I. Cheng-Sánchez et al. / Tetrahedron Letters 57 (2016) 3392–3395
3393
NH2
Peptidic
Coupling
HO
Boc
RCM
NH Me
O
a) HATU, DIPEA
H
N
(This work)
Functional
Gr oup
O
R
NH Me
O
O
Heck Reaction
(Reddy´s wor k)
NHBoc
N
H
NH2
Me
H
N
O
1
5
6
Manipulations
6
OH NH
NH HN
Me
Me
O
NHBoc
HO
OH NH
8
O
Me
O
BnO
O
7
HO
NH HN
O
Me
O
c) TFA, CH2Cl2
b) TEMPO/BAIB
OH
Me
O
O
Me
10
d) HATU, DIPEA,
3
Solomonamide A, 1: R = OH
Solomonamide B, : R = H
OH
9
OH OH
10
2
Boc
Me
O
Me
Me
NH
O
O
O
OH NH
OH NH
e) RCM
OH NH
NH HN
Me
OH NH
NH HN
Me
NH HN
Me
O
NH HN
O Me
O
O
(See text for reaction
conditions)
BnO
O
O
O
O
5
4
11
5
Scheme 1. Molecular structures of the solomonamides and retrosynthetic analysis.
Scheme 2. RCM approach to the solomonamides: first generation. Reagents and
conditions: (a) 1.5 equiv HATU, 1.0 equiv DIPEA, DMF, 25 °C, 12 h, 78%; (b) 0.5 equiv
TEMPO, 5.0 equiv BAIB, CH3CN/H2O 1/1, 25 °C, 7 h, 65%; (c) 8% TFA in CH2Cl2,
0 °C ? 25 °C, 3 h; (d) 1.0 equiv 10, 1.0 equiv HATU, 3.0 equiv DIPEA, DMF,
molecules based on an olefin metathesis reaction as the key step
for the construction of the macrocycle,9 which could potentially
satisfy the established objectives. Thus, as depicted in Scheme 1,
our analysis began with a straightforward amide disconnection
25 °C, 12 h, 92% over
2 steps; (e) see text for different reaction conditions.
BAIB = (diacetoxyiodo)benzene, DIPEA = N,N-diisopropylethylamine, HATU = N-
[(dimethylamino)-1H-1,2,3-triazolo-[4,5-b]pyridine-1-ylmethylene]-N-methyl-
methanaminium hexafluorophosphate N-oxide, TEMPO = 2,2,6,6-tetramethyl-1-
piperidinyloxy free radical.
of the L-serine residue and functional group manipulations at the
C5 and C6 positions to generate the cyclic derivative 3, which rep-
resents a common precursor for both solomonamides. Cyclopep-
tide 3, in turn, could be obtained from the acyclic diolefin 4 via a
ring-closing metathesis (RCM) process. In order to demonstrate
the viability and efficiency of the planned olefin metathesis
approach to the synthesis of the solomonamides, we decided to
commence this synthetic study with a model system represented
by the diolefin 5 (Scheme 1).
its acyclic precursor, diene 14, given the favourable reactivity
effected by the hydroxyl group at the allylic position in facilitating
the ring closing metathesis reaction of olefins14 (Scheme 3). Thus,
we designed a synthetic plan to be executed in two phases, the first
initiated with a cyclization phase, via a ring closing metathesis, and
the second, an oxidation phase of the resulting macrocyclic pro-
duct to instal the functional groups in order to obtain the final
compound. It is noteworthy to point out the advantages of this
new synthetic strategy in that utilizes simple starting materials
and avoids the preparation of the complex ADMOA residue, which
would be constructed at the later stages of the synthesis. In addi-
tion, the relatively simple macrocyclic intermediate 13 may repre-
sent an interesting scaffold that could provide access to the
generation of analogues from late stage intermediates, allowing
the divergent entry to numerous scaffolds. Encouraged by these
appealing features, we initiated the synthetic route with the prepa-
ration of the readily accessible dipeptide 18 from the known
iodonitrobenzene derivative 1515 according to the synthetic
sequence depicted in Scheme 3. Thus, the introduction of the allylic
group in 15, via a Stille reaction, was followed by the reduction of
the nitro group to produce aniline 17. Coupling of 17 with dipep-
tide 7 was carried out under the same conditions as those
described before for 8 to obtain in good yield dipeptide 18. Prior
to the synthesis of the acyclic precursor 14, we decided to check
the olefin metathesis reaction by use of the model compound 20,
which was prepared by coupling of the amine derived from the
Boc derivative 18 with commercial acid 19. Thus, when 20 was
treated with 10 mol % of HG-II catalyst (A) in refluxing dichloro-
methane in the presence of p-benzoquinone,16 the macrocycle 21
was obtained as the sole product in an excellent 79% yield. To
For the synthesis of the model compound 5, we started from the
simple aniline 6,10 which was coupled with dipeptide 711 to obtain
derivative 8 in 78% yield. The coupling with the hydroxy acid 10,
obtained from the known diol 912 by selective oxidation with
TEMPO/BAIB,13 was achieved by conventional amide bond synthe-
sis to provide the model olefin metathesis precursor 5 in excellent
yield (92% over two steps). With the dialkene 5 in hand, we pro-
ceeded with the olefin metathesis reaction utilizing the
Hoveyda–Grubbs 2nd generation catalyst (HG-II, A) in refluxing
dichloromethane. However, after 24 h the reaction failed to afford
any macrocyclic product, leading instead to decomposition and/or
polymerization, together with the recovery of some starting mate-
rial (ꢀ12%). In an effort to obtain the ring-closing metathesis pro-
duct, more forcing conditions (toluene at 65 °C or 100 °C) and other
catalysts (Grubbs 1st and 2nd generations, Hoveyda–Grubbs, 1st
generation) were used, but the results were similarly unsuccessful
in all the cases, with no detection of the formation of the desired
macrocyclic product 11 (Scheme 2). Shortly after the execution of
these synthetic studies carried out by us in this direction, Reddy
et al. published a synthetic variant, based on an intramolecular
Heck reaction, that provided a synthetic precursor closely related
to compound 36c (see Scheme 1).
In reevaluating our approach to the solomonamides, we decided
to maintain the olefin metathesis reaction as a key step for the con-
struction of the macrocyclic structure. However, in this new ret-
rosynthetic scenario, we envisaged the C4AC5 bond of the
natural product for bond disconnection, which would require the
removal of functional groups present at these positions. This ret-
rosynthetic action would then lead to epoxy alcohol 12 as a poten-
tial precursor for solomonamide A (1), which, in turn, can be traced
back to the olefin 13. At this stage, the preparation of the macro-
cyclic olefin 13 was envisioned to proceed without difficulties from
our delight, the newly formed double bond (D
4,5) of 21 was present
exclusively as the Z-isomer, as demonstrated by the coupling con-
stant J = 12.9 Hz.
With the formation of the macrocyclic system demonstrated in
an efficient manner, we then proceeded to extend this reaction to
the desired system. To this aim, olefinic acid 29 was previously
prepared from the described epoxy alcohol 22,17 according to the
methodology reported in the literature for related compounds.18