Total synthesis of an atropdiastereomer of RP-66453 and determination of its absolute configuration

Article abstract of DOI:10.1002/anie.200351996

The absolute configuration (aR, S, S, S, S, S) was assigned to the natural product RP-66453 (1) after the total synthesis of its atropdiastereomer and spectroscopic studies on the two compounds. A sequence of S_NAr-based cyclo-etherification and

Full text of DOI:10.1002/anie.200351996

Communications
Natural Product Synthesis
assignment of the absolute configuration of the natural
product as aR(M),S,S,S,S,S.
Total Synthesis of an Atropdiastereomer of
RP-66453 and Determination of Its Absolute
Configuration**
Our approach is depicted retrosynthetically in Scheme 1.
In forward sense, a sequence of SNAr-based cycloetherifica-
tion[8] for the formation of the highly strained 14-membered
Mich›le Bois-Choussy, Pierre Cristau, and Jieping Zhu*
RP-66453 (1), a secondary metabolite, was isolated from an
Actinomycetes strain by Helynck and co-workers at Rhône–
Poulenc Rorer.[1] This new bicyclic compound binds very
selectively to the neurotensin receptor from guinea pigs
(IC50 = 30 mgmLÀ1). Subsequent structural modification has
led to the discovery of potent neurotensin antagonists claimed
to be useful for treating psychosis and Alzheimer's and
Parkinson's diseases.[2] Structurally, RP-66453 belongs to a
growing family of macrocycles that include the vancomycin
class of antibiotics,[3] chloropeptin,[4] and kistamine.[5] The
presence of strained macrocycles with both endo aryl–aryl
and aryl–aryl ether bonds is the common structural feature of
these complex and medicinally relevant natural products.
Scheme 1. Retrosynthetic analysis of RP-66453 (1).
Figure 1. Structure of RP-66453 (1).
B-O-C cyclophane 3 followed by an intramolecular Suzuki–
Miyaura coupling reaction[9] for the construction of the 15-
membered ring in 2 would complete the construction of the
elusive A-B-O-C bicyclic skeleton of RP-66453. To proceed
with the synthesis, we arbitrarily assigned all asymmetric
carbon centers the S configuration, keeping in mind that the
convergent approach would facilitate the incorporation of
each fragment with desired stereochemistry.
The connectivity of RP-66453 (1) was deduced from
detailed spectroscopic studies. However, the absolute config-
uration of the five asymmetric carbon centers as well as the
possible atropisomerism of the biaryl axis remained to be
solved. Intrigued by its potent bioactivity and with the hope to
determine its absolute configuration, we and Boger's group
embarked on the total synthesis of RP-66453.[6,7] From a
synthetic perspective, RP-66453 (1) is a challenging target
because of the inherent ring strain associated with its two
bridged macrocyclic ring systems. We communicate herein
the first total synthesis of all-S-configured RP-66453, which,
in combination with spectroscopic studies, also led to the
The synthesis began with commercially available (S)-dopa
and (S,S)-isoleucine (Scheme 2). Coupling of (S)-dopa methyl
ester (6) with (S,S)-N-Boc isoleucine (7) provided the
dipeptide 8 in 94% yield. Subsequent hydrolysis of the
methyl ester followed by amidation with the hydrochloride
salt of (S)-methyl 4-fluoro-3-nitrophenylalanate (9)[10]
afforded the tripeptide 5 in 78% overall yield. The size-
selective ring-forming process based on the intramolecular
SNAr reaction proceeded smoothly.[11] Thus, stirring a solution
of 5 (0.0026m) in DMSO in the presence of cesium fluoride at
room temperature gave the corresponding 14-membered
meta,para-cyclophane. Without purification, the crude reac-
tion mixture was submitted to one-pot sequential iodination
and methylation to provide the two separable atropisomers
10a and 11a in 48% overall yield (1.3:1 ratio) from the linear
tripeptide 5. Similarly, a sequence of cycloetherification/
bromination/methylation provided the two atropisomers 10b
and 11b in 62% overall yield. It is worthy of note that the
halogenation reaction must be performed before the O-
methylation, otherwise the halogenation takes place exclu-
[*] Dr. J. Zhu,Dr. M. Bois-Choussy,P. Cristau
Institut de Chimie des Substances Naturelles
CNRS
91198 Gif-sur-Yvette Cedex (France)
Fax: (+33)1-6907-7247
E-mail: zhu@icsn.cnrs-gif.fr
[**] We thank Dr. G. Helynck for helpful discussions. We are grateful to
Drs. G. Helynck and J. C. Cavalier of Aventis Pharma S.A. for an
authentic sample of RP-66453. We are grateful to J. F. Gallard for
recording the NMR spectra of our final compounds. Financial
support from CNRS is gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
4238
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200351996
Angew. Chem. Int. Ed. 2003, 42, 4238 –4241

Angewandte
Chemie
Scheme 2. Synthesis of the B-O-C ring. a) EDC,HOBt,Et 3N,CH 2Cl2,
RT,94%; b) LiOH,THF/H 2O,RT,then compound 9,EDC,HOBt,
Et3N,DMF,RT,77%; c) CsF,DMSO (0.0026 m),RT,2 h; d) NIS,DMF,
then K2CO3,MeI,48% overall yield of 10a and 11a from 5; e) NBS,
DMF,then MeI,62% overall yield of 10b and 11b from 5. Abbrevia-
tions: DMF=N,N-dimethylformamide,DMSO =dimethyl sulfoxide,
EDC=1-ethyl-(3-dimethylaminopropyl)carbodiimide hydrochloride,
HOBt=1-hydroxybenzotriazole hydrate,NBS =N-bromosuccinimide,
NIS=N-iodosuccinimide,RT =room temperature.
sively at the position para to the diaryl ether linkage.[12] Thus
judicious selection of the reaction sequence allowed us to use
the readily available and cheap (S)-dopa as starting material,
reducing significantly the length of the overall synthetic
sequence.[13]
Scheme 3. Completion of the synthesis of bicycle 2. a) [PdCl2(dppf)],
bis(pinacolato)diboron,KOAc,DMSO,80%; b) LiOH,THF/H 2O (1:1),
RT,quantitative; c) 1) 11a, 7% HCl in MeCN,RT,aqueous (KHCO
workup,then 13,EDC,HOBt,Et 3N,DMF,RT,87%; d) [PdCl 2(dppf)],
toluene/H2O (30:1),K 2CO3,90 8C,40%; e) Pd/C,H 2,MeOH; f)
)
3
NaNO2, H3PO2,Cu 2O,THF/H 2O (6:1),42%; g) AlBr ,EtSH,55%.
3
The production of two atropisomers is of no consequence
since the planar chirality will be destroyed in subsequent
synthetic operations. We nevertheless determined their
planar chirality by NOE studies[14] and continued the syn-
thesis with the atropisomerically pure compounds 10a and
11a. Removal of the N-Boc function from 10a followed by
coupling with arylboronate 13, obtained in turn from the (S)-
3-iodotyrosine derivative 12 by the method of Miyaura et
al.,[15] furnished the tetrapeptide 14 (Scheme 3). Due to the
high ring strain imposed by the B-O-C ring, closing the second
ring by means of an intramolecular Suzuki–Miyaura coupling
was expected to be difficult.[16,17] In the event, the best results
were obtained when we heated a solution of 14 in toluene/
H2O to 908C in the presence of potassium carbonate and a
catalytic amount of [PdCl2(dppf)]. Under these conditions the
desired bicycle 16 was obtained reproducibly in about 40%
yield as a single atropstereoisomer. The choice of solvent was
critical since the cyclization carried out in other solvents
(DMF, DMSO, toluene/methanol) but under otherwise iden-
tical conditions afforded a complex reaction mixture. Finally,
Abbreviation: dppf=1,1’-bis(diphenylphosphano)ferrocene.
reductive removal of the nitro function provided the fully
protected all-S-configured bicycle 17. Following the same
sequence, the atropisomer 11a was converted into 17 in a
similar overall yield. On the other hand, cyclization of
bromide 15 gave the bicycle 16 in only about 5% yield
under identical conditions. Treatment of 17 with AlBr3 in the
presence of EtSH removed all the protecting groups (two
methyl ethers, one methyl ester, and one N-Boc function) to
provide the desired bicyclic compound 2 in 55% yield.
The spectroscopic data (1H NMR, COSY) and the high-
resolution mass spectrum of bicycle 2 are in accord with its
structure, but they, as well as the optical rotation, differ from
those of the natural product (2: + 208 (c = 1, MeOH); RP-
66453 (1): À181 (c = 1, MeOH)). However, a simple experi-
ment demonstrated that 1 and 2 differ only in their axial
chirality and that the chiral carbon centers of the natural
product 1 are all S-configured as in compound 2. Indeed,
Angew. Chem. Int. Ed. 2003, 42, 4238 –4241
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4239

Communications
heating a solution of 1 in DMSO at 1508C for 3 h gave
exclusively the compound 2 without any detectable decom-
position. Conversely, compound 2 was perfectly stable at the
same temperature, indicating that compound 2 is the thermo-
dynamically more stable atropstereoisomer. Similarly, com-
pound 18, which was synthesized from the natural product by
a two-step sequence (Scheme 4) was irreversibly converted
Scheme 4. Derivatization of RP-66453 (1). a) Boc2O,MeOH,Et 3N;
b) CsF,DMF,MeI,50 8C,80%. Boc 2O=di-tert-butyl dicarbonate.
into 17 when a a solution in DMSO was heated at 1508C. The
similarity of the spectra of the pairs 1 and 18, 2 and 17
indicated that there was no major conformational perturba-
tion upon protection of the peripheral functional groups of
the bicyclic structure. These results demonstrate that axial
chirality was preserved during the deprotection of 17 to give
2. Scheme 5 summarizes the chemical transformations that
Figure 2. Stereochemistry of the natural product RP-66453 (1):
(aR,S,S,S,S,S) and the synthetic compound 2: (aS,S,S,S,S,S).
hence that of RP-66453 (1), was evidenced by the following
characteristic NOEs: H27 with H23, H25, H13, H15, H9; H13 with
H27, H28’, H25, H17, H20, H21 and H15 with H28, H27, H31, H9. This
stereochemical assignment also nicely explains the unusual
upfield shift of protons H27 (d = 3.63 ppm in [D6]DMSO) and
NH25 (d = 4.22 ppm in [D6]DMSO) of RP-66453, both of
which are located under the aromatic C ring. Although the
ROESY spectra of compound 17 did not provide useful NOE
correlations for the determination of its axial chirality, an
aS(P) configuration was assigned, since chemical experiments
indicated that 17 is an atropisomer of 18. Drieding models
showed that for atropisomer 17, the proton H27 is completely
out of the influence of aromatic ring current in contrast to that
of 18.
In conclusion, we have developed a convergent synthesis
of all S-configured diastereoisomer of RP-66453
(aS,S,S,S,S,S). The synthesis is notable for its brevity, partly
because functionalized amino acids can be used directly in our
approach. By combination of chemical evidence and NMR
data, the absolute configuration of RP-66453 (1) was deter-
mined to be (aR,S,S,S,S,S). It is interesting to note that
nature created RP-66453 with a thermodynamically less
stable atropisomer, while laboratory synthesis using intra-
molecular Suzuki–Miyaura coupling as the last ring-closure
step produced the thermodynamically more stable isomer.
Scheme 5. Evidence of atropisomerism between RP-66453 (1) and
compound 2. a) DMSO,150 8C,3 h,quantitative; b) Boc 2O,MeOH,
Et3N,then CsF,DMF,MeI,50 8C,80%; c) AlBr ,EtSH,CH 2Cl2,55%.
3
proved the atropisomerism between RP-66453 (1) and the
synthetic compound 2. The activation energy (Ea) for the
isomerization of 1 to 2 was calculated to be 16.9 kcalmolÀ1.
The 1H NMR spectra of both 1 and 2 exhibit a single set of
signals in [D6]DMSO or in CD3OD indicating that they exist
as single conformers. Based on the known absolute config-
uration of the peptide chain, the axial chirality of both
compounds 17 and 18 was deduced from NOE studies as
shown in Figure 2.[18] For compound 18, it is evident from the
ROESY spectrum that all amide bonds are trans by observing
the NOEs between NH36–H2, NH29–H31, NH25–H27. The
aR(M) configuration of the axial chirality of compound 18,
Received: May 28, 2003 [Z51996]
Keywords: atropisomerism · macrocycles · structure
.
elucidation · Suzuki–Miyaura reaction · total synthesis
4240
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org Angew. Chem. Int. Ed. 2003, 42, 4238 –4241

Angewandte
Chemie
[1] G. Helynck, C. Dubertret, D. Frechet, J. Leboul, J. Antibiot.
1998, 51, 512 – 514.
[2] F. F. Clerc, M. C. Dubroeucq, G. Helynck, J. Leboul, J. P. Martin ,
FR 2720066 (WO 9532218), 1995 [Chem. Abstr. 1996, 124,
233154x and 124, 233155y].
[3] a) D. H. Williams, B. Bardsley, Angew. Chem. 1999, 111, 1264 –
1286; Angew. Chem. Int. Ed. 1999, 38, 1173 – 1193; b) K. C.
Nicolaou, C. N. C. Boddy, S. Bräse, N. Winssinger, Angew. Chem.
1999, 111, 2230 – 2287; Angew. Chem. Int. Ed. 1999, 38, 2096 –
2152; c) J. Zhu, Expert Opin. Ther. Pat. 1999, 9, 1005 – 1019.
[4] K. Matsuzaki, H. Ikeda, T. Ogino, A. Matsumoto, H. B. Wood-
ruff, H. Tanaka, S. Omura, J. Antibiot. 1994, 47, 1173 – 1174.
[5] N. Naruse, M. Oka, M. Konishi, T. Oki, J. Antibiot. 1993, 46,
1812 – 1818.
[6] a) S. Boisnard, A.-C. Carbonnelle, J. Zhu, Org. Lett. 2001, 3,
2061 – 2064; b) S. Boisnard, J. Zhu, Tetrahedron Lett. 2002, 43,
2577 – 2580.
[7] a) P. J. Krenitsky, D. L. Boger, Tetrahedron Lett. 2002, 43, 407 –
410; b) P. J. Krenitsky, D. L. Boger, Tetrahedron Lett. 2003, 44,
4019 – 4022.
[8] a) J. Zhu, Synlett 1997, 133 – 144; b) J. S. Sawyer, Tetrahedron
2000, 56, 5045 – 5065.
[9] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457 – 2483.
[10] C. Vergne, M. Bois-Choussy, J. Ouazzani, R. Beugelmans, J. Zhu,
Tetrahedron: Asymmetry 1997, 8, 391 – 398.
[11] a) A. Bigot, J. Zhu, Tetrahedron Lett. 1998, 39, 551 – 554; b) A.
Bigot, E. Tran Huu Dau, J. Zhu, J. Org. Chem. 1999, 64, 6283 –
6296.
[12] Bromination of (S)-dopa gave exclusively the undesired (S)-2-
bromo-4,5-dihydroxyphenylalanine, see: S. Kirschbaum, H.
Waldmann, J. Org. Chem. 1998, 63, 4936 – 4946.
[13] For the synthesis of selective protected l-dopa from tyrosine, see
ref. [7] and a) D. L. Boger, D. Yohannes, J. Org. Chem. 1987, 52,
5283 – 5286; b) M. E. Jung, T. I. Lazarova, J. Org. Chem. 1997,
62, 1553 – 1555; c) C. Chen, Y.-F. Zhu, K. Wilcoxen, J. Org.
Chem. 2000, 65, 2574 – 2576.
[14] In the NOESY spectra, characteristic NOE crosspeaks between
protons Ha and Hb, between protons Ha and Hc were observed
for compounds 10 and 11, respectively.
[15] T. Ishiyama, M. Murata, N. Miyaura, J. Org. Chem. 1995, 60,
7508 – 7510.
[16] Aryl–Aryl bond formation by intramolecular Suzuki–Miyaura
coupling see: a) A. M. Elder, D. H. Rich, Org. Lett. 1999, 1,
1443 – 1446; b) W. Li, K. Burgess, Tetrahedron Lett. 1999, 40,
6527 – 6530; c) A.-C. Carbonnelle, J. Zhu, Org. Lett. 2000, 2,
3477 – 3480; d) V. Lobrꢀgat, G. Alcaraz, H. Bienaymꢀ, M.
Vautier, Chem. Commun. 2001, 817 – 818.
[17] In our hands formation of the B-O-C cyclophane followed by
macrolactamization failed to afford the complete A-B-O-C
bicycle (S. Boisnard, Ph.D. Thesis, Universitꢀ Paris-Sud, 2001),
although the 15-membered A-B ring can be prepared efficiently
by macrolactamization (refs. [6a,7b]).
[18] Due to the overlap of several key signals in the 1H NMR spectra
of RP-66453 (1) and 2 in DMSO, we performed NOE studies on
their derivatives 18 and 17, respectively.
Angew. Chem. Int. Ed. 2003, 42, 4238 –4241
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4241