A bicyclic cispentacin derivative as a novel reverse turn inducer in a GnRH mimetic

Article abstract of DOI:10.1021/jo025857o

A cispentacin-derived bicyclic β-amino acid (Bic) has been synthesized and incorporated into the 6-position of GnRH. The new GnRH analogue has been characterized with respect to its structure in solution and its activity and affinity toward the human GnRH receptor.

Full text of DOI:10.1021/jo025857o

A Bicyclic Cisp en ta cin Der iva tive a s a Novel Rever se Tu r n
In d u cer in a Gn RH Mim etic
Oliver Langer, Hanspeter Ka¨hlig, Karen Zierler-Gould, J an W. Bats,† and J ohann Mulzer*
Institut fu¨r Organische Chemie der Universita¨t Wien, Wa¨hringer Strasse 38, A-1090 Wien, Austria, and
Institut fu¨r Organische Chemie der Universita¨t Frankfurt, Marie-Curie-Str. 11,
D-60439 Frankfurt am Main, Germany
johann.mulzer@univie.ac.at
Received April 25, 2002
A cispentacin-derived bicyclic â-amino acid (Bic) has been synthesized and incorporated into the
6-position of GnRH. The new GnRH analogue has been characterized with respect to its structure
in solution and its activity and affinity toward the human GnRH receptor.
The synthesis and application of rigid amino acid
templates to generate peptidomimetics of various kinds
is a research area of ever-increasing volume and impor-
tance in bioorganic chemistry.1 Most of these templates
are derived from R-amino acids, in particular, proline
derivatives,2 and have been introduced into peptides in
order to induce the formation of â-turn-like geometries.3
Usually, incorporation of â-amino acids results in
enhanced stability against enzymatic cleavage, but at the
expense of lower biological activity. Therefore, consider-
ably less is known about the conformational preferences
of linear peptides containing one or more â-amino acids
as turn inducers.4 To get more insight into the potential
of such an approach, we have synthesized a cispentacin5-
like unnatural rigid bicyclic â-amino acid, called Bic (1).
F IGURE 1. X-ray crystal structure of Boc-protected â-amino
ester 1b (racemic material, enantiomer of 1b shown).
functionalities. Owing to the pseudoaxial and -equatorial
arrangement of the appendages, defined turn angles can
be obtained as demonstrated in the graphical representa-
tion of the X-ray structure of Boc-protected â-amino ester
1b (Figure 1). The distance between the ester-carbonyl-C
(the later C-terminus) and the Boc-carbamate nitrogen
(the later N-terminus) is 2.90 Å, and the dihedral angle
of the appendages around the Bic core is 39°, which is
reflected in the peptide turn structure later.
As the â-hydrogen of â-amino acids is not acidic and
the carboxyl function is in a thermodynamically favored
exo arrangement, epimerization at these centers, which
is a major concern in peptide synthesis with R-amino
acids, is not problematic for 1.
The Bic-core provides a rigid template for a variety of
different alignments of carboxyl, amino, and side-chain
Fmoc-Bic(Ketal)-OH6 (1a ) was synthesized in multi-
gram amounts from enantiomerically pure â-hydroxy
ester 27 in five steps and 65% overall yield (Scheme 1).
For the introduction of the â-amino group, a conversion
of â-hydroxy ester 2 into the corresponding nosylate,
substitution by azide, and reduction with H2 and Pd/C
was found to be the most effective and convenient
* To whom correspondence should be addressed.
† Universita¨t Frankfurt.
(1) Hanessian, S.; McNaughton-Smith, G.; Lombart, H.-G.; Lubell,
W. D. Tetrahedron 1997, 53, 12789-12854 and references therein. See
also Mulzer, J .; Schu¨lzchen, F.; Bats, J .-W. Tetrahedron 2000, 56,
4289-4298.
(2) Baures, P. W.; Ojala, W. H.; Gleason, W. B.; J ohnson, R. L. J .
Peptide Res. 1997, 50, 1-13.
(3) (a) Nagai, U.; Sato, K.; Nakamura, R.; Kato, R. Tetrahedron
1993, 47, 17, 3577-3592. (b) Mueller, R.; Revesz, L. Tetrahedron Lett.
1994, 35, 4091-4092.
(4) See, for examples: (a) Hibbs, D. E.; Hursthouse, M. B.; J ones, I.
G.; J ones, W.; Abdul Malik, K. M.; North, M. J . Org. Chem. 1998, 63,
1496-1504 and references therein. (b) J ones, I. G.; J ones, W.; North,
M. J . Org. Chem. 1998, 63, 1505-1513 and references therein.
(5) Cispentacin ) 2-aminocyclopentanecarboxylic acid, 2-ACPC.
(6) Nomenclature and abbreviations for amino acid residues
follow the recommendations of the IUPAC-IUB Commission on Bio-
chemical Nomenclature (Biochemistry 1970, 9, 3471-3479 and http://
www.chem.qmw.ac.uk/iupac/AminoAcid/).
(7) Petzold, K.; Dahl, H.; Skuballa, W.; Gottwald, M. Liebigs Ann.
Chem. 1990, 1087-1091.
10.1021/jo025857o CCC: $22.00 © 2002 American Chemical Society
Published on Web 09/11/2002
6878
J . Org. Chem. 2002, 67, 6878-6883

Cispentacin Derivative as a Reverse Turn Inducer in a GnRH Mimetic
SCHEME 1. Syn th esis of F m oc-P r otected Bic 1a a
a
Key: (a) Et3N, CH2Cl2, 4-NO2C6H4SO2Cl, 0 °C, 3 h, 95%; (b)
NaN3, DMF, rt, 8 h, 95%; (c) (i) H2, Pd/C, rt, 2 d, (ii) HCl/Et2O,
(iii) NEt3, 82%; (d) LiOH, 1,4-dioxane, H2O; (e) Na2CO3, Fmoc-Cl,
1,4-dioxane, rt, 36 h, 88%.
F IGURE 2. GnRH analogues 9a and 9b.
loop-inducing unnatural amino acid and the other parts
of the decapeptide are left unchanged. Thus, Freidinger
et al.12 reported that GnRH derivative 7 in which the
central Gly6-Leu7 amino acids had been replaced by a
γ-lactam dipeptide induced a 8.9-fold LH release in adult
ovariectomized female rats and in an in vitro pituitary
cell culture compared to native GnRH. However, no
detailed conformational or receptor affinity studies have
been communicated. A GnRH derivative 8 with a bicyclic
5,6-dipeptide mimetic was described in ref 3a. The LH
release in this case was 50% compared to native GnRH.
Again, no conformational analysis was performed. The
same is true of a number of other acyclic GnRH ana-
logues, in which the postulated turn structure has been
addressed by a more extensive exchange of amino acids.8
Nevertheless, the assumption of a â-turn conformation
for native GnRH is now generally accepted.
method, completely suppressing elimination to R,â- and
â,γ-unsaturated compounds. Amino ester 5 was saponi-
fied and Fmoc-protected under standard reaction condi-
tions. The N-protected amino ester 1b was obtained from
5 by Boc-protection with Boc2O under standard conditions
and was used for the crystal structure determination
(Figure 1; for data, see the Supporting Information).
On this basis, we have incorporated amino acid 1c into
the 6-position of native GnRH to generate novel acyclic
GnRH analogue 9a (Figure 2) with a structurally defined,
highly oriented turnlike conformation around position 6.
In particular, the presumed conformational induction
exerted by 1 with its carbocyclic framework should be
different from the usual bicyclic dipeptide turn-mimetics
(Figure 2), which have an amide bond in the backbone
mimic.
Scheme 2 shows the solid-phase synthesis of [Bic6-
(OH)]-GnRH (9a ) and [Bic6(OH)]-GnRH-OH (9b). Both
peptides were obtained by automated solid-phase peptide
synthesis using a conventional Fmoc temporary protec-
tion and acid-labile resin anchors and protecting groups.
Under the acidic, reductive cleavage conditions (TFA/Et3-
SiH/H2O), the acetal on the bicyclic amino acid is removed
and the resulting ketone is stereoselectively reduced to
the endo-alcohol (Scheme 3). This OH group was left
unprotected; it was envisaged as a potential anchoring
group for attaching additional side chains, e.g., to form
a glycopeptide.
To explore the properties of 1 for generating interesting
peptidomimetic structures, a target with a â-turn-
containing active conformation had to be selected. In our
case, the gonadotropin releasing hormone8 [GnRH (6)
luteinizing hormone-releasing hormone (LH-RH)] was
chosen. GnRH is a decapeptide amide that is produced
in the hypothalami of all mammals. Its pulsed secretion
stimulates the release of luteinizing hormone (LH) and
follicle-stimulating hormone (FSH) in the anterior pitu-
itary gland. LH and FSH are responsible for the produc-
tion of sex steroids and for the gametogenesis. Several
GnRH superagonists and antagonists are used for clinical
treatment of endocrine-based diseases such as prostate
cancer, breast cancer, and endometriosis.9 Regarding the
active conformation of 6 and its analogues, a â-turn
around the central 6-position (Gly) was postulated by
Monahan10 in the early 1970s. Conformational studies
by Momany,11 based on empirical energy calculations,
have been accepted as a confirmation of this assumption.
The most direct way, however, to check this structure
model experimentally is the synthesis of acyclic GnRH
analogues in which the 6-Gly section is replaced by a
Con for m a tion a l An a lysis. It was thought important
to perform the conformational analysis under conditions
close to those found physiologically. As amide 9a is
sparingly soluble in water, the free acid 9b was chosen
for the envisaged NMR study (5 mmol/L peptide in H2O/
D2O (9:1, v/v), pH 6.5, 283 K). The assignment and
analysis of the NMR spectra was done with the program
ANSIG13 resulting in distance constraints based upon 162
(8) Kutscher, B.; Bernd, M.; Beckers, T.; Polymeropulos, E. E.; Engel,
J . Angew. Chem. 1997, 109, 2240-2254; Angew. Chem., Int. Ed. Engl.
1997, 36, 2148-2161.
(9) Cho, N.; Harada, M.; Imaeda, T.; Imada, T.; Matsumoto, H.;
Hayase, Y.; Sakasi, S.; Furuya, S.; Suzuki, N.; Okubo, S.; Ogi, K.; Endo,
S.; Onda, H.; Fujino, M. J . Med. Chem. 1998, 41, 4190-4195.
(10) Monahan, M. W.; Amos, M. S.; Anderson, H. A.; Vale, V.
Biochemistry 1973, 12, 4616-4620.
(12) Freidinger, R. M.; Veber, D. F.; Perlow, D. S.; Brooks, J . R.;
Saperstein, R. Science 1980, 210, 656-658.
(13) Kraulis, P. J . J . Magn. Reson. 1989, 24, 627-633.
(11) Momany, F. A. J . Am. Chem. Soc. 1976, 98, 2990-2996.
J . Org. Chem, Vol. 67, No. 20, 2002 6879

Langer et al.
SCHEME 2. Syn th esis of P ep tid e Am id e 9a a n d
P ep tid e Acid 9ba
a
F IGURE 3. Superposition of 17 low energy NMR conformers
of 9b (hydrogen omitted for clarity, backbone trace through
R-carbon atoms).
Synthesis of 9a : (a) TentaGel S RAM Fmoc amide resin (X )
-NH-; 0.2 mmol/g), (i) Fmoc-deprotection with 20% piperidine
in DMF, (ii) double coupling with 2-5.0 equiv of Fmoc-Xaa-OH
or Glp-OH, using PyBOP as coupling agent and NMM as base;
Xaa ) Gly, Pro, Arg(Pbf), Leu, Bic (1a ), Tyr(tBu), Ser(tBu),
Trp(Boc), His(Trt); (b) final deprotection and cleavage from resin
with TFA/triethylsilane/H2O ) 19:1:1, 2 h, rt. Synthesis of 9b: (a)
Fmoc-Gly peptide acid resin (X ) -O-Gly-), analogous to 1a with
Xaa ) Pro, Arg(Pbf), Leu, Bic (1a ), Tyr(tBu), Ser(tBu), Trp(Boc),
His(Trt); (b) final deprotection and cleavage from resin with TFA/
triethylsilane/H2O ) 19:1:1, 2 h, rt.
SCHEME 3. Ster eoselective Ca r bon yl Red u ction
of th e Resin -Bou n d P ep tid e
F IGURE 4. NH-proton exchange rates for 9b.
NOE contacts and a few backbone dihedral constraints
based upon NH-CRH coupling constants (see the Sup-
porting Information). The program XPLOR14 was used
for structure calculations starting from a linear peptide
using a restrained molecular dynamics/refinement/energy
minimization protocol. The resulting lowest energy struc-
tures show a partially ordered C terminus, a less ordered
N terminus, and a well-defined loop region around the
newly introduced amino acid (Figure 3).
1H,15N HSQC spectra show low exchange rates for the
central (positions 5-8) region (Figure 4). This criterion
is considered typical of amino acids within a loop
conformation.16
To gain insight into the conformational behavior of
[Bic6(OH)]-GnRH (9a ) itself, a computer model was
constructed and examined using the Tinker17 software
package. Force-field parameters were taken from the
OPLS-AA force field, using the generalized Born/surface
area (GB/SA) water solvation model.18 Figure 5 shows a
graphical representation of the three lowest energy
conformers. They represent more than 99% population
of the conformational space of the peptide at room
temperature.
As the direction of the peptide chain is reversed and
no intramolecular hydrogen bond between Tyr5 and Arg8
is formed, this motive can be defined as a kind of an open
â turn. The distance between the R carbon atoms of these
two residues is about 500 pm and thus lies well within
the limit for a reverse turn (700 pm).15 The experimental
data show strong sequential CRH-NH NOEs for the 5-8
region as well as weak sequential NH-NH NOEs (see the
NOE list in the Supporting Information). In addition,
hydrogen exchange measurements using saturation trans-
fer techniques in combination with gradient-selected
Similar to the NMR study of 9b, the reverse turn
induced by Bic can be clearly seen in the computer model
of 9a . The distances from Ser-CR to Leu-CR in these three
conformers are 420, 610, and 690 pm and thus lie within
the limit of 700 pm for a reverse turn. In accordance with
(14) Nilges, M.; Kuszewski, J .; Brunger, A. T. In Computational
Aspects of the Study of Biological Macromolecules; Hoch, J . C., Ed.;
Plenum Press: New York, 1991.
(15) Ball, J . B.; Hughes, R. A.; Alewood, P. F.; Andrews, P. R.
Tetrahedron 1993, 49, 3467-3478 and literature cited therein.
(16) Creighton, T. E. In ProteinssStructures and Molecular Proper-
ties, 2nd ed.; W. H. Freeman and Co.: New York, 1993; p 283.
(17) Kong, Y.; Ponder, J . W. J . Chem. Phys. 1997, 107, 481-492.
(18) Qiu, D.; Shenkin, P. S.; Hollinger, F. P.; Still, W. C. J . Phys.
Chem. A 1997, 101, 3005-3014.
6880 J . Org. Chem., Vol. 67, No. 20, 2002

Cispentacin Derivative as a Reverse Turn Inducer in a GnRH Mimetic
F IGURE 5. Superposition of three conformers of 9a (>99%
population/300 K, hydrogens and atoms of residues 1-4
omitted for clarity, backbone trace through R-carbon atoms).
F IGURE 7. Superposition of the minimum energy conformers
of 9a (red) and 9b (blue).
(IC50 ) 1 × 10-10 mol/L). Application in concurrency with
busereline showed that 9a is not able to completely block
the receptor and thus suppress busereline-induced IP3
release. This is in accordance with its agonistic character.
Compound 9a effects Ca2+-release in CHO#3 cells in
concentrations above 10-8 mol/L. Application 4 min before
application of 10-9 mol/L busereline partly inhibits
busereline-induced Ca2+-release, which can be explained
by a partial antagonism of 9a effected by de-sensitization
of the GnRH receptor or its signal transduction system.
Con clu sion
In conclusion, we have shown that the novel â-amino
acid Bic in the form of 1c is suitable for inducing turn
structures in oligopeptides. Thus, the GnRH analogue 9a
shows the profile of an agonist with respect to the human
GnRH receptor and has an activity on a par with the
natural hormone. This result is in accordance with the
established assumption of a â-turn conformation of native
GnRH. Additionally, the conformation of 9a has been
determined in detail by computational methods and its
free acid analogue 9b has been examined by NMR
methods. Both studies independently demonstrate that
a reverse turn is induced into the peptide backbone by
the rigid amino acid 1c at position 6 and that the central
5-8-section of the peptide is conformationally relatively
well defined.
F IGURE 6. Superposition of one conformer of 9a (blue) with
Mamony’s structure of 6 (red).
the NMR results, a relatively ordered C-terminal branch
and a less ordered N terminus can be found. The
structures are more compact than the NMR-derived
conformers, though. The NMR-derived conformers are
averaged in time by the NMR time scale and hence differ
less from each other than the computationally generated
ones, which are structure candidates on a molecular
motion resolution. A superimposition of one of our
calculated low energy structures of 9a with the GnRH
model presented by Momany shows a significant overlap
in the core region (Figure 6).
Exp er im en ta l Section
Gen er a l P r oced u r es. Reaction progress was monitored by
analytical thin-layer chromatography (TLC) using silica gel
60 F254 aluminum foils. Visualization was accomplished by UV
illumination (254 nm), 0.2% ninhydrin in ethanol, anisalde-
hyde/sulfuric acid in ethanol (1:2:98), or phosphomolybdic acid/
ceric sulfate/sulfuric acid in water (2 g/1 g/10 mL/90 mL). Flash
chromatography was performed using 40-63 µm silica gel
packing unless otherwise noted. 1H NMR and 13C NMR spectra
were recorded at 250, 400, or 600 MHz. Chemical shifts (δ)
are reported as parts per million from an internal tetrameth-
ylsilane (TMS) standard in the solvent indicated or by using
A similar calculation as for 9a was also performed for
amide 9b, and it can be seen from Figure 7 that the
calculated minimum energy conformers of 9a and 9b are
almost superimposable. This result demonstrates that,
at least in the calculation, the nature of the end group is
of secondary importance.
Biologica l Activity. [Bic6(OH)]-GnRH (9a ) was tested
on the human GnRH receptor expressed in intact Chinese
hamster ovary (CHO) cells. It showed a distinct affinity
(IC50 ) 2 × 10-8 mol/L) for the human receptor (indicated
by release of IP3), which is about the same as for GnRH
(IC50 ) 1 × 10-8 mol/L), but less than for busereline19
(19) Busereline ) Glp-His-Trp-Ser-Tyr-d-Ser(tBu)-Leu-Arg-Pro-Gly-
NHEt; Glp-OH ) pyroglutamic acid.
J . Org. Chem, Vol. 67, No. 20, 2002 6881

Langer et al.
the respective residual signal at 25 °C unless otherwise noted.
Coupling constants were determined from 1H NMR. High-
pressure liquid chromatography (HPLC) was performed using
ultraviolet detection at 254 nm. For analytical reversed-phase
(RP) chromatography, a 4 mm × 32 cm C-18 column was used
with the solvent system indicated. Elemental analyses were
performed by the Mikroanalytisches Laboratorium, Institut
fu¨r Physikalische Chemie der Universita¨t Wien. ESI spectra
were recorded at the Institut fu¨r Organische Chemie der
Universita¨t Frankfurt am Main and at the Institut fu¨r
Analytische Chemie der Universita¨t Wien.
All reactions using air- or water-sensitive reagents were
conducted under an Ar atmosphere with dry solvents. Solvents
were purified as follows: ethyl acetate and hexanes were
fractionally distilled, CH2Cl2 was distilled and filtered through
alumina B (activity grade super I), DMF was distilled from
CaH2. Triethylamine and N-methylmorpholine (NMM) were
refluxed for several hours over CaH2 and then slowly distilled.
All other reagents were purchased from commercial suppliers
and used without further purification.
Automated peptide synthesis was performed on a peptide
synthesizer using commercially available amide resin columns
for peptide amides or columns preloaded with the first amino
acid for peptide acids. Synthesis was performed using the Fmoc
strategy and a 5-fold excess of protected amino acids. PyBOP
was used as coupling reagent, NMM as base. Fmoc-deprotec-
tion was achieved using 20% piperidine in DMF. Final depro-
tection and cleavage from resin was accomplished using a
cocktail of 0.35 mL of triethylsilane and 0.35 mL of distilled
water in 6.3 mL of TFA. The resulting peptide amides were
precipitated as TFA-salt using precooled tert-butyl methyl
ether and separated by centrifugation.
temperature in portions during 5 min. The dark red suspension
was stirred at room temperature for 8 h, during which time it
turned yellow with a colorless precipitate. Another 7.5 g (0.12
mol) of NaN3 was added, and the suspension was allowed to
stir until complete consumption of starting material according
to DC analysis. Water (30 mL) was added, the resulting
solution was extracted 3 × 60 mL with hexanes, and the
combined organic layers were dried over MgSO4 and concen-
trated to an amber residue. Chromatography on silica gel (20%
ethyl acetate/hexanes) afforded a viscous oil, which was
dissolved in a small amount of hexanes and cooled in a freezer
to -30 °C. Crystallization was initiated by carefully warming
the bottom of the flask in a warm water bath. Following this
procedure, azide 4 was obtained as a slightly yellowish solid
1
(68.1 g, 95%): [R]20 ) -62.1 (c ) 3.3, CHCl3). H NMR (250
D
MHz, CDCl3): δ 4.29 (ψtd, J ) 3.8, 0.8 Hz, 1H), 3.75 (s, 3H),
3.47 (s, 4H), 2.98 (ψquint d, J ) 9.3, 3.5, 1H), 2.91 (dd, J )
8.9, 4.3 Hz, 1H), 2.81 (ψquint d, J ) 9.2, 4.1 Hz, 1H), 2.18
(ddd, J ) 13.4, 8.4, 1.5 Hz, 1H), 2.10-1.90 (m, 3H), 1.74 (ddd,
J ) 13.5, 8.9, 4.5 Hz, 1H), 0.98 (s, 3H), 0.96 (s, 3H). 13C NMR
(62.5 MHz, CDCl3): δ 172.6, 110.2, 72.9, 71.9, 67.0, 56.0, 52.0,
40.7, 39.8, 39.1, 38.5, 38.1, 30.5, 22.9. IR (thin film): 2954 s,
2865 m, 2110 vs, 1740 vs 1436 m, 1330 m, 1264 m, 1106 vs,
1019 m, 881 m, 739 m, 610 vs, 513 m cm-1. MS (EI, 70 eV, 60
°C): m/z 309 (0.8), 281 (2.3), 267 (18.2), 238 (8.6), 213 (8.2),
195 (2.5), 180 (20.5), 168 (17.0), 155 (16.5), 128 (34.4), 94 (21.3),
79 (20.2), 69 (89.3), 56 (25.8). Anal. Calcd for C15H23N3O4: C,
58.24; H, 7.49; N, 13.58. Found: C, 58.46; H, 7.35; N, 13.72.
(1R,2S,3R,5S)-7,7-(2′,2′-Dim eth yltr im eth ylen ed ioxy)-3-
a m in obicyclo[3.3.0]octa n e-2-ca r boxylic Acid Meth yl Es-
ter (5). To a solution of azide 4 (30.9 g, 0.10 mol) in ethyl
acetate (900 mL) was added Pd/C (10%, 350 mg), and hydrogen
was allowed to bubble through the solution in a weak stream
for 2 days. The solution was filtered through a plug of Celite,
concentrated under vacuum, coevaporated 1 × 100 mL, 1 ×
50 mL with toluene, again dissolved in diethyl ether (500 mL),
and cooled to 0 °C. A solution of HCl in ether (110 mL, 1 M)
was added, and the resulting hydrochloride was filtered. The
solid was dissolved and partitioned between a solution of 10 g
of triethylamine in 250 mL of water and 10 g of triethylamine
in 250 mL of diethyl ether. The phases were separated, and
the aqueous phase was again extracted 2 × 100 mL with
diethyl ether. Pooled organic phases were dried (MgSO4) and
concentrated to a yellowish oil that solidified after several
hours to an almost colorless solid 5 (23.2 g, 82%). 1H NMR
(250 MHz, CDCl3): δ 3.68 (mc, 1H), 3.66 (s, 3H), 3.42 (s, 4H),
2.89 (mc, 1H), 2.77 (ddd, J ) 16.4, 8.2, 2.7 Hz, 1H), 2.70 (dd,
J ) 7.7, 5.1, 1H), 2.13 (ddd, J ) 18.9, 9.0, 1.4, 1H), 2.07 (ddd,
J ) 18.7, 8.9, 1.3, 1H), 1.89-1.55 (m, 4H), 1.17 (s, 2H), 0.93
(s, 3H), 0.91 (s, 3H). 13C NMR (62.5 MHz, CDCl3): δ 174.2,
109.7, 72.1, 71.8, 56.8, 55.9, 51.4, 41.8, 41.1, 39.9, 38.8, 37.8,
30.0, 22.5. IR (thin film): 3383 w, 2952 vs, 2865 s, 1730 s,
1617 br, 1473 m, 1435 s, 1203 s, 1170 s, 1115 vs, 1020 s, 907
m, 882 m, 610 vs, 511 m cm-1. MS (EI, 70 eV, 60 °C): m/z 283
(13.2), 268 (8.6), 251 (7.0), 240 (3.2), 213 (15.3), 207 (24.4), 196
(54.5), 179 (84.3), 166 (19.5), 152 (58.9), 139 (21.3), 128 (32.3),
(1R,2S,3S,5S)-7,7-(2′,2′-Dim eth yltr im eth ylen ed ioxy)-3-
(4-n itr oben zylsu lfon yloxy)bicyclo[3.3.0]octa n e-2-ca r box-
ylic Acid Meth yl Ester (3). To a solution of â-hydroxy ester
2 (56.9 g, 0.20 mol) in DCM (500 mL) at 0 °C was added
triethylamine (57.0 g, 0.56 mmol, 2.8 equiv). The solution was
allowed to stir for 5 min, and then 4-nitrobenzenesulfonyl
chloride (62.1 g, 0.28 mol, 1.4 equiv) was added in portions
during 10 min. The dark reddish brown solution was allowed
to stir for 3 h at 0 °C, until complete consumption of starting
material (monitored by DC analysis). A 1:1 mixture of hexanes
and ethyl acetate (300 mL) was added, and the resulting
suspension was filtered through a plug of silica gel. The plug
was washed with hexanes/ethyl acetate (100 mL). The solvent
was removed to two-thirds under reduced pressure, and the
resulting suspension was again filtered. The filtrate was
concentrated to a brown oil, which was dissolved in warm tert-
butyl methyl ether (750 mL), and warm hexanes (500 mL)
were added. The product was allowed to crystallize overnight,
and fine yellow needles (83.3 g) were obtained. A second
crystallization from the mother liquor afforded another 5.4 g
product, altogether 88.7 g (95%) of 3. [R]20 ) -11.6 (c ) 2.2,
D
1
CHCl3). H NMR (400 MHz, CDCl3): δ 8.37 (ddd, J ) 9.2, 2.2,
2.2 Hz, 2H), 8.08 (ddd, J ) 9.1, 2.2, 2.2 Hz, 2H), 4.96 (ddd, J
) 9.4, 9.4, 6.7 Hz, 1H), 2.92 (ψt, J ) 8.8 Hz, 1H), 3.57 (s, 3H),
3.48 (s, 2H), 3.43 (s, 2H), 2.60-2.45 (m, 2H), 2.41-2.32 (m,
1H), 2.12-2.05 (m, 1H), 2.05-1.97 (m, 2H), 1.93-1.86 (m, 1H),
1.84-1.75 (m, 1H), 0.96 (s, 3H), 0.94 (s, 3H). 13C NMR (100
MHz, CDCl3): δ 173.1, 150.7, 142.4, 129.4, 124.3, 109.4, 84.4,
72.5, 71.7, 55.0, 52.1, 42.0, 39.7, 38.5, 38.3, 36.0, 30.1, 22.5.
IR (thin film): 3107 w, 2955 s, 2868 m, 1736 vs, 1608 w, 1534
vs, 1436 m, 1372 s, 1351 vs, 1313 s, 1188 vs, 1116 vs, 1094 vs,
1014 m, 964 s, 919 s, 738 vs, 685 m, 619 vs, 557 m, 512 w, 466
w cm-1. MS (EI, 70 eV, 130 °C): m/z 469 (17.4), 410 (2.1), 368
(5.5), 267 (84.8), 207 (30.7), 181 (31.8), 128 (70.8), 93 (40.6),
69 (90.4), 41 (100). Anal. Calcd for C21H27NO9S: C, 53.72; H,
5.80; N, 2.98. Found: C, 53.98; H, 5.38; N, 3.04.
96 (30.6), 81 (18.7), 69 (100.0), 56 (53.4). Anal. Calcd for C15H25
-
NO4: C, 63.58; H, 8.89; N, 4.94. Found: C, 63.52; H, 8.68; N,
4.83.
(1R,2S,3R,5S)-7,7-(2′,2′-Dim eth yltr im eth ylen ed ioxy)-3-
(9H -flu or en -9-ylm et h oxyca r b on yla m in o)-7-oxob icyclo-
[3.3.0]octa n e-2-ca r boxylic Acid (1a ). Amino ester 5 (23.24
g, 81.9 mmol) was dissolved in 1,4-dioxane (200 mL), and a
solution of LiOH (6.70 g, 160 mmol) in water (200 mL) was
added. The solution was allowed to stir at room temperature
for 2.5 h (complete consumption of starting material). The
solution was neutralized with KHSO4 (170 mL of 1 M solution
in water) and cooled to 0 °C. Na2CO3 was added to about 10%
w/v. The solution of Fmoc-Cl (32.4 g, 125 mmol) in 1,4-dioxane
(150 mL) was added under stirring within 60 min. The reaction
mixture was allowed to stir at room temperature for 36 h. The
solution was extracted with ether (2 × 300 mL), and the
(1R,2S,3R,5S)-7,7-(2′,2′-Dim eth yltr im eth ylen ed ioxy)-3-
a zid obicyclo[3.3.0]octa n e-2-ca r boxylic Acid Meth yl Es-
ter (4). To a suspension of nosylate 3 (109.2 g, 0.23 mol) in
DMF (150 mL) was added NaN3 (37.7 g, 0.58 mol) at room
6882 J . Org. Chem., Vol. 67, No. 20, 2002

Cispentacin Derivative as a Reverse Turn Inducer in a GnRH Mimetic
organic phase was discarded. The aqueous phase was neutral-
ized with 1 M KHSO4 solution and extracted with ether (2 ×
300 mL). The aqueous phase was acidified to pH 2-3 with
KHSO4 solution and again extracted with ether (2 × 300 mL).
Pooled organic phases were dried (MgSO4), concentrated, and
chromatographed twice on flash silica gel (EtOAc/hexanes )
1:1 containing 0.5% HOAc) to afford the Fmoc-protected
Double couplings of the Fmoc-protected amino acids (2 × 5
equiv) were performed with NMM as base, PyBOP as activator,
and piperidine (20% in DMF) as deprotecting agent. Histidine
was protected as trityl-, tryptophane as Boc-, arginine as Pbf-
derivative, and serine and tyrosine as tert-butyl ethers. Final
capping was performed with pyroglutamic acid/PyBOP/NMM.
Cleavage and deprotection with concomitant reduction of the
bicycle’s keto functionality to the corresponding endo-alcohol
was accomplished using a cocktail (7 mL) of TFA/water/
triethylsilane (90:5:5). The resulting peptide was precipitated
with ice cold tert.-butyl methyl ether, centrifuged and purified
by gel-permeation chromatography on Sephadex LH-20, using
MeOH/H2O ) 1:1 as eluent. The peptide was further purified
by RP-HPLC using MeOH/H2O ) 1:1 as eluent. ESI+: m/z
1292.1 ([MH]+ ) C62H86N17O14 requires 1292.6).
â-amino acid as a colorless foam (35.5 g, 88%): [R]20 ) -7.5
D
(c ) 2.0, CHCl3). 1H NMR (250 MHz, CDCl3): δ 11.5-10.5 (br
m, 1H), 7.75 (d, J ) 7.1 Hz, 2H), 7.57 (d, J ) 6.9 Hz, 2H), 7.35
(mc, 4H), 6.89 (d, J ) 7.5 Hz) and 5.53 (d, J ) 7.1 Hz, together
1 H), 4.56-4.14 (m, 4H), 3.49 (s, 2H), 3.46 (s, 2H), 3.00-2.62
(m, 3H), 2.37-1.94 (m, 3H), 1.84-1.53 (m, 3H), 0.98 (s, 3H),
0.96 (s, 3H). 13C NMR (carbamate cis/trans-rotamers, 62.5
MHz, CDCl3): δ 178.0 and 177.4, 158.1, 155.9, 143.9, 143.7,
143.6, 141.2, 127.7, 127.1, 124.9 and 124.8, 108.9, 72.1 and
71.9, 67.5 and 66.8, 54.3 and 54.0, 53.5 and 53.0, 47.0, 42.5
and 42.1, 40.4 and 40.1, 39.5 and 39.0, 38.3, 37.2 and 37.0,
30.0, 22.5. IR (thin film): 3318 br, 2955 s, 2867 m, 1711 vs,
1522 m, 1450 m, 1333 m, 1248 m, 1227 m, 1113 vs, 1004 w,
907 m, 878 m, 759 s, 741 vs, 610 vs, 514 w cm-1. MS (EI, 70
eV, 230 °C): m/z 491 (0.3), 446 (0.3), 404 (0.5), 387 (0.1), 378
(0.1), 361 (0.3), 345 (0.1), 313 (0.2), 295 (29.0), 250 (4.9), 208
(9.0), 196 (37.8), 165 (91.0), 128 (29.4), 69 (35.8), 41 (29.0).
HRMS (EI, 70 eV, 190 °C): m/z 491.2321 (C29H33O6N requires
491.2307). Anal. Calcd for C29H33O6N: C, 70.86; H, 6.76; N,
2.85. Found: C, 70.61; H, 6.68; N, 2.58.
6Bic(OH)-LH-RH-OH (9b). Synthesis of this compound
was performed analogously to the procedure above on a 0.1
mmol scale, but instead of the amide-resin, a Fmoc-glycine pre-
loaded acid resin, supplied by Abimed, was used. After
cleavage/deprotection/reduction (see 9a ), the resulting solution
of crude 9b was concentrated under reduced pressure, purified
by gel-permeation chromatography on Sephadex LH-20 (MeOH/
H2O ) 1:1) and subsequent RP-HPLC (MeOH/H2O ) 1:1).
ESI+: m/z 1293.0 ([MH]+ ) C62H85N16O15 requires 1293.6).
NMR Str u ctu r e Deter m in a tion of 9b. All measurements
of a 5 mmol solution of the peptide in 90% H2O/10% D2O at
pH 6.5 were done at 283 K on a BRUKER Avance DRX 600
spectrometer equipped with a TBI triple resonance probe and
actively shielded triple axis gradient accessory. The assign-
ment and analysis of the NMR spectra was done on Silicon
Graphics workstations with the program ANSIG13 resulting
in distance contraints based upon 162 NOE contacts and a
few backbone dihedral constraints based upon NH-CRH cou-
pling constants. The program XPLOR14 was used for structure
calculations starting from a linear peptide using a restrained
molecular dynamics/refinement/energy minimization protocol.
In an analogous way, the Boc-amino ester 1b was prepared
from 5. 1H NMR (carbamate rotamers, 250 MHz, CDCl3): δ
4.85 (br), 4.33 (br), 3.67 (s), 3.42 (m), 4.14 (m, 4H), 2.83 (m),
2.67 (m), 2.2 (m), 2.0 (m), 1.40 (s), 0.96 (s). 13C NMR (carbamate
cis/trans-rotamers, 62.5 MHz, CDCl3): δ 174.16, 155.21,
109.12, 79.29, 72.17, 71.99, 54.07, 53.59, 51.66, 42.32, 39.99,
39.12, 38.82, 37.20, 30.06, 28.34, 22.48. Anal. Calcd for
C
20H33O6N: C, 62.64; H, 8.67; N, 3.65. Found: C, 62.39; H,
8.68; N, 3.58.
Cr ysta l Da ta of 1b. A single crystal was measured on an
Enraf-Nonius CAD4 diffractometer. Three standard reflections
remeasured every 5500 s decreased 1% during data collection.
The data were rescaled with respect to the standards. An
empirical absorption correction was made on the basis of the
ψ-scans of four reflections. The relative transmission factor
ranged from 0.92 to 1.00. Equivalent reflections were averaged
(R(F)internal ) 0.016). The structure was determined by direct
methods using program SIR92. A difference Fourier synthesis
showed the positions of the hydrogen atoms. The H atoms were
refined with isotropic thermal parameters. The non-H atoms
were refined with anisotropic thermal parameters. The struc-
ture was refined on F values using weighting scheme: w(F)
) 4F2/[σ2(F2) + (0.03F2)2]. For further data, see the Supporting
Information.
Ack n ow led gm en t. This paper is dedicated to Pro-
fessor Peter Welzel on the occasion of his 65th birthday.
We thank Schering AG (Berlin) for providing compound
2, Dr. Muhn (Schering AG) and Dr. Scho¨llkopf (J ena-
pharm) for performing the bioassays, and the Fonds
der Chemischen Industrie and the Graduiertenkolleg
Chemische und biologische Synthese von Wirkstoffen
of the University of Frankfurt/Main for scholarships.
Su p p or tin g In for m a tion Ava ila ble: X-ray data for 1b
and 1H and 13C NMR spectra including NOE and dihedral
constraint lists of 9b. This material is available free of charge
via the Internet at http://pubs.acs.org.
6Bic(OH)-LH-RH (9a ). Solid-phase synthesis of 9a was
performed on a 0.1 mmol scale on 0.5 g of Abimed amide-resin.
J O025857O
J . Org. Chem, Vol. 67, No. 20, 2002 6883