
HO
H
H
O
H
H
plumarellide (36)
verrillin (29)
[O]
35
38
40
[4+2]

[2+2]
[O]
[O]
Michael
Cyclisation
9

A proposed biosynthesis of providencin is shown in the Scheme 1.1. The formation of the
cyclobutane ring in providencin appears to involve a Norrish type II rearrangement of
bipinnatin E. This proposed biosynthesis has recently been validated synthetically by the
Pattenden group
13
. Irradiation of unsaturated aldehyde 43 in benzene produced the
cyclobutanol 44 in 19% yield by a C-H insertion reaction (Scheme 1.5).


Scheme 1.5 Pattenden’s cyclobutanol synthesis of providencin.
It is difficult to determine an all-encompassing biosynthetic pathway or propose an exact
chemical sequence to each of these natural products. Clearly, there appears to be
considerable structural diversity (e.g., induced by seasonal or chemotype variations in the sea
plumes) by way of concerted cycloadditions, photochemical oxidations, rearrangements and
singlet-O
2
oxidations.
1.1.5 Relevant Synthetic Efforts
1.1.5.1 Butenolide Construction
Butenolides are oxidised derivatives of furans. Apart from the direct oxidation of furans by
using traditional methods, there is interest in making butenolides 47 in optically pure form
(Scheme 1.6). Marshall adopted allenic esters 46 to construct the butenolide moiety in the
total synthesis of kallolide A

16
and rubifolide
17
. Other methods include selenoxide
elimination
18,19,20
of butyrolactone 49, Wittig reaction
21
, alkylation of silyloxy furan
22
and
ring closingmetathesis
23
(RCM) of vinyl esters 50.
10










Scheme 1.6 Stereo-defined syntheses of chiral γ-butenolides

1.1.5.2 Furan Construction
The strategy for constructing the furan moiety was reported by Paquette in his first synthesis
of psuedopterolide

19
. The furan in question was constructed by condensing glyceraldehyde
51 and a β-keto ester to produce the 2,3,5-substituted furan 52 (Scheme 1.7). Careful
extension of side chains and macrocyclisation led to the furanocembranes. The disadvantage
of this approach was to carry the reactive furan ring throughout the synthesis. Alternatively,
the macrocycle could be constructed first and the furan ring formed later in the synthesis.



Scheme 1.7 Paquette’s synthesis of the furan moiety of acersolide.
OMs
R
2
R
1
HR
1
R
2
RO
O
O
O
R
2
R
1
75-85% yield
1. Hy drolysis
2. AgNO

3
,
Hexanes
70-90% yield
O
O
R
2
R
1
O
O
R
2
R
1
SePh
O
O
R
2
R
1
PhSeBr, LHMDS
60-80% yield
H
2
O
2
, THF

O
O
R
2
R
1
O
O
R
2
R
1
RCM
45 46 47
4748 49
50 47
CO, ROH
Pd(PPh
3
)
4
O
O
O
O
OMe
O
PhS
HOAc, H
2

O, EtOH O
CO
2
Me
OH
SPh
51 52
80 °C
11

Wipf’s group
24
followed a base or palladium catalysed cyclisation of the α-propargyl-β-keto
ester 53 to form the required alkenyl furan 54. A disadvantage of this approach was the poor
E/Z selectivity, which was improved via the facial selectivity of the allene protonation step,
by introducing a TMS group (Scheme 1.8).



Scheme 1.8 Wipf’s late-stage construction of a furan moiety in lophotoxin and pukalide.
Marshall suggested two alternative approaches. The first was to react allenylstannane to an
aldehyde at the macrocyclisation stage (Scheme 1.9). After DMP oxidation of an
allenylalcohol, the allenone 55 was treated with silver nitrate and acid to form the furanyl
macrocycles 56,
16,17,25
.



Scheme 1.9 Marshall’s synthesis of the furan moiety of rubifolide.

In another approach, the macrocyclic ynone 57 was treated with silica gel to provide the furan
58. This method is both mild and flexible (Scheme 1.10).

Scheme 1.10 Marshall’s synthesis of furan moiety from an alkynone β-ketoester
TMS
BzO
CO
2
Me
O
O
CO
2
Me
TMS
K
2
CO
3
, Pd(OAc)
2
, dppf
53 54
CH
3
CN:H
2
O, 84 °C
O
CO

2
t
Bu
O
O
O
CO
2
t
Bu
SiO
2
57 58
12

1.1.5.3 Macrocycle Construction
Paquette
19
was one of the first to construct furanocembranoids using the Nozaki-Hiyama-
Kishi reaction as an elegant key macrocyclisation step. This Cr mediated coupling is
powerful enough to form the 11-oxa and 13-membered carbocycle 60 with high
stereoselectivity (Scheme 1.11).




Scheme 1.11 Paquette’s Cr-mediated Nozaki-Hiyama-Kishi macrocyclisation
17
.
Marshall prepared cembrane macrocycles (Scheme 1.12) via cyclisation of the

allenylstannane 61 with aldehydes, followed by furan ring formation in the synthesis of
rubifolide
17
. During the synthesis of kallolide A, Marshall made the macrocyclic propargylic
ether 64 (Scheme 1.13) by displacement of the propargylic halide 63 with allylic sodium
alkoxide under S
N
2 conditions
16
.



Scheme 1.12 Marshall’s allenylstannane macrocyclisation to the rubifolide framework
15
.

O
CO
2
Me
O
O
OH
H
CrC l
2
, THF
O
CO

2
Me
O
O
O
Br
59 60
11
rt
O
CO
2
Me
O
O
Cr
L
L
O
H
H
Me
H
13

O
Cl
HO
O
H

3
C
O
NaH, 18-C-6
63 64
PhCH
3

Scheme 1.13 Marshall’s macrocyclic etherification of kallolide B
23
.
The Pattenden
18
group utilised Stille coupling to construct the 13 membered macrocycle 66
of deoxylophotoxin from 65 (Scheme 1.14).



Scheme 1.14 Pattenden’s arsenine mediated Stille macrocyclisation to deoxy-lophotoxin
16
.
More recently, Trauner, Rawal and Pattenden adopted Paquette’s macrocyclisation strategy in
their independent syntheses of bipinnatin J 26 (Scheme 1.15).




Scheme 1.15 Trauner, Rawal and Pattenden’s Cr-mediated macrocyclisation of bipinnatin J.

O

O
I
O
OTBS
B u
3
S n
OH
O
O
O
OTBS
OH
A sPh
3
, Pd
2
dba
3
65 66
13
40 °C
14

1.1.6 Sulikowski’s Synthetic Study
Sulikowski
27
and Doroh were the first to report the tetracyclic core 78 of bielschowskysin by
a stereoselective intramolecular [2+2] photocycloaddition of the alkylidene furanone 77
(Scheme 1.16). Their strategy was to make and test a congested precursor for

photocycloaddition studies. Here, the major issue was the geometry of the double bond,
which would affect the course of a stereoselective photocycloaddition. On irradiation of the
alkylidene furanone 77 in acetone the bi-radical species produced obeyed the rule-of-five and
favoured the desired single stereoisomer. Their synthesis (Scheme 1.16) began with the
acetonide methyl ester 68 that was prepared from L-malic acid in three steps. Conversion of
methyl ester to its Weinreb amide followed by methylmagnesium bromide addition yielded
the methyl ketone 69 in 91% yield. Under chelation control, the Grignard addition to ketone
69 with ethynylmagnesium bromide provided a 4:1 separable mixture of propargylic alcohols
70, giving the major diasteromer in good yield. Treatment of the ethynyl carbinol 70 with
mesitaldehyde dimethylacetal in the presence of CSA transformed the acetonide smoothly to
the more stable 6-membered mesitylene acetal 71. Aldehyde formation from alcohol 71
followed by Wittig-Horner homologation with trifluoroethyl phosphono methylacetate and
KHMDS gave the cis-α,β-unsaturated ester 73.
Sonogashira coupling with vinyl iodide 74 with 73 provided the vinyl alcohol 75. Double
oxidation of the alcohol yielded the carboxylic acid 76, and silver nitrate catalysed cyclisation
under Negishi’s conditions produced an alkylidene furanone as a single geometric isomer.
The γ-butenolide 77 was then prepared by cleavage of the mesitylene acetal followed by
cyclisation treatment with aqueous acetic acid at room temperature. Finally, irradiation of the
alkylidene furanone 77 in acetone produced the intramolecular [2+2] photocycloaddition
product 78 in 50% yield as a single stereoisomer.
15

Scheme 1.16. Sulikowski’s synthesis of the tetracyclic core of bielschowskysin
25
.
1.2 Approaches toward related diterpenes
1.2.1 Leo Paquette Approach
The Paquette group was the first to synthesize a furanocembranoid with the synthesis of
dihydropseudopterolide in 1990
19
. Their synthesis began with the substituted furan 52 and
elaboration to a substituted butenolide (Scheme 1.17). This strategy lead to the synthesis of
16

psuedopterolides, gorgiacerone (Scheme 1.17) and acersolide (Scheme 1.18). Isopropylidene
glyceraldehyde was condensed with α-thiomalonate in acetic acid to provide the starting
material, the substituted furan 52. Swern oxidation of alcohol followed by addition of 2-
propenyl magnesium bromide afforded the allylic alcohol 79. Acetylation of the secondary
alcohol and palladium-catalyzed allylic stannylation gave the isomeric tributylstannane 80
(82: 18, E/Z). Erythro selective condensation of allylic stannane with aldehyde 81 in the
presence of BF
3
·Et
2
O produced an alcohol adduct, which followed by CSA mediated
lactonisation, gave the lactone 82 in good yield and diastereoselectivity (7.5: 1 ratio, 78%).
In order to make the butenolide and furan carboxaldehyde, a dianion was generated using
KHMDS and reacted with 2 eq. of phenylselenium chloride to give an α-selenolactone –
selenothioacetal. Following hydrolysis of the selenothioacetal with silver perchlorate, the

selenolactone was oxidatively eliminated with sodium periodate to give the butenolide 83.
Reduction of aldehyde with NaBH
4
followed by mesylation and subsequent displacement
with bromide yielded the furyl bromide 84. Addition of vinyl stannane 85 to the bromide in
the presence of Pd(0) provided all the requisite carbons in 86. Conversion of the
tetrahydropyranyl ether into its bromide was achieved with bis(diphenylphosphino)ethane
tetrabromide, followed by deprotection of TBDPS group to give the alcohol 59.
Prior to macrocyclisation, the alcohol 59 was oxidised with PDC. Finally, Nozaki-Hiyama-
Kishi stereoselective macrocyclisation of 59 was achieved using 10 eq. of CrCl
2
to provide
the dihydropseudopterolide 60 in 20-25% yield. Oxidation of the alcohol provided the
natural product, gorgiacerone 87. Paquette and Astles published a related furanocembranoid
synthesis of the natural product acersolide 92 in 1993 (Scheme 1.18).


17

Scheme 1.17 Paquette’s synthesis of dihydropseudopterolide and gorgiacerone
17
.
18












Scheme 1.18 Paquette’s synthesis of acersolide
26
.
1.2.2 Marshall Approach
Along with Leo Paquette, the Marshall group was actively involved toward the synthesis of
furanocembranoids. Their group strategy was to make the macrocyclic propargylic ether
followed by [2,3]-Wittig rearrangement to construct the furanocembranoid framework
25
. In
addition, the butenolide was constructed via an allenoic acid intermediate in the synthesis of

kallolide B (Scheme 1.19). The synthesis began with epoxidation of the chiral perillyl
alcohol 93 to afford epoxide 94. Periodic oxidative cleavage gave the acid, which was
19

esterified using diazomethane to yield the methyl ester-aldehyde 95. Treatment of 95 with L-
bromo-2-butyne in the presence of tin(II)chloride gave the allenyl alcohol 96. Swern
oxidation and subsequent silver nitrate mediated cyclisation afforded the substituted furan 97.
Conversion of methyl ester to propargylic alcohol 97-99 was affected via Dibal-H reduction,
Corey-Fuchs olefination, addition of BuLi and trapping the acetylide with paraformaldehyde
to afford the alcohol 99. Vilsmeyer-Haack formylation of the furan in 99 with sec-BuLi and
DMF gave the substituted furfuraldehyde 100, and homologation under Still-Gennari
olefination yielded the cis enoate 101. Propargylic alcohol was converted into chloride using
MsCl and Dibal-H reduction of the ester gave the unsaturated alcohol 63. Then the key
macrocyclisation step was solved with NaH in toluene in the presence of 18-C-6 to afford the
O-alkylated macrocyclic propargylic ether 64. [2,3]-Wittig rearrangement with n-BuLi
afforded the allylic propargylic alcohol 102 as a single product with syn diastereoselectivity
in good yield.
With failures in making an anti-diastereomer, the propargylic alcohol 102 was converted into
its mesylate then reacted with Pd, CO and β-trimethylsilyl ethanol to afford the allenoic ester
103. Isomerisation of allenoate with triphenyl phosphine followed by TBAF deprotection of
the TMS group of desired epimer gave the allenoic acid. Finally, silver nitrate catalysed
cyclisation yielded the natural product kallolide B 104.





20

Scheme 1.19 Marshall’s synthesis of kallolide B
23
.
OH
O
HO
O
H
3
CO
O
H
3
CO
2
C
HO

CH
3
C
H
2
C
H
3
CO
2
C
O
OHC
O
O O
OHC
O
CO
2
E t
O
Cl
HO
O
H
3
C
O
O
HO

V O(acac)
2
CH
2
N
2
, Et
2
O
1-bromo-2-butyne
(COCl)
2
, DMSO
A gNO
3
, acetone
DIBAL-H
n-BuLi, THF, -78 °C,
sec-BuLi
KHMDS, 18-C-6
MsCl, LiCl, 2,6-lutidine
DIBAL-H, CH
2
Cl
2
,
-78 °C
NaH, 18-C-6
THF/pentane
OH OH

OH
DMF, -78 °C
93 94 95
96 97
98 99 100
101
102
63
64
t
BuOOH
H
5
IO
6
S nCl
2
, NaI, DMPU
CH
2
Cl
2
, Et
3
N, -78 °C
P hCH
3
Et
3
N, CH

2
Cl
2
CBr
4
, PPh
3
(CH
2
O)
n
,-78 °C to rt
(F
3
CH
2
CO)
2
P
O
CO
2
Et
Me
THF, -78 °C
PhCH
3
n-butyllithium
-78 °C
O

O
O
3. AgNO
3
, acetone
K allolide B (104)
O
H
CO
2
CH
2
CH
2
TMS
MsCl, Et
3
N, -78 °C
Pd
2
dba
3
, PPh
3
, lutidine
103
HOCH
2
CH
2

TMS
CO, THF
1. P Ph
3
, CH
3
CN
2. TBAF, DMF, THF
21

1.2.3 Wipf Approach
Wipf developed
24
a new method for synthesising 2-alkenylfurans using Pd mediated
conditions that could tolerate various functional groups. In order to explore this strategy,
they extended their synthesis toward furanocembranoids (Scheme 1.20). The synthesis began
with TBDPS protection of 2-iodoethanol followed by alkylation with TMS-dithiane to 107.
After several attempts to cleave the dithiane moiety, Fe(NO
3
)
2
on basic alumina as a support
provided the acylsilane 108 in good yield. The lithiated propargylic chloride of 109 was
reacted with the acylsilane to afford the tertiary alcohol, which following benzoylation, gave
the benzoate. Reaction of propargylic chloride with NaI and acetone gave the iodide 110.
The other fragment, the β-ketoester 114 was prepared in 8 steps from 1,4-butanediol. Mono
MEM protection of diol 111 followed by Swern oxidation, Wittig homologation and Dibal-H
reduction, gave the unsaturated alcohol 112. The Johnson ortho ester-Claisen rearrangement
of alcohol 112 with trimethyl orthoacetate and propionic acid yielded the isopropenyl ester
113. Hydrolysis and condensation with Meldrum’s acid gave the β-ketoester 114. Alkylation

of β-ketoester with the propargylic iodide 110 produced the α-propargylic β-ketoester 115.
Pd-mediated cyclisation of β-ketoester afforded the substituted the furan 116. Silane-iodine
exchange of 116 then gave the vinyl iodide 117. Cross coupling of iodide with dimethylzinc
under Pd mediation yielded the advanced C(1)-C(18) segment 118 of lophotoxin and
pukalide.




22

















Scheme 1.1.20 Wipf’s approach towards the fragment for lophotoxin and pukalide
22
.

23

1.2.4 Pattenden Approach
The Pattenden group
18
have contributed greatly to the chemistry of the furanocembranoids.
Their strategy parallels Paquette’s synthesis, by first constructing the substituted
butyrolactone vinyl iodide and then performing a Stille macrocyclisation of an organotin
derived furan (Scheme 1.21). The synthesis commences with ring opening of (R)-
epichlorohydrin 119 with lithiated acetylide followed by TMS deprotection to give
chloropentynol 120. Carbometalation and iodination of pentynol gave the vinyl iodide 121.
Subsequent chloride displacement by neighbouring -OH group using NaOH gave an epoxide.
Lithiated ethoxyacetylene was reacted with the epoxide, followed by PTSA catalysed
cyclisation, affording the enantiopure butyrolactone. The lithium enolate of butyrolactone
was then selenated using PhSeBr to give the α-phenylseleno lactone 123. The stannylfuran
substituted aldehyde building block was prepared from Evan’s chiral oxazolidinone.
Treatment of the imide 124 with NaHMDS gave the allylic deprotonated sodium enolate;
following addition of ethyl 2-bromomethyl-3-furoate 125, this gave the deconjugative
alkylated product. Super hydride reduction of the imide resulted in reductive cleavage of
Evan’s auxiliary to afford the alcohol 126. Conversion of the primary alcohol into its tosylate
followed by ester reduction gave 127. Subsequent cyanation and TBS protection then gave
nitrile 128 in four straightforward steps. Dibal-H reduction of nitrile followed by NaBH
4

reduction afforded the alcohol. Lithiation of the furan was achieved using excess of n-BuLi
and treatment with Me
3
SnCl gave the desired C5 substituted organotin compound. Finally,
perruthenate oxidation of the alcohol gave the aldehyde 129 for coupling. Next, the lithiated
selenolactone was reacted with aldehyde 129, followed by selenoxide elimination using

H
2
O
2
/Py/CH
2
Cl
2
, to give the butenolide 65.
24















Scheme 1. 21 Pattenden’s synthesis of deoxylophotoxin
16
.
25

Arsenine mediated Stille macrocyclisation yielded the furanocembranoid 130 as mixture of
epimeric alcohol. Acetylation followed by TBS deprotection and subsequent DMP oxidation
gave the bis-deoxylophotoxin 131 with major α isomer (α:β / 2:1).
1.2.4 Trauner Approach
Trauner reported the first total synthesis of racemic bipinnatin J
14b
without protecting groups
in 9 steps from commercially available materials (Scheme 1.22). With their asymmetric total
synthesis of bipinnatin J, they next unveiled the synthesis of intricarene 23 through a
biosynthetic proposal
14a
(c.f. scheme 1.4). Furthermore, bipinnatin J was converted into
rubifolide through deoxygenation and (+)-isoephilophodione using m-CPBA. Their synthesis
began with DMP oxidation of vinyl iodide 132 followed by lithium acetylide addition to yield
a racemic alkynol. Oxidation and (S)-Alpine-borane asymmetric reduction gave the
enantiopure propargylic alcohol 133 in good yield. TMS-acetylene deprotection and TES-O-
protection of 133, followed by treatment with ethylchloroformate and TES deprotection gave
the propargylic alcohol 134. Trost is enyne reaction with allyl alcohol gave an aldehyde,
which was homologated a Wittig reaction to give the unsaturated aldehyde 135.
Chemoselective reduction of 135 produced an allyl alcohol; following Stille coupling of the
vinyl iodide 136 with the furan moiety 137, and all 20 carbons of the target molecule were in
place. After conversion of the alcohol to bromide using NBS and PPh
3
, Nozaki-Hiyama-
Kishi conditions were applied to give 26. Bipinnatin J 26 was formed as a single
diastereomer, confirming the conformational rigidity of the 13 membered ring, which
contains one cis-double bond and two 1,3-disubstituted five membered rings.