Divergent Synthesis of Aeruginosins Based on a C(sp3)￡H Activation Strategy

Article abstract of DOI:10.1002/chem.201501370

A general and scalable access to the aeruginosin family of marine natural products, exhibiting potent inhibitory activity against serine proteases, is reported. This was enabled by the strategic use of two recently implemented Pd-catalyzed C(sp³)￡H activation reactions. The first method allowed us to obtain the common 2-carboxy-6-hydroxyoctahydroindole (Choi) core of the target molecules on a large scale, whereas the second method provided a rapid and divergent access to various hydroxyphenyllactic (Hpla) subunits, including halogenated ones. This unique strategy, together with an optimization of the fragment coupling sequence allowed the synthesis of four aeruginosins, that is, 98A-C and 298A from the chiral pool. Among them, aeruginosin 298A was synthesized on an unprecedentedly large scale. In addition, halogenated aeruginosins 98A and 98C were synthesized for the first time, thanks to a fine-tuning of the final hydrogenation step. Go natural! A general and scalable access to the aeruginosin family of marine natural products (see graphic), exhibiting potent inhibitory activity against serine proteases, is described. The strategic use of two different Pd-catalyzed C(sp³)￡H activation reactions led to the synthesis of aeruginosins98A-C and 298A.

Full text of DOI:10.1002/chem.201501370

DOI: 10.1002/chem.201501370
Full Paper
&
Natural Products |Hot Paper|
Divergent Synthesis of Aeruginosins Based on a C(sp3)ÀH
Activation Strategy
David Dailler, GrØgory Danoun, Benjamin Ourri, and Olivier Baudoin*[a]
Chem. Eur. J. 2015, 21, 9370 – 9379
9370
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Abstract: A general and scalable access to the aeruginosin
family of marine natural products, exhibiting potent inhibito-
ry activity against serine proteases, is reported. This was ena-
bled by the strategic use of two recently implemented Pd-
catalyzed C(sp3)ÀH activation reactions. The first method al-
lowed us to obtain the common 2-carboxy-6-hydroxyoctahy-
droindole (Choi) core of the target molecules on a large
scale, whereas the second method provided a rapid and di-
vergent access to various hydroxyphenyllactic (Hpla) subu-
nits, including halogenated ones. This unique strategy, to-
gether with an optimization of the fragment coupling se-
quence allowed the synthesis of four aeruginosins, that is,
98A–C and 298A from the chiral pool. Among them, aerugi-
nosin 298A was synthesized on an unprecedentedly large
scale. In addition, halogenated aeruginosins 98A and 98C
were synthesized for the first time, thanks to a fine-tuning of
the final hydrogenation step.
Introduction
a terminal guanidine, a hydrophobic amino acid, and a d-hy-
droxyphenyllactic (Hpla) subunit which may include halogen
(Cl, Br) atoms (1b, 1d, 1e). The structure of aeruginosin 298A
(1a) was initially elucidated by Murakami and co-workers, who
misassigned the configuration of the hydrophobic amino acid
as l-leucine.[2] Later, the first total syntheses of this molecule
were independently reported by the groups of Bonjoch[5] and
Wipf,[6] and revealed a d-configuration for the leucine moiety.
To access aeruginosins 1a–e, we first considered classical
retrosynthetic disconnections at the amide bonds, leading to
the four fragments 2–5 (Scheme 1). Different approaches have
been reported to construct the bicyclic Choi core 2 in previous
total syntheses of aeruginosins 298A (1a). The groups of Bon-
joch,[5] Wipf,[6] and Shibasaki[7] employed intramolecular Mi-
chael-type additions to build the CÀN bond of the pyrrolidine
ring from precursors that were either obtained from l-tyrosine
by a reductive[5] or an oxidative[6] process, or by catalytic asym-
metric phase-transfer alkylation of a glycine derivative.[7] These
Michael addition based strategies all furnished a mixture of
diastereoisomers, which required an additional equilibration
step to obtain the desired stereoisomer. In their synthesis of
aeruginosin 98B (1c), Trost and co-workers, employed a differ-
ent strategy, based on an intramolecular asymmetric Tsuji–
Trost reaction, which directly led to a hexahydroindole inter-
mediate possessing the required configuration.[8] In light of
these precedents,[9] we envisioned that our recently developed
intramolecular palladium(0)-catalyzed C(sp3)ÀH alkenyla-
tion[10,11] would allow access to compound 2 in a straightfor-
ward and scalable manner, thereby enabling the collective syn-
thesis of aeruginosins. The required bromocyclohexene precur-
sor 6 should be accessible from abundant l-alaninol 10 in
a few steps only. In order to access Hpla fragments 5, which
may include potentially labile halogen atoms, in a straightfor-
ward manner, a divergent approach from a common precursor
would be ideal. We envisioned that the intermolecular palla-
dium(II)-catalyzed directed CÀH arylation of a suitable deriva-
tive of d-lactic acid would fulfill such an objective. Indeed,
Daugulis and co-workers initially introduced the use of the bi-
dentate 8-aminoquinoline directing group to perform the
direct C(sp3)ÀH arylation of alkyl carboxylic acids at the b posi-
tion.[12] Since this discovery, important developments were ach-
ieved in this field, including the introduction of new directing
groups, thereby establishing it as one of the most powerful
and practical strategies to construct valuable arylated alkyl
Aeruginosins are marine natural products that have been iso-
lated from sponges and cyanobacterial water blooms, and
which include more than 20 congeners.[1] Among these, aerugi-
nosin 298A (1a), 98A–C (1b–d), and 101 (1e) have shown
potent in vitro inhibition of various serine proteases, including
thrombin and trypsin (Scheme 1, Table 1). These enzymes are
involved in a number of important physiological processes, in
particular, the blood coagulation cascade. Aeruginosins 1a–
e have been isolated from dried algae with yields in the range
of 0.01—0.05%.[2–3] This low availability, combined with their
interesting biological properties, stresses the need for an effi-
cient and modular total synthesis of these marine com-
pounds.[4] Structurally, aeruginosins display four different units
linked by three amide bonds: a 2-carboxy-6-hydroxyoctahy-
droindole (Choi) core containing a free (1a) or sulfated (1b–e)
hydroxy group, a C-terminus (l-argol or agmatine) containing
Table 1. Structures of target aeruginosins.
Compound
R1
R2
R3
R4 R5
aeruginosin 298A (1a)
H
H
H
H
À
À
aeruginosin 98A (1b)
SO3
Cl
H
aeruginosin 98B (1c)
aeruginosin 98C (1d)
aeruginosin 101 (1e)
SO3À agmatine
SO3À agmatine
d-allo-Ile
d-allo-Ile Br
d-allo-Ile Cl Cl
H
H
SO3
agmatine
[a] D. Dailler, Dr. G. Danoun, B. Ourri, Prof. Dr. O. Baudoin
UniversitØ Claude Bernard Lyon 1, CNRS UMR 5246
Institut de Chimie et Biochimie MolØculaires et SupramolØculaires
CPE Lyon, 43 Boulevard du 11 Novembre 1918
69622 Villeurbanne (France)
E-mail: olivier.baudoin@univ-lyon1.fr
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201501370.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9371
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. Retrosynthetic analysis. PG=protecting group; DG=directing group.
scaffolds.[13] This methodology has been already applied in sev-
eral total syntheses.[14,15]
supporting an SN1-type mechanism. In the case of amine 13,
we surmised that the use of the much bulkier alaninol-derived
nucleophile might allow an inversion of the diastereoselectivity
in favor of the trans-product by virtue of steric effects.
In this article, we report in detail our investigations that led
to the general and scalable total synthesis of the four aerugi-
nosin congeners 1a–d from simple chiral pool precursors.[16]
Halogenated aeruginosins at the Hpla fragment (including 1b
and 1d) had never been synthesized before. This achievement
was enabled by a combination of two key C(sp3)ÀH activation
reactions with the optimization of the fragment coupling se-
quence and protecting-group strategy.
In order to test this hypothesis, a study was conducted on
the influence of both the nucleophile and the protecting-
group (R) carried by the dibromocyclohexenol moiety (9). To
this purpose, two dibromocyclohexenols protected with a TBS
or TBDPS silyl group (9a,b) were prepared from 1,6-heptane-
dien-3-ol 15 (Table 2). Compounds 9a,b were then submitted
to nucleophilic substitution with various primary amines and
with sodium azide (Table 2). First, an inversion of diastereose-
lectivity in favor of the trans product 13 was observed when
a primary amine was used instead of the azide (Table 2, en-
Results and Discussion
Synthesis of the choi core
The construction of the Choi core through C(sp3)ÀH activation
required the synthesis of the enantiomerically pure precursor 6
possessing a trans relationship between the hydroxy and
amino groups on the cyclohexene ring. We hypothesized that
the dibromocyclopropanation of a protected cyclopent-3-enol
12 followed by thermal electrocyclic ring-opening[17] would
provide a short and easy access to this precursor despite a po-
tential diastereoselectivity issue (Scheme 2). Indeed, Banwell
and co-workers used a similar sequence to obtain TBS-protect-
ed dibromocyclohexenol 9a, which was submitted to nucleo-
philic substitution with sodium azide.[18] The cis-configured
product (14aa) was obtained with a diastereoselectivity of 9:1
starting from a 6:1 diastereoisomeric mixture of 9a, thereby
Table 2. Study of the allylic nucleophilic substitution.[a]
Entry
R
Nucleophile (Nu)
d.r.[b]
À
1
2
3
4
5
6
7
8
TBS
TBS
TBS
TBS
TBDPS
TBDPS
TBDPS
TBDPS
N3
1:8
[c]
nPrNH2
iBuNH2
3:1
3:1
6:1
1:15
7:1
[c]
[c]
iPrNH2
À
N3
[c]
nPrNH2
iBuNH2
iPrNH2
[c]
8:1
11:1
[c]
[a] TBDPS=tert-butyldiphenylsilyl; TEBAC=Et3BnNCl. [b] Ratio of 13/14
determined by 1H NMR spectroscopy of the crude reaction mixture. Rela-
tive configurations of the major diastereoisomers obtained with sodium
azide and isopropylamine were determined by X-ray analysis of ferrocene
derivatives 17 and 18 (Figure 1).[19] For other primary amines, the trans
configuration was ascribed to major diastereoisomers by analogy of their
chemical shifts in 1H NMR spectra. [c] In combination with 1.1 equiv
K2CO3.
Scheme 2. Synthetic plan for precursor 13. TBS=tert-butyldimethylsilyl.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9372
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tries 2–4 vs. 1; entries 6–8 vs. 5). Then, increasing the steric hin-
drance of the amine in the a-position to the N atom increased
the diastereoselectivity in favor of the trans product (entries 2–
4 and 6–8). In addition, the steric hindrance of the silyl protect-
ing group also had a major influence on the diastereoselectiv-
ity (entries 5–8 vs. 1–4), with the bulkier TBPDS group consis-
tently providing higher d.r. than the TBS group in favor of the
major cis or trans diastereoisomer for each considered nucleo-
phile. This study shows that both cis and trans diastereoiso-
mers of 5-aminocyclohex-3-en-1-ols can be accessed with
useful levels of diastereoselectivity by modifying the type of
nucleophile (i.e. azide vs. amine) and by introducing a suffi-
ciently bulky protecting group on the alcohol. These findings
might be useful in other syntheses of alkaloids and related
compounds (Figure 1).[20]
then protected as a trifluoroacetamide, thus leading to the
C(sp3)ÀH activation precursor 6. Despite the production of two
diastereoisomers in the nucleophilic substitution step (out of
the four possible), this synthetic pathway proved to be effi-
cient and robust, since it allowed the preparation of 8 g of bro-
mocyclohexene 6 with only one purification and 33% overall
yield for 6 steps (83% average yield per step).
The key Pd-catalyzed intramolecular C(sp3)ÀH alkenylation of
compound 6 was then examined, using previously reported
conditions as a starting point (Table 3, entry 1).[10] These condi-
Table 3. Optimization of the C(sp3)ÀH alkenylation.
Entry
[Pd]
Ligand
Base
T [8C]
Yield [%][a]
1
2
3
4
5
6
7
Pd(OAc)2
[Pd2(dba)3]
[Pd(PPh3)4]
Pd(PCy3)2
Pd(PCy3)2
Pd(PCy3)2
Pd(PCy3)2
Pd(PCy3)2
Pd(PCy3)2
PCy3·HBF4
PCy3·HBF4
Rb2CO3
Rb2CO3
Rb2CO3
Rb2CO3
Rb2CO3
Rb2CO3
Cs2CO3
K2CO3
120
120
120
120
110
100
120
120
120
45
45
32
75
70
18
40
85
50
–
–
–
–
–
–
–
8
9[b]
K2CO3
Figure 1. X-ray structures of major diastereoisomers 17 and 18 (thermal el-
lipsoids at the 50% probability level, some H atoms are omitted for clari-
ty).[19]
1
[a] NMR yield, determined by H NMR spectroscopy using trichloroethylene
as an internal standard. [b] Using 5 mol% of Pd(PCy3)2.
The application of the above trans-selective nucleophilic
substitution to TBS-protected l-alanilol 19 and TBDPS-protect-
ed dibromocyclohexenol 9b provided a 1:1 mixture of two dia-
stereoisomers 20a,b possessing the desired trans configura-
tion at the cyclohexene ring (Scheme 3). Indeed, the symmetri-
cal allyl cation, which is expected from the SN1-type mecha-
nism, should be attacked by the enantiopure nucleophile indis-
criminately on both electrophilic positions, opposite to the
OTBDPS group. Gratifyingly, diastereoisomers 20a,b could be
easily separated by column chromatography on silica gel.[21]
The amine group of the desired diastereoisomer (20a) was
tions afforded 45% of the desired cyclic compound 21. A rapid
screening of palladium sources and ligands (entries 2–4) al-
lowed the yield to be increased to 75% when engaging the
well-defined complex Pd(PCy3)2 (entry 4). A decrease of the
temperature induced a drop of the reaction efficiency (en-
tries 5–6). The nature of the base was then investigated (en-
tries 7 and 8), and K2CO3 furnished the highest yield (85%) of
cyclic product 21 (entry 8). Finally, an attempt at decreasing
the catalyst loading to 5 mol% Pd was met with little success
(entry 9).
With these new conditions in hand, the reaction was gradu-
ally and successfully scaled up, thereby allowing the bicyclic
product 21 to be obtained in approximately 70% isolated
yield in
a reproducible manner on a multigram scale
(Scheme 4). The absolute stereochemistry of this compound
was confirmed by X-ray analysis.[19]
Then, we studied the diastereoselective reduction of the tri-
substituted alkene by catalytic hydrogenation. Surprisingly,
among the homogeneous (Table 4, entries 1–2) and heteroge-
neous (entries 3–10) catalysts tested, only Rh/C led to com-
plete conversion (entries 7–10). Of note, Pearlman’s catalyst,
which we previously employed on a similar hexahydroindole
scaffold lacking the OTBDPS group,[10] proved completely ineffi-
cient in the present case (entry 3). Moreover, the pressure of
hydrogen affected the diastereoselectivity of the reaction.
Indeed, a decrease of the d.r. in favor of the desired product
was observed when the pressure was increased (entries 7–10).
Scheme 3. Synthesis of the CÀH activation precursor 6. TFAA = trifluoroacetic
anhydride.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9373
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 4. Scale-up of the C(sp3)ÀH alkenylation. [a] Thermal ellipsoids at
the 50% probability level, some H atoms are omitted for clarity.[19]
Scheme 5. Synthesis of the Choi core 2.
Table 4. Optimization of the hydrogenation of 21.
tate moiety in light of a precedent on the directed arylation of
this motif.[12b] We initially chose to install a triisopropylsilyl
ether (TIPS) protecting group on the phenol by analogy to pre-
vious syntheses of aeruginosin 298A.[6,7] With these coupling
partners, a rapid screening of various directing groups such as
the quinolinyl and methylthiophenyl groups introduced by
Daugulis and co-workers[12] and the 2-pyridinylisopropyl (PIP)
group introduced by Shi and co-workers[22] was performed.
The latter turned out to be the best choice, especially because
it allowed the preservation of the integrity of the sensitive lac-
tate stereogenic center. The corresponding CÀH arylation pre-
cursor 7 was synthesized in three steps from commercially
available methyl d-lactate 11 (Scheme 6).[23] Of note, directed
C(sp3)ÀH arylation is known to give rise to various byproducts
such as polyarylated compounds, thereby complicating purifi-
cation and up-scaling.[12–13] We thus performed the reaction op-
timization with the main objective to devise a practical, scala-
ble process (Table 5).
Entry
Catalyst
Solvent
PH
[bar]
Conv.
[%][a]
d.r.[a]
2
1
2
3
4
5
6
7
8
9
[RhCl(PPh3)3]
[Ir(cod)(C5H5N)(PCy3)]PF6
Pd(OH)2/C
PtO2
Ru/Al2O3
Ru/C
Rh/C
Rh/C
Rh/C
Rh/C
AcOEt
AcOEt
THF
MeOH
AcOEt
AcOEt
AcOEt
AcOEt
AcOEt
AcOEt
1
1
1
1
1
1
1
3
–
–
–
–
–
–
–
–
<10
<10
<10
100
100
100
100
–
>20:1
10:1
8:1
6:1
10
20
10
1
[a] Determined by H NMR spectroscopy.
Gratifyingly, a complete diastereocontrol of the hydrogenation
on the convex face of the hexahydroindole system was ach-
ieved under one atmosphere of hydrogen (entry 7), and the cis
configuration of the corresponding product 22 was confirmed
by X-ray analysis (Scheme 5).[19]
Standard conditions were first tested using Pd(OAc)2 as the
catalyst, K2CO3 as the base in t-amyl alcohol at 1108C, which
provided the desired arylated product 26a in 50% yield, to-
gether with 10% of unreacted starting material 7 (Table 5,
entry 1). The investigation of various solvents (entries 1–5) re-
vealed acetonitrile as the most effective to maintain the mass
balance of the reaction (entry 5). The amount of aryl iodide 8a
The isolation of product 22 showed a partial but inconse-
quential deprotection of the TBS group, probably due to the
residual acidity of the rhodium catalyst. Consequently, the hy-
drogenation and the acid-mediated deprotection of the TBS
group were performed in a one-pot manner, yielding 3.5 g of
alcohol 23 in 96% yield (Scheme 5). The primary alcohol was
then oxidized under Jones’ conditions. Finally, the resulting
carboxylic acid 24 was esterified and the trifluoroacetyl group
was removed under reductive conditions to give over 1 g of
amino ester 2 in good overall yield (15% for 11 steps) from al-
cohol 15.
Synthesis of the Hpla fragments
We then investigated the second key C(sp3)ÀH activation step,
relevant to the synthesis of the various Hpla fragments. First of
all, a benzyl group was chosen as protecting group of the lac-
Scheme 6. Synthesis of CÀH arylation substrate 7. HBTU=O-(benzotriazol-1-
yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate; DIPEA=diisopropyle-
thylamine.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9374
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 5. Optimization of the intermolecular C(sp3)ÀH arylation.
Entry
ArI [equiv]
Base
Solvent
Yield [%]
26a/7[a]
1
2
3
4
5
6
7
8
2
2
2
2
2
4
4
4
4
4
4
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
Li2CO3
Na2CO3
Rb2CO3
Cs2CO3
K3PO4
tert-amylOH
DMF
toluene
dioxane
MeCN
MeCN
MeCN
MeCN
MeCN
50:10
0:67
12:66
20:60
42:45
65:7
0:85
3:75
35:41
0:90
58:20
Scheme 8. Cleavage of the PIP directing group.
A, arising from the nitrosation of 26a, should be unstable at
room temperature and should rearrange to diazoester B,
which should spontaneously loose N2 to lead to the formation
of carboxylic acid 5a and propenyl pyridine 27.[25] Several tests
were performed using NOBF4 and pyridine as reagents. Two
parameters were found to be important: the temperature and
the amount of employed pyridine. Indeed, an optimal yield of
69% was obtained when the reaction was performed in pyri-
dine as the solvent at À308C and warming to room tempera-
ture (Scheme 8).
9
10
11
MeCN
MeCN
[a] NMR yields determined by 1H NMR spectroscopy using trichloroethy-
lene as an internal standard.
was then increased to four equivalents, which allowed 26a to
be obtained in 65% yield (entry 6). The countercation of the
carbonate base was then varied (entries 7–10), but K2CO3 re-
mained the most efficient base. Potassium phosphate was also
less effective (entry 11).
The corresponding carboxylic acid 5a was then submitted
to peptidic coupling with d-leucine methyl ester (4a) to pro-
vide the southern part of aeruginosin 298A (28a) in 72% yield
(Scheme 9). However, the hydrolysis of methyl ester 28a under
mild conditions led to a partial cleavage of the TIPS group. As
a result, we decided to change the phenol protecting group
for a benzyl group, which is insensitive to saponification condi-
tions.
With these optimized conditions (Table 5, entry 6), a scale-up
was performed, and 1.4 g of the desired CÀH arylated product
26a obtained in 63% yield after a single chromatographic pu-
rification (Scheme 7).[24] Moreover, the HPLC analysis of 26a on
a chiral stationary phase revealed no epimerization of the ste-
reocenter during the reaction.
Scheme 9. Attempt at synthesizing the southern part of aeruginosin 298A
from TIPS-protected lactate 5a.
Scheme 7. Scale-up of the C(sp3)ÀH arylation.
For the cleavage of the PIP directing group, Shi and co-
workers employed a three-step sequence (Scheme 8, top), in-
volving an initial nitrosation with a mixture of NaNO2/AcOH/
Ac2O.[22] Unfortunately, applying these conditions to compound
26a only led to the recovery of the starting material. Various
precursors of NO+, such as tBuONO (with or without acid) and
NOBF4, were tested, and only the latter allowed the conversion
of the starting material, albeit into an intractable mixture of
products. Consequently, we decided to modify the hydrolysis
process while keeping NOBF4 as the source of NO+. According
to literature elements, the putative N-nitrosamide intermediate
Thus, benzyl-protected iodophenol 8b was prepared and
submitted to the optimized CÀH activation conditions
(Scheme 10). Remarkably, the corresponding arylated product
26b was isolated in 78% yield on a 1.9 g scale without detect-
able racemization. The cleavage of the directing PIP group was
also performed on 1.3 g scale with an excellent yield of 92%
for compound 5b. The overall yield of this sequence was thus
better than the previous one featuring the TIPS protecting
group, which can be imputed to the higher sensitivity of the
latter to the employed reaction conditions. The same sequence
was then applied to PIP-amide 7 and halogenated iodoarenes
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9375
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 10. Final route to Hpla fragments.
Scheme 12. Synthesis of the C-termini. Cbz=carboxybenzyl; Tf=trifluoro-
methanesulfonyl; IBCF=isobutyl chloroformate; NMM=N-methylmorpho-
line; TMS=trimethylsilyl.
8c–e, relevant to the synthesis of aeruginosins 98A, 98C, and
101, thus furnishing halogenated Hpla fragments 5c–e in
good overall yields.
Hpla fragments 5b–d were then submitted to peptidic cou-
pling with the requisite amino esters 4a and 4b,[26] thus giving
rise to the various southern parts of aerugniosins 298A (28b)
and 98A–C (28c–e) (Scheme 11).[27]
group was removed in the presence of TMSOTf, to afford
1.25 g of l-argol fragment 3b in 4 steps, 56% overall yield.
Fragment coupling and completion of total syntheses
With the required fragments in hand, we investigated our first
coupling strategy, which consisted of coupling the C-termini
with N-trifluoroacetyl carboxylic acid 24 (Scheme 13). The latter
was thus reacted with C-termini 3a,b under standard peptidic
coupling conditions. The resulting fragments 36 and 37 were
then submitted to deprotection reactions for both trifluoroace-
tyl and TBDPS groups. With regard to fragment 36, various
mild conditions to cleave the trifluoroacetyl group, such as
K2CO3/MeOH and NH3/MeOH, were employed, but no conver-
sion was observed, which might be imputable to excessive
steric hindrance. Reductive conditions were also evaluated
(NaBH4 in MeOH), but only led to degradation. With regard to
fragment 37, several attempts at cleaving the TBDPS group
using fluorides or acidic conditions led to degradation or re-
Scheme 11. Synthesis of the southern parts of aeruginosins. PyBOP=(benzo-
triazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate.
Synthesis of the C-termini
In parallel, we synthesized the two C-termini required for aeru-
ginosins 1a–d (Scheme 12). For the C-terminus of aeruginosin-
s 98A–C (1b–d), an optimized two-step sequence was applied,
as described by Trost and co-workers[8] and afforded Cbz-pro-
tected agmatine 3a in quantitative overall yield.[28]
The l-argol fragment 3b, relevant to the synthesis of aerugi-
nosin 298A (1a), was prepared by a sequence starting with the
guanidinylation of tert-butoxycarbonyl (Boc)-protected l-orni-
thine 32. The carboxylic acid of the resulting protected l-argi-
nine fragment 33 was then reduced, and alcohol 34 was pro-
tected as a TBS silyl ether (compound 35). Finally, the Boc
Scheme 13. First coupling strategy.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9376
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
covery of the starting material. The same deprotection was
also attempted on the trifluoroacetyl-protected methyl ester
38, but only poor or moderate yields were obtained for the
corresponding product 39.
These experiments provided us with guidelines on the best
coupling strategy to adopt. The southern part of aeruginosins
(28) should be first coupled to the Choi core (2), in order to
avoid the problematic cleavage of the trifluoroacetyl group. In
addition, the cleavage of the TBDPS group should occur
before the installation of the sensitive C-terminus (3).
Accordingly, the ester part of fragment 28b was hydrolyzed,
and the corresponding carboxylic acid was subjected without
intermediate purification to peptidic coupling with the Choi
core 2 (Scheme 14). The resulting product 40 was obtained on
a 1.5 g scale in 76% yield as an 85:15 mixture of two rotamers,
as shown by 2D NMR spectroscopic experiments, and in ac-
cordance with previous data.[5] In contrast to the above-men-
tioned deprotection issues (Scheme 13), the TBDPS group of
40 was cleaved in high yield using Olah’s reagent (HF·pyridine).
The resulting methyl ester 41 was then hydrolyzed and cou-
pled with l-argol fragment 3b. This synthetic strategy allowed
the production of 1.9 g of advanced intermediate 42 in quanti-
tative manner from 40. Finally, acid-mediated cleavage of the
TBS group on the argol fragment and hydrogenolysis of Bn
and Cbz groups afforded 700 mg of aeruginosin 298A (1a),
which was isolated as the salt formed with trifluoroacetic acid,
as described previously.[2,5c,6,7a] Remarkably, the synthetic prod-
uct was obtained in 8.2% overall yield (86% average yield per
step) for the longest linear sequence of 17 steps starting from
alcohol 15. Both the overall yield and scale of this total synthe-
sis are unprecedented.
Following this achievement, we turned to the syntheses of
aeruginosin 98A–C (1b–d) (Scheme 15). The same sequence of
ester hydrolysis, peptidic coupling with the core Choi 2, and
cleavage of the TBDPS group was applied to the three differ-
ent southern parts 28c–e, thereby furnishing intermediates
43a–c in good overall yields. After hydrolysis of the methyl
ester of the Choi core, the agmatine fragment 3a was coupled
to the resulting carboxylic acids. The free hydroxy group was
then sulfated using the SO3·pyridine complex as previously de-
scribed (compounds 44a–c).[8] In the case of compound 44a,
the final hydrogenolysis was performed using H2 (1 bar) and
Pd(OH)2/C as above for 1a, thereby furnishing aeruginosin 98B
(1c) with 85% yield. This final deprotection step was much
Scheme 14. Synthesis of aeruginosin 298A.
Scheme 15. Syntheses of aeruginosin 98A, 98B, and 98C.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9377
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
more challenging for halogenated intermediates 44b,c, due to
a total concomitant dehalogenation under the same condi-
tions. This is a recurrent issue in total synthesis, especially
when hydrogenolysis is employed in the final steps. This prob-
lem was sometimes avoided by internal poisoning of the cata-
lyst by heteroatoms present on the targets.[29] Studies have
been conducted to avoid this overreduction,[30] but they often
lack generality. We were particularly interested in reports by
Wu and co-workers, prescribing the use of a combination of
Pd/C and ZnBr2.[31] We first performed model experiments with
Bn-protected 2-chlorophenol and 2-bromophenol. Various pa-
rameters were studied such as the ratio between ZnBr2 and
Pd/C, the pressure of H2, and the solvent. The best results
were obtained using a 1:1 ratio of ZnBr2 and Pd/C, especially in
the case of the most sensitive brominated compound, in ace-
tonitrile under 1 bar of hydrogen. This system was then ap-
plied to advanced intermediates 44b,c (Scheme 15). Chlorinat-
ed compound 44b was successfully converted into aeruginosi-
n 98A (1b) without observable dechlorination (50% yield). In
contrast, brominated substrate 44c afforded a 1:1 ratio of aer-
uginosin 98B (1c) and 98C (1d), as observed in the crude mix-
ture, and these products were separated by preparative HPLC.
Synthetic aeruginosins 98A–C showed identical spectroscopic
data to previously reported compounds.[3,8]
tions, Dr. J.-Y. Ortholand and Dr. H. Lemoine, Edelris, and Mrs.
O. Thoison, ICSN, for preparative HPLC.
Keywords: aeruginosin · CÀH activation · marine natural
products · palladium catalysis · total synthesis
[1] Leading review: K. Ersmark, J. R. Del Valle, S. Hanessian, Angew. Chem.
Int. Ed. 2008, 47, 1202–1223; Angew. Chem. 2008, 120, 1220–1242.
[2] M. Murakami, Y. Okita, H. Matsuda, T. Okino, K. Yamaguchi, Tetrahedron
Lett. 1994, 35, 3129–3132.
[3] a) M. Murakami, K. Ishida, T. Okino, Y. Okita, H. Matsuda, K. Yamaguchi,
Tetrahedron Lett. 1995, 36, 2785–2788; b) K. Ishida, Y. Okita, H. Matsuda,
T. Okino, M. Murakami, Tetrahedron 1999, 55, 10971–10988.
[4] Selected reviews on bioactive marine natural products: a) O. K. Radjasa,
Y. M. Vaske, G. Navarro, H. Vervoort, K. Tenney, R. G. Linington, P. Crews,
Bioorg. Med. Chem. 2011, 19, 6658–6674; b) W. H. Gerwick, B. S. Moore,
Chem. Biol. 2012, 19, 85–98.
[5] a) N. Valls, M. López-Canet, M. Vallribera, J. Bonjoch, J. Am. Chem. Soc.
2000, 122, 11248–11249; b) N. Valls, M. López-Canet, M. Vallribera, J.
Bonjoch, Chem. Eur. J. 2001, 7, 3446–3460; for a solid-phase synthesis
of aeruginosin 298A employing a similar strategy: c) T. Doi, Y. Hoshina,
H. Mogi, Y. Yamada, T. Takahashi, J. Comb. Chem. 2006, 8, 571–582.
[6] P. Wipf, J.-L. Methot, Org. Lett. 2000, 2, 4213–4216.
[7] a) T. Ohshima, V. Gnanadesikan, T. Shibuguchi, Y. Fukuta, T. Nemoto, M.
Shibasaki, J. Am. Chem. Soc. 2003, 125, 11206–11207; b) Y. Fukuta, T.
Ohshima, V. Gnanadesikan, T. Shibuguchi, T. Nemoto, T. Kisugi, T. Okino,
M. Shibasaki, Proc. Natl. Acad. Sci. USA 2004, 101, 5433–5438.
[8] B. M. Trost, T. Kaneko, N. G. Andersen, C. Tappertzhofen, B. Fahr, J. Am.
Chem. Soc. 2012, 134, 18944–18947.
[9] For other strategies employed to build the Choi core in the total syn-
theses of other aeruginosin congeners: a) S. Hanessian, R. Margarita, A.
Hall, S. Johnstone, M. Tremblay, L. Parlanti, J. Am. Chem. Soc. 2002, 124,
13342–13343 (dysinosin A, ring-closing metathesis); b) S. Hanessian, M.
Tremblay, J. F. W. Petersen, J. Am. Chem. Soc. 2004, 126, 6064–6071 (os-
cillarin, aza-Prins reaction); c) S. Diethelm, C. S. Schindler, E. M. Carreira,
Org. Lett. 2010, 12, 3950–3953 (microcin SF608, nucleophilic opening
of an oxabicyclic system).
[10] J. Sofack-Kreutzer, N. Martin, A. Renaudat, R. Jazzar, O. Baudoin, Angew.
Chem. Int. Ed. 2012, 51, 10399–10402; Angew. Chem. 2012, 124, 10545–
10548.
[11] Reviews: a) R. Jazzar, J. Hitce, A. Renaudat, J. Sofack-Kreutzer, O. Bau-
doin, Chem. Eur. J. 2010, 16, 2654–2672; b) O. Baudoin, Chem. Soc. Rev.
2011, 40, 4902–4911; c) H. Li, B.-J. Li, Z.-J. Shi, Catal. Sci. Technol. 2011,
1, 191–206; d) N. Dastbaravardeh, M. Christakakou, M. Haider, M.
Schnürch, Synthesis 2014, 1421–1439.
Conclusion
We have provided a general and scalable access to the aerugi-
nosin family of marine natural products possessing interesting
pharmacological properties, albeit low availability from natural
sources. To this purpose, two recently discovered Pd-catalyzed
C(sp3)ÀH activation methods were employed in a strategic
manner. Indeed, an intramolecular CÀH alkenylation reaction
enabled the large-scale synthesis of the common (Choi) hetero-
cyclic core of the target molecules, and an intermolecular di-
rected CÀH arylation furnished a rapid and divergent access to
diversely decorated Hpla fragments. The power of this strategy
was demonstrated in the synthesis of four aeruginosins, that is,
aeruginosin 298A (1a), 98B (1c), which had been synthesized
before, and for the first time aeruginosins 98A (1b) and 98C
(1d), which include a halogenated Hpla fragment. Importantly,
these syntheses can be scaled up, as demonstrated with aeru-
ginosin 298A which was obtained in unprecedentedly large
quantities (700 mg). This study should facilitate the production
of a library of analogues, and allow the conduction of more
advanced pharmacological studies.
[12] a) V. G. Zaitsev, D. Shabashov, O. Daugulis, J. Am. Chem. Soc. 2005, 127,
13154–13155; b) D. Shabashov, O. Daugulis, J. Am. Chem. Soc. 2010,
132, 3965–3972; c) E. T. Nadres, G. I. F. Santos, D. Shabashov, O. Daugu-
lis, J. Org. Chem. 2013, 78, 9689–9714; see also : d) B. V. S. Reddy, L. R.
Reddy, E. J. Corey, Org. Lett. 2006, 8, 3391–3394.
[13] For a review: G. Rouquet, N. Chatani, Angew. Chem. Int. Ed. 2013, 52,
11726–11743; Angew. Chem. 2013, 125, 11942–11959.
[14] a) Y. Feng, G. Chen, Angew. Chem. Int. Ed. 2010, 49, 958–961; Angew.
Chem. 2010, 122, 970–973; b) G. He, G. Chen, Angew. Chem. Int. Ed.
2011, 50, 5192–5196; Angew. Chem. 2011, 123, 5298–5302; c) W. R. Gu-
tekunst, P. S. Baran, J. Am. Chem. Soc. 2011, 133, 19076–19079; d) W. R.
Gutekunst, R. Gianatassio, P. S. Baran, Angew. Chem. Int. Ed. 2012, 51,
7507–7510; Angew. Chem. 2012, 124, 7625–7628; e) W. R. Gutekunst,
P. S. Baran, J. Org. Chem. 2014, 79, 2430–2452; f) C. P. Ting, T. J. Mai-
mone, Angew. Chem. Int. Ed. 2014, 53, 3115–3119; Angew. Chem. 2014,
126, 3179–3183; g) G. He, S.-Y. Zhang, W. A. Nack, R. Pearson, J. Rabb-
Lynch, G. Chen, Org. Lett. 2014, 16, 6488–6491.
Acknowledgements
Financial support was provided by RØgion Rhône-Alpes (ARC
SantØ), Fondation Pierre Potier pour le DØveloppement de la
Chimie des Substances Naturelles et ses Applications, and Insti-
tut Universitaire de France. We thank Dr. G. Pilet, UCBL, for X-
ray diffraction analysis, the group of Prof. M. Lemaire, (in par-
ticular Dr. E. MØtay and Dr. C. Guyon), UCBL, for helpful sugges-
[15] For reviews on CÀH functionalization strategies in total synthesis: a) L.
McMurray, F. O’Hara, M. J. Gaunt, Chem. Soc. Rev. 2011, 40, 1885–1898;
b) W. R. Gutekunst, P. S. Baran, Chem. Soc. Rev. 2011, 40, 1976–1991;
c) J. Yamaguchi, A. D. Yamaguchi, K. Itami, Angew. Chem. Int. Ed. 2012,
51, 8960–9009; Angew. Chem. 2012, 124, 9092–9142.
[16] Total synthesis of nonhalogenated aeruginosins 1a and 1c was report-
ed in a preliminary communication: D. Dailler, G. Danoun, O. Baudoin,
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9378
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Angew. Chem. Int. Ed. 2015, 54, 4919–4922; Angew. Chem. 2015, 127,
5001–5004.
Zhang, J.-W. Xu, F. Hu, B.-F. Shi, Chem. Commun. 2014, 50, 13924–
13927.
[17] B. Halton, J. Harvey, Synlett 2006, 1975–2000.
[25] a) E. H. White, J. Am. Chem. Soc. 1955, 77, 6011–6014; b) P. Romea, F.
Urpi, J. Vilarrasa, J. Org. Chem. 1989, 54, 3209–3211.
[26] d-allo-Isoleucine methyl ester was obtained from l-isoleucine through
a described epimerization/resolution process: H. Noda, K. Sakai, H. Mur-
akami, Tetrahedron: Asymmetry 2002, 13, 2649–2652.
[27] Synthesis of dichlorinated aeruginosin 101 employing Hpla fragment
5e could not be completed within the time frame of this study.
[28] For a highly efficient guanidinylation reagent: K. Feichtinger, C. Zapf,
H. L. Sings, M. Goodman, J. Org. Chem. 1998, 63, 3804–3805.
[29] See for instance: a) K. C. Nicolaou, H. J. Mitchell, N. F. Jain, N. Winssinger,
R. Hughes, T. Bando, Angew. Chem. Int. Ed. 1999, 38, 240–244; Angew.
Chem. 1999, 111, 253–255; b) J. Li, X. Chen, A. W. G. Burgett, P. G.
Harran, Angew. Chem. Int. Ed. 2001, 40, 2682–2685; Angew. Chem.
2001, 113, 2754–2757.
[18] L. Petit, M. G. Banwell, A. C. Willis, Org. Lett. 2011, 13, 5800–5803.
[19] CCDC-1058439(17), 1058440(18), 1058441(21), and 1029844(22) contain
the supplementary crystallographic data for this paper. These data can
be obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[20] See for instance: a) H. Ishibashi, T. S. So, T. Sato, K. Kuroda, M. Ikeda, J.
Chem. Soc. Chem. Commun. 1989, 762–764; b) H. Ishibashi, H. Nakatani,
S. Iwami, T. Sato, N. Nakamura, M. Ikeda, J. Chem. Soc. Chem. Commun.
1989, 1767–1769; c) T. H. Jepsen, A. A. Jensen, M. H. Lund, E. Glibstrup,
J. L. Kristensen, ACS Med. Chem. Lett. 2014, 5, 766–770; d) P. Wipf, S. R.
Spencer, J. Am. Chem. Soc. 2005, 127, 225–235.
[21] Attempts at performing an asymmetric Tsuji–Trost reaction to control
the site of nucleophilic attack were unsuccessful. For a review: B. M.
Trost, T. Zhang, J. D. Sieber, Chem. Sci. 2010, 1, 427–440.
[22] a) F.-J. Chen, S. Zhao, F. Hu, K. Chen, Q. Zhang, S.-Q. Zhang, B.-F. Shi,
Chem. Sci. 2013, 4, 4187–4192; b) Q. Zhang, K. Chen, W. Rao, Y. Zhang,
F.-J. Chen, B.-F. Shi, Angew. Chem. Int. Ed. 2013, 52, 13588–13592;
Angew. Chem. 2013, 125, 13833–13837.
[30] See for instance: J. Li, S. Wang, G. A. Crispino, K. Tenhuisen, A. Singh,
J. A. Grosso, Tetrahedron Lett. 2003, 44, 4041–4043.
[31] a) G. Wu, W. Tormos, J. Org. Chem. 1997, 62, 6412–6414; b) G. Wu, M.
Huang, M. Richards, M. Poirier, X. Wen, R. W. Draper, Synthesis 2003,
1657–1660.
[23] C. Dubost, B. Leroy, I. E. Markó, B. Tinant, J.-P. Declercq, J. Bryans, Tetra-
hedron 2004, 60, 7693–7704.
[24] After this work was completed, a scalable C(sp3)ÀH arylation reaction
Received: April 8, 2015
Published online on June 1, 2015
using the aminoquinoline directing group was reported: K. Chen, S.-Q.
Chem. Eur. J. 2015, 21, 9370 – 9379
www.chemeurj.org
9379
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim