Syntheses of both the putative and revised structures of aeruginosin EI461 bearing a new bicyclic α-amino acid

Article abstract of DOI:10.1021/ol0273250

(Matrix presented) The putative structure of the naturally occurring aquatic peptide aeruginosin EI461 has been prepared from D-tyrosine. A corrected structure for aeruginosin EI461 is proposed, and the structure is proven by synthesis, which was accompli

Full text of DOI:10.1021/ol0273250

ORGANIC
LETTERS
2003
Vol. 5, No. 4
447-450
Syntheses of Both the Putative and
Revised Structures of Aeruginosin EI461
Bearing a New Bicyclic r-Amino Acid
,†
,†
Nativitat Valls,* Merce` Vallribera,† Shmuel Carmeli,‡ and Josep Bonjoch*
Laboratory of Organic Chemistry, Faculty of Pharmacy, UniVersity of Barcelona,
08028-Barcelona, Spain, and School of Chemistry, Raymond and BeVerly Sackler
Faculty of Exact Sciences, Tel AViV UniVersity, Ramat AViV, Tel AViV 69978, Israel
bonjoch@farmacia.far.ub.es
Received November 21, 2002
ABSTRACT
The putative structure of the naturally occurring aquatic peptide aeruginosin EI461 has been prepared from D-tyrosine. A corrected structure
for aeruginosin EI461 is proposed, and the structure is proven by synthesis, which was accomplished using the new r-amino acid (2S,-
3aR,6R,7aR)-6-hydroxy-2-carboxyoctahydroindole, prepared from L-tyrosine. Succesive couplings of the dipeptide D-Leu-3a,7a-diepi-L-Choi with
L-Hpla and NH OH and a deprotection step gave aeruginosin EI461.
4
Aeruginosins are a class of protease inhibitors isolated from
freshwater cyanobacterial blooms, which incorporate in their
peptide backbone the bicyclic R-amino acid 2-carboxy-6-
hydroxyoctahydroindole (L-Choi, Figure 1). The 14 initially
reported members of this family of natural products1 have
the configuration 2S,3aS,6R,7aS in their azabicyclic core,
involving an endo relationship with the substituent at C-2.2
Recently, the aquatic peptide aeruginosin EI461 was isolated
from the cyanobacterium Microcyctis aeruginosa and struc-
ture 1 was proposed (Figure 2).3 The aeruginosin EI461
structure attracted our attention4,5 due to the fact that the
† University of Barcelona.
‡ University of Tel Aviv.
(1) (a) Ishida, K.; Okita, H.; Matsuda, H.; Okino, T.; Murakami, M.
Tetrahedron 1999, 55, 10971-10988 and references therein. (b) Banker,
R.; Carmeli, S. Tetrahedron 1999, 55, 10835-10844.
Figure 1. Structures of the Choi nucleus in aeruginosins.
(2) For the synthesis and reassignment of aeruginosins 298-A and 298-
B, see: Valls, N.; Lo´pez-Canet, M.; Vallribera, M.; Bonjoch, J. Chem. Eur.
J. 2001, 7, 3446-3460.
(3) Ploutno, A.; Shoshan, M.; Carmeli, S. J. Nat. Prod. 2002, 65, 973-
978.
(4) For our previous work in the aeruginosin synthesis, see: (a) Valls,
N.; Lo´pez-Canet, M.; Vallribera, M.; Bonjoch, J. J. Am. Chem. Soc. 2000,
122, 11248-11249. (b) ref 2. (c) Valls, N.; Vallribera, M.; Lo´pez-Canet,
M.; Bonjoch, J. J. Org. Chem. 2002, 67, 4945-4950.
configuration of the Choi was different from that found in
all other aeruginosins previously described (Figure 1). In the
(5) For an alternative synthesis of the L-Choi and the aeruginosin 298-
A, see: Wipf, P.; Methot, L.-L. Org. Lett. 2000, 2, 4213-4216.
10.1021/ol0273250 CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/18/2003

derivative of the Choi 4, after removal of the N-Boc group,
was coupled with Boc-D-Leu to give in excellent yield the
dipeptide 5, which, in turn, was assembled with the L-Hpla
fragment.4c The resulting peptide 6, after simultaneous
saponification of the three esters, was coupled with am-
monium hydroxide to give, after a last debenzylation step,
compound 1, [R]D +15.9 (c 0.5 MeOH), which has the
proposed structure for aeruginosin EI461.
Figure 2. Originally assigned (1) and corrected (2) structures of
aeruginosin EI461.
We are sure that a better yield can be achieved in the last
coupling en route to 1, but our attention was diverted by the
fact that the NMR spectra of 1 did not match those of
aeruginosin EI461 (See Supporting Information). Most
notably, in DMSO-d6 the two methine protons H-2 and H-7a
of the Choi unit both resonate at δ 4.23, whereas those of
aeruginosin EI461 appear at δ 4.71 and 3.94, respectively.
Also, the H-2 of D-Leu resonates at δ 4.57 in 1, whereas it
appears at δ 4.14 in the natural product. Moreover, significant
differences in the 13C NMR spectra of 1 and aeruginosin
EI461 were observed for the Choi carbons (C-3, C-3a, C-7),
which resonate in 1 at δ 30.6, 36.2, and 39.2, respectively,
whereas those in the natural compound appear at δ 33.3,
32.6, and 36.1, respectively.
The multiplicity (doublet) of H-2 in the 1H NMR spectrum
of aeruginosin EI461, which differs from that of 1 (triplet),
would be explained by structure 2 (Figure 2).8 Structure 2
is also more consistent with the structures of previously
isolated aeruginosins, which always have the configuration
S at C-2.
Building on this point of view, we synthesized 2 to confirm
the new structural assignment. First we prepared the new
bicyclic R-amino acid 3a,7a-diepi-L-Choi (10). Starting from
â-amino ketone 79 to prepare the hydroxycarbamate 10, we
performed the required two-step sequence (change of the
protective group and the reduction step) in two different ways
(Scheme 2). When the debenzylative process with tert-
butoxycarbonylation in situ was carried out from 7, the
conversion to ketone 8 was achieved in only 56% yield.10
So, although the later reduction of ketone 8 was highly
stereoselective, alcohol 10 being isolated with 90% yield,
this protocol was discarded.
aeruginosin EI461, the hydroxyl group at C-6 in the new
bicyclic R-amino acid is located equatorially (NMR data),3
although this fact does not necessarily imply that the
stereocenter at C-6 was configurationally different from that
observed in the L-Choi in the other aeruginosins6 since the
octahydroindole ring is conformationally mobile.
Herein we describe the total synthesis of the proposed
structure of aeruginosin EI461, and after revealing that the
structure of the natural product has been assigned incorrectly,
we suggest an alternative structure for this natural product
and prove its correctness with a total synthesis.
As starting material for the synthesis of the proposed
aeruginosin EI461 (1) shown in Figure 2, we used D-tyrosine,
since the configuration of the Choi was 2R, and following
the protocol developed by us in the L-series,2,7 we obtained
the azabicyclic ketone 3. Reduction of this ketone with
NaBH4 gave stereoselectively alcohol 4 (Scheme 1). Interest-
ingly, the hydroxyl group required protection before the
following peptide-bond formation steps, which had not been
necessary working in the axial series.2 The O-acetylated
Scheme 1. Synthesis of Putative Aeruginosin EI461
(6) Recently, the closely related marine natural product dysinosin A,
incorporating a 5â-hydroxy-L-Choi core, has been isolated6a and synthesized.6b
(a) Carroll, A. R.; Pierens, G. K.; Fechner, G.; Leone, P. A.; Ngo, A.;
Simpson, M.; Hyde, E.; Hooper, J. N. A.; Bostro¨m, S.-L.; Musil, D.; Quinn,
R. J. J. Am. Chem. Soc. 2002, 124, 13340-13341. (b) Hanessian, S.;
Margarita, R.; Hall, A.; Johnstone, S.; Tremblay, M.; Parlanti, L. J. Am.
Chem. Soc. 2002, 124, 13342-13343.
(7) For the first generation synthesis of Choi derivatives, see: Bonjoch,
J.; Catena, J.; Isa´bal, E.; Lo´pez-Canet, M.; Valls, N. Tetrahedron:
Asymmetry 1996, 7, 1899-1902.
(8) For a discussion about the multiplicity of H-2 in Choi derivatives in
1
their H NMR spectra, see ref 2.
(9) Compound 7 was synthesized from O-methyl-L-tyrosine in three steps
(33% yield) following our procedure described in ref 2, but using a 5 N
solution of HCl in MeOH in the cyclization step to avoid the formation of
minor quantities of the trans isomer.
(10) The reason for the moderate yield in the transformation (7 f 8) is
the formation of the cyclohexanone i, due to the retro-Michael process of
the starting â-amino ketone 7, the product of which suffers hydrogenation
in the reaction medium. This unwanted process did not occur during the
formation of N-Boc derivative 3 in the endo series.
448
Org. Lett., Vol. 5, No. 4, 2003

Removal of the Boc group from Choi derivative 11,
followed by coupling to Boc-D-Leu by using pyBOP, gave
dipeptide 12, which after removal of the Boc group and
coupling with the L-Hpla derivative4c by using pyBOP gave
compound 13. From this L-Hpla-D-Leu-L-3a,7a-diepi-Choi
derivative, a two-step sequence consisting of coupling with
ammonium hydroxide by using PyBOP to give the amide
and a deprotecting debenzylation gave aeruginosin EI461
(Scheme 3).
Scheme 2. Synthesis of 3a,7a-diepi-L-Choi, a New Natural
R-Amino Acid
Compound 2 had NMR spectral data matching those
reported for the isolated aeruginosin EI461,13 which allowed
us to conclude that the stereostructure of the natural product
corresponds to that of 2. Moreover, when samples of natural
and synthetic aeruginosin EI461 (2) underwent acid hydroly-
sis and derivatization with Marfey’s reagent,14 the following
HPLC analysis revealed an identical retention time for the
diepi-L-Choi, L-Hpla, and D-Leu moieties in a coelution assay
and consequently demonstrated that both compounds were
identical.15
It is noteworthy that aeruginosin EI461 (2) in DMSO-d6
solution appears as a 2:3 mixture of trans and cis rotamers
around the 3a,7a-diepiChoi-Leu peptide bond (Figure 3) and
hence is the unique member of this family of natural products
in which the cis rotamer population is higher than that
observed for the trans rotamer.16 This fact is undoubtedly
related to the exo relationship between the carboxamido
group at C-2 and the carbocyclic ring in the Choi nucleus,
since structure 1, having an endo relationship in the Choi
As expected, when the interconversion of the N-benzyl
derivative to the N-Boc compound was carried out after the
reduction step, the unwanted opening byprocess observed
in 7 did not occur and the yield of the transformation
(9 f 10) increased up to 90%. The previous reduction of
ketone 7 was carried out with NaBH4 to diastereoselectively
afford the axial alcohol 9.11 This stereocontrol is due to the
conformation of the cyclohexanone ring in 7, which adopts
a half-boat form that favors the hydride attack on the â-face.12
With the new R-amino acid 10 prepared, final assembly
of 2 could begin (Scheme 3). It was also necessary in this
exo series to protect the equatorial hydroxyl group before
the coupling processes. We used the acetate group as the
protecting group because it could be removed simultaneously
with the other ester protecting groups during the saponifica-
tion step, before the final coupling (13 f 2).
(11) The reduction process of 7 using NaBH4 at -78 °C gave 9 (72%
yield), and its C-6 epimer was also formed in 11%.
(12) The preferred conformation of 7 was ascertained by NMR NOESY
data and coupling constant analyses, and its twisted conformation (depicted
below) was supported by molecular modeling studies (AM1 method). For
details of the structure of compound 7, see ref 2.
Scheme 3. Synthesis of Aeruginosin EI461
(13) The NOESY spectrum of 2 showed slight differences from the
reported data.3 We went over the ROESY spectrum of the natural product
and found that the correlation between H-2(mj) and H-7a(mj) does exist in
the spectrum. However, we found some correlations that contradict this
correlation: H-2(mj) with H-3proS(mj), H-7a(mj) with H-3proR(mj), and
H-7a(mj) with H-3a(mj). No correlation was found between H-2 and
H-3proR. Therefore, we assume that the correlation between H-2(mj) and
H-7a(mj), which of course was not observed in 2, is due to a small impurity
peak located at 4.72 ppm. The notation mj refers to the major rotamer
observed; see Figure 3.
(14) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591-596.
(15) Optical rotation for synthetic 2, [R]D -28.4, was very different from
the reported value for the natural aeruginosin EI461, [R]D +5. Since the
original sample of the natural product was slightly impure, its optical rotation
was again measured in Tel Aviv from a purified aeruginosin EI461 used
for Marfey’s analysis (See Supporting Information). The sample was still
not perfectly pure, and the quantity of material was low, 0.9 mg. The optical
rotation measured was -11((5) (c 0.0009, MeOH). Despite the remaining
difference and considering all experimental data, we concluded that
aeruginosin EI461 is levorotatory and that the real value of its optical rotation
is that measured from the pure synthetic aeruginosin 2.
(16) For the steric effects on the amide isomer equilibrium of prolyl
peptides, see: (a) Beausoleil, E.; Lubell, W. D. J. Am. Chem. Soc. 1996,
118, 12092-12098. (b) Halab, L.; Lubell, W. D. J. Am. Chem. Soc. 2002,
124, 2474-2484.
Org. Lett., Vol. 5, No. 4, 2003
449

enabled the absolute configuration of the bicyclic core of
aeruginosin EI461 to be established as 2S,3aR,6R,7aR and
has led to the reporting of the new natural R-amino acid
3a,7a-diepi-L-Choi.
Acknowledgment. Support of this research was provided
by the Spanish Ministry of Science and Technology (Project
2001BQU-3551). Thanks are also due to the DURSI (Cata-
lonia) for Grant 2001SGR-00083. M.V. is a recipient of a
fellowship (MEC, Spain).
Figure 3. Amide isomer equilibrium of 3a,7a-diepiChoi-Leu
peptide bond in the aeruginosin EI461 (2).
Supporting Information Available: Experimental pro-
cedures and spectroscopic and analytical data for compounds
2 and 9-13 as well as NMR data comparative tables of 1
and aeruginosin EI461 (2). This material is available free of
charge via the Internet at http://pubs.acs.org.
core (see Figure 1), in DMSO-d6 solution appears as a 10:1
mixture of trans and cis rotamers around the same peptide
bond.
In summary, the first synthesis of the revised structure of
aeruginosin EI461 has been accomplished. This work has
OL0273250
450
Org. Lett., Vol. 5, No. 4, 2003