Synthesis of 1-(m-hydroxybenzyl)-substituted 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid derivatives as opioid peptide mimetics - Unexpected amide bond cleavages under mild conditions

Article abstract of DOI:10.1002/ejoc.200300175

N-Glycyl-(1R,3S)-1-(m-hydroxybenzyl)-1,2,3, 4-tetrahydroisoquinoline-3-carboxylic acid (Tic) was prepared as a Tyr-Tic dipeptide mimetic for exploration of its potential as a delta opioid receptor selective ligand. The compound was successfully obtained b

Full text of DOI:10.1002/ejoc.200300175

FULL PAPER
Synthesis of 1-(m-Hydroxybenzyl)-Substituted 1,2,3,4-Tetrahydroisoquinoline-
3-carboxylic Acid Derivatives as Opioid Peptide Mimetics ؊ Unexpected
Amide Bond Cleavages under Mild Conditions
[a]
[b]
[a]
[a]
´
Els Mannekens, Marco Crisma, Sylvia Van Cauwenberghe, and Dirk Tourwe*
Keywords: Alkylations / Asymmetric synthesis / Peptidomimetics / Receptors / Tetrahydroisoquinolines
N-Glycyl-(1R,3S)-1-(m-hydroxybenzyl)-1,2,3,4-tetrahydro- from L-Tic. In the course of the reactions, unusual peptide
isoquinoline-3-carboxylic acid (Tic) was prepared as a Tyr-
Tic dipeptide mimetic for exploration of its potential as a
delta opioid receptor selective ligand. The compound was
successfully obtained by a stereoselective synthesis starting
bond cleavages were observed under mild conditions, and
reaction mechanisms have been proposed.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
The development of peptide mimetics starting from the
structure of a bioactive peptide remains a considerable chal-
lenge. In the past decade, delta opioid receptor antagonists
have attracted much attention because of their potential to
suppress the addiction and tolerance associated with the use
of mu opioid agonists such as morphine as analgesics.[1]
Non-peptide delta antagonists such as naltrindole (NTI) as
well as peptide antagonists have been reported. In the
prototype Tyr-Tic-Phe-Phe-OH peptide antagonist se-
quence, the antagonism has been shown to be due to the
cyclic, and therefore conformationally restricted, nature of
the Tic (1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid)
residue. Several pharmacophoric models have been pro-
posed, based on comparison of the calculated low-energy
conformations of the flexible peptide with the rigid struc-
ture of NTI.[2,3,4] There is general agreement about the im-
portance of the relative orientation of three pharmacoph-
ores: the N-terminal amine, the Tyr1 phenolic ring, and the
Tic2 aromatic ring. The model for the bioactive confor-
mation of the Tyr-Tic-Phe-OH (TIP) tripeptide proposed
by Wilkes[2] shows a close proximity of the Tyr1 phenolic
ring and the saturated ring of the tetrahydroisoquinoline
ring. This observation led us to the design of N-glycyl-
1-(m-hydroxybenzyl)-1,2,3,4-tetrahydroisoquinoline-3-
carboxylic acid (2) as a mimetic for the Tyr-Tic dipeptide
1 (Figure 1).
Figure 1. Design of a new Tyr-Tic dipeptide mimetic 2
Results and Discussion
The 1-substituted (1R,3S)-stereoconfigured 1,2,3,4-tetra-
hydroisoquinoline-3-carboxylic acid derivative was stereo-
selectively prepared as depicted in Scheme 1.[5]
Scheme 1. 5 (a) tBuCOCl, Et3N (98%); (b) H2, Pd/C (99%);
(c) tBuLi, Ϫ78 °C, 3-benzyloxybenzyl bromide (80Ϫ84%)
(S)-2-Pivaloyl-Tic 4 was prepared from enantiomerically
pure (S)-Tic-OBn,[6] followed by hydrogenolysis as de-
scribed by Seebach.[5] Direct pivaloyl protection of unpro-
tected (S)-Tic to provide 4 was tried, but failed. 2-Pivaloyl-
Tic 4 was subsequently stereoselectively alkylated[5] by use
of tBuLi and m-benzyloxybenzyl bromide[7] in THF at Ϫ78
°C to give (1R,3S)-1-(m-benzyloxybenzyl)-2-pivaloyl-Tic 5
in good yields (80Ϫ84%). The reaction needed 3 days for
[a]
Laboratorium voor Organische Chemie, Vrije Universiteit
Brussel,
Pleinlaan 2, 1050 Brussel, Belgium
Institute of Biomolecular Chemistry, CNR, Via Marzolo 1,
[b]
35131 Padova, Italy
3300
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200300175
Eur. J. Org. Chem. 2003, 3300Ϫ3307

1,2,3,4-Tetrahydroisoquinoline-3-carboxylic Acid Derivatives as Opioid Peptide Mimetics
FULL PAPER
completion at Ϫ78 °C. The reaction time could be de- OMe in 57% and 25% yields, respectively. When the cleav-
creased to 1 h by addition of 2 equiv. of 1,3-dimethyl- age reaction was performed at 0 °C for 21 h, pure (1R,3S)-
3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU), but the 1-(m-benzyloxybenzyl)Tic-OMe 7 was obtained as its HCl
yield then dropped to 40Ϫ62%.
salt in 70 to 90% yield (Scheme 2). These mild cleavage con-
The stereochemical outcome of the alkylation reaction ditions are remarkable, since Seebach reported difficult piv-
was confirmed after coupling of Phe-OBn to 5 with O- alamide cleavages of 2-pivaloyl-1,2,3,4-tetrahydroisoquinol-
(benzotriazol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoro- ines.[5,8]
borate (TBTU) as coupling reagent (Scheme 2). Crystalliza-
Coupling of benzyloxycarbonyl-protected glycine (Z-Gly)
tion by vapor diffusion from ethyl acetate/petroleum ether to (1R,3S)-1-(m-benzyloxybenzyl)Tic-OMe 7 failed when
gave single crystals of 6 suitable for X-ray diffraction analy- TBTU or benzotriazol-1-yloxytris(dimethylamino)phos-
sis. The determined X-ray structure confirmed the (1R,3S) phonium hexafluorophosphate (BOP) were used as coup-
stereochemistry (Figure 2).
ling reagents, but was successful when N-ethyl-NЈ-(3-di-
methylaminopropyl)carbodiimide (EDC) in dichlorometh-
ane was used. The glycine-coupled compound 8 was pre-
pared in yields of 86 to 97% after purification by flash
column chromatography. Removal of the benzyl ether and
the benzyloxycarbonyl groups by hydrogenolysis resulted in
immediate cyclization to diketopiperazine 9.
An alternative pathway (Scheme 3), making use of the
mild simultaneous pivaloyl deprotection and benzyl ester
formation to 10 catalyzed by p-toluenesulfonic acid
(Scheme 2), followed by coupling of Boc-Gly to give 11 and
subsequent hydrogenolysis to 12, indicated that after Boc-
deprotection, lactamization to 9 by attack of the amine
onto the free carboxylic acid in 2 occurred on standing.
Scheme 2. (a) MeOH, HCl(g), 0 °C (90%); (b) Z-Gly-OH, EDC
(97%); (c) H2, Pd/C (95%); (d) Phe-OBn, TBTU, NMM (94%);
(e) BnOH, pTosOH, benzene reflux, then NaHCO3 (99%)
Removal of the pivaloyl group of (1R,3S)-1-(m-benzyl-
oxybenzyl)-2-pivaloyl-Tic 5 was unexpectedly easy. Stirring
of the compound in HCl(g)-saturated MeOH at room tem-
perature resulted in a mixture of (1R,3S)-1-(m-benzyloxy-
Scheme 3. (a) Boc-Gly-OH, EDC (60%); (b) H2, Pd/C (92%);
benzyl)Tic-OMe 7 and (1R,3S)-1-(m-hydroxybenzyl)Tic- (c) TFA, H2O (99%)
Figure 2. X-ray diffraction structure of (6) with numbering of the atoms; thermal ellipsoids are traced at the 30% probability level
Eur. J. Org. Chem. 2003, 3300Ϫ3307
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3301

´
E. Mannekens, M. Crisma, S. Van Cauwenberghe, D. Tourwe
FULL PAPER
The dipeptide mimetic 2 could not therefore be obtained
for biological testing.
We therefore focussed our attention on the tripeptide mi-
metic 15, containing an additional C-terminal phenylala-
nine. (Scheme 4). Saponification of the methyl ester in 8 re-
sulted in partial epimerization of the α-C atom indepen-
dently of the conditions used: 1.1 equiv. of KOH in EtOH/
water at room temperature for 6 days, 1.5 equiv. of KOS-
iMe3 in Et2O at room temperature for 30 min,[9] or 1.5
equiv. of nBu4NOH in THF at 0 °C for 50 min.[10] LC/MS
analysis after coupling of Phe-OBn with TBTU and NMM
in DMF showed a diastereomeric ratio of 9:1 in which the
major isomer was the (1R,3S)-stereoconfigured Z-Gly-1-
(m-benzyloxybenzyl)Tic-Phe-OBn 14. This was verified by
transformation of the epimeric mixture of 13 into the dike-
topiperazine 9. The major isomer was identical to the one
formed from ester 8 (Scheme 2). Crystallization of the mix-
ture of diastereomers obtained after the coupling of Phe-
OBn to 13 yielded 70% of the pure Z-Gly-[(1R,3S)-1-(m-
benzyloxybenzyl)Tic]-Phe-OBn 14. Subsequent hydro-
genolysis afforded the tripeptide mimetic Gly-[(1R,3S)-1-
(m-hydroxybenzyl)Tic]-Phe 15 in 45% yield after column
chromatography.
Scheme 5. (a) MeOH, HCl(g), 0 °C
Scheme 6. (a) MeOH, HCl(g), 0 °C
These observations cannot be ascribed to a twisted amide
bond involving the pivaloyl group. Indeed the X-ray data
of 6 reveal a normal geometry for the Tic-Phe amide bond
(which is cleaved) and a slightly distorted pivaloyl-Tic
amide bond with a torsion angle of Ϫ19.3°and a COϪN
˚
bond length (1.372 A) slightly longer than a typical amide
[11]
bond (1.346 A). Moreover, the 13C NMR chemical shift
˚
of the pivaloyl carbonyl in 2-pivaloyl-Tic-OH
4 is
177.8 ppm, and in the corresponding 1-(m-benzyloxyben-
zyl) substituted compound 5 it is 178.9 ppm, revealing very
little change in resonance overlap.
These observations can be explained by assuming a
mechanism involving intramolecular formation of a tetra-
hedral intermediate, as demonstrated for acylated N-
methyl-α-aminoisobutyryl residues by Goodman.[12]
(Scheme 7)
Scheme 4. (a) saponification (84Ϫ94%); (b) Phe-OBn, TBTU, N-
methylmorpholine, then crystallization (70%); (c) H2, Pd/C,
AcOH (45%)
In (1R,3S)-1-(m-benzyloxybenzyl)-2-pivaloyl-Tic 5, the
To avoid the saponification step, which causes epimeriz- pivaloyl carbonyl experiences an acid-catalyzed nucleophilic
ation, and to exploit the anticipated mild cleavage of the addition reaction by the oxygen of the 3-carboxylic acid,
pivaloyl group, a more convergent pathway starting from 6 producing an oxazolidinone intermediate. This rearranges
was proposed (Scheme 5). Treatment of 6 in HCl-saturated with cleavage of the pivaloylϪTic amide bond to a mixed
methanol, however, did not result in the selective cleavage anhydride, which is converted into the methyl ester. This
of the pivaloyl-Tic amide bond, but the Tic-Phe amide mechanism is consistent with the fact that the methyl ester
bond was rather unexpectedly cleaved to yield 16. More- 16 is stable under these reaction conditions. In the case of
over, prolonged treatment of this ester also did not result 6, it is the pivaloyl carbonyl oxygen that attacks the car-
in the removal of the pivaloyl group to give 7. To investigate bonyl carbon atom of the Tic residue, creating a cyclic ox-
the nature of these amide bond cleavages in more detail, azolinium/oxazolidinium ion intermediate that loses Phe-
other test reactions were performed. In the case of Z-Gly- OBn and that subsequently opens by attack of a solvent
(1R,3S)-1-(m-benzyloxybenzyl)Tic-Phe-OBn 14, no cleav- molecule (Scheme 8).
age of the Tic-Phe amide bond was observed under these
The target compound 15 was tested for its ability to dis-
conditions. Similarly, in Z-Gly-(1R,3S)-1-(m-benzyloxyben- place [3H]Ile5,6deltorphin II as a delta opioid selective li-
zyl)Tic-OH 13, no cleavage of the Gly-Tic amide bond was gand in rat brain homogenates, but displays very weak af-
observed, indicating a role of the pivaloyl group in both finity (IC50 ϭ 17 µ).[13] The signal transduction properties
cases. A minor influence of the 1-(m-benzyloxybenzyl) sub- of 15 were evaluated by DPDPE-receptor-activated
stituent was indicated by partial cleavage in 2-pivaloyl-Tic- [35S]GTPγS binding to cloned human delta receptors in C6
OH 4 (Scheme 6).
glioma cells, and indicated that 15 was a weak delta opioid
3302
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3300Ϫ3307

1,2,3,4-Tetrahydroisoquinoline-3-carboxylic Acid Derivatives as Opioid Peptide Mimetics
FULL PAPER
Scheme 7. Proposal for the mechanism involved in the cleavage of the pivaloylϪTic amide bond
Scheme 8. Proposal for the mechanism involved in the cleavage of the TicϪPhe amide bond
antagonist (Kb ϭ 1.3 µm).[14] The in vivo analgesic effect vivo analgesic potency is about six times less than that of
was tested in the tail withdrawal test after intrathecal injec- morphine. These data indicate that the peptide mimetic ef-
tion in rats. At a dose of 40 µg per rat, full analgesia was fectively interacts with the target receptor. However the
reached at 15 min after injection, and lasted for more than orientation of the pharmacophoric groups is not optimal.
2 h. In comparison to intrathecally administered morphine, Further modifications based on this design principle are
compound 15 is about six times less potent.[14]
currently under investigation.
Conclusion
Experimental Section
A convenient stereoselective method has been developed
for the synthesis of the tripeptide mimetic Gly-(1R,3S)-1-
(m-hydroxybenzyl)Tic-Phe-OH 15, starting from 2-pivaloyl-
-Tic 4. The key step in the pathway was the stereoselective
alkylation of 2-pivaloyl-Tic 4 to yield the trans- or (1R,3S)-
configured 1-(m-benzyloxybenzyl)-2-pivaloyl-Tic derivative
5. Confirmation of the stereoselectivity of the reaction was
obtained by the X-ray structure after coupling of Phe-OBn
to 10 to give 6. The pivalamide bond of (1R,3S)-1-(m-
benzyloxybenzyl)-2-pivaloyl-Tic 5 was cleaved under sur-
prisingly mild acidolytic conditions. In contrast, the piva-
loyl group of (1R,3S)-1-(m-benzyloxybenzyl)-2-pivaloyl-
Tic-Phe-OBn 6 was not cleaved under these mild conditions
but rather cleavage of the Tic-Phe peptide bond was ob-
served. These cleavages under unusually mild conditions
could be explained by an acid-catalyzed intramolecular nu-
cleophilic addition of a carbonyl oxygen to the adjacent car-
bonyl carbon. The tripeptide mimetic 15 was shown to have
General Methods: Unless stated otherwise, solvents and reagents
were used as supplied. Ethanol and methanol were dried by distil-
˚
lation from magnesium and stored over activated 3 A molecular
sieves (10 wt%). THF was heated at reflux with 0.1% w/v LiAlH4
for 24 h, distilled over sodium wire, and distilled again just before
use. DMF was distilled from benzene-1,2,4-tricarboxylic anhydride
and stored over activated molecular sieves. Diethyl ether and ben-
zene were shaken with CaCl2 (5Ϫ10% w/v) before being dried on
sodium wire. DMSO was purified and dried by distillation and over
activated molecular sieves (5% w/v). Triethylamine was distilled
from CaH2 and stored over KOH pellets. Dichloromethane was
heated at reflux with CaH2 (5% w/v) and distilled onto activated
molecular sieves.
Flash column chromatography was performed on Merck Kieselgel
60 (0.040Ϫ0.063 mm). TLC was performed on precoated polyethyl-
ene-backed silica plates (Merck Kieselgel 60 F254) with viewing by
UV light or by use of iodine. Mass spectra were recorded on a VG
Quattro II spectrometer (ESP ionization, cone voltage 70 V, capil-
lary voltage 3.5 kV, source temperature 80 °C). Data collection was
weak delta opioid affinity and antagonist properties. Its in done with the aid of Masslynx software. Exact mass measurements
Eur. J. Org. Chem. 2003, 3300Ϫ3307
www.eurjoc.org  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 3303

´
E. Mannekens, M. Crisma, S. Van Cauwenberghe, D. Tourwe
FULL PAPER
(EMM) were recorded al low resolution by electrospray ionization
with PEG-Na or PEG-Me as standard.[15] RP-HPLC analyses were
performed on a Spectra-Physics P-4000 analytical system. A Vydac
cooled to 0 °C. Dry HCl gas was flushed through the mixture until
the medium was saturated. The reaction mixture was stirred over-
night at 0 °C, then again some HCl gas was flushed through it for
218TP54RP column (C18, 5 µm, d ϭ 0.46 cm, l ϭ 25 cm) was used final completion and the mixture was stirred for an additional 4 h.
with a flow rate of 1 mL/min. and a gradient going from 100% The solvent was evaporated. MeOH (40 mL) was added to the resi-
water containing 0.1% TFA to a water/CH3CN mixture (80:20) due, and the solution was evaporated again. Finally, the residue was
containing 0.1% TFA in 30 min, followed by 10 min isocratic run
under these conditions. UV detection was performed at 215 nm. 1H
NMR and 13C NMR spectra were recorded at 25 °C on a Bruker
Avance DRX 250 or AMX 500 spectrometer, with the residual iso-
topic solvent signal as internal standard. Chemical shifts (δ) are
crystallized from EtOAc. Additional purification was not generally
necessary but could be carried out by flash chromatography with
cyclohexane/EtOAc (6:1 to 3:1) containing 1% Et3N as eluent.
Yields varied from 1.19 to 1.52 g or 70 to 90%, beige solid; m.p.
the compound decomposed at 240 °C. Rf (EtOAc) 0.60; Rf
(CH3CN/MeOH/water, 4:1:1) 0.86. HPLC: Rt ϭ 23.2 min. MS
(ESϩ): 388 [M ϩ Hϩ]. EMM (m/z): [M ϩ Hϩ] calcd. for
C25H26NO3, 388.1912; found, 388.1932. 1H NMR (CDCl3, free
amine): δ ϭ 2.54 (br. s, 1 H, NH), 2.89Ϫ3.10 (m, 4 H, Hβ, HβЈ,
1
given in ppm, the coupling constants (J) in Hz. H NMR signals
were assigned by 2D COSY and NOESY. Melting points were de-
termined on a Büchi B-540 melting point apparatus with a tem-
perature gradient of 2 °C/min. Optical rotations (αD) were meas-
ured on an automatic polarimeter (Optical Activity AA-5) at a C-CH2-Ar), 3.71 (s, 3 H, COOCH3), 3.97 (dd, 1 H, J1 ϭ 5.8, J2 ϭ
wavelength of 289 nm with a cell length of 10 cm, concentration is
in g/mL. In vitro δ-opioid receptor binding studies were performed
on homogenized rat (Wistar) brain cells with use of Ile5,6deltorphi-
13.7, Hα), 4.37 (dd, 1 H, J1 ϭ 3.7, J2 ϭ 10.6, N-CH), 5.06 (s, 2 H,
Ph-CH2-O), 6.87 (m, 3 H, H arom.), 7.16 (m, 5 H, H arom.),
7.20Ϫ7.46 (s, 5 H, H arom.). 13C NMR (CDCl3, the free amine):
n II as δ-selective ligand, as described before.[13] Antagonist proper- δ ϭ 31.9 (CH2), 43.1 (CH2), 51.2 (CH), 52.1 (CH3), 56.2 (CH),
ties were determined in the GTPγS assay on human δ-receptor
transfected C6 glioma cells, while the in vivo analgesia was deter-
mined in rats after intrathecal injection.[14]
70.1 (CH2, OBn), 113.3 (CH, arom.), 116.1 (CH, arom.), 122.0
(CH, arom.), 126.1 (CH, arom.), 126.6 (CH, arom.), 126.8 (CH,
arom.), 127.6 (2CH, arom.), 128.0 (CH, arom.), 128.6 (2CH,
arom.), 129.2 (CH, arom.), 129.8 (CH, arom.), 133.1 (C, arom.),
137.2 (C, arom.), 137.7 (C, arom.), 140.5 (C, arom.), 159.2 (C-O,
arom.), 173.5 (CϭO) ppm.
(1R,3S)-1-(m-Benzyloxybenzyl)-2-pivaloyl-1,2,3,4-tetrahydroiso-
quinoline-3-carboxylic Acid (5): tBuLi (1.7  solution in pentane,
7.8 mL, 11.49 mmol) was added under Ar atmosphere at Ϫ78 °C
to a solution of -2-pivaloyl- 1,2,3,4-tetrahydroisquinoline-3-car-
boxylic acid[5] (4, 1.50 g, 5.75 mmol) and DMPU (1.4 mL,
11.49 mmol) in anhydrous THF (40 mL). After the mixture had
been stirred for 1.5 h, a solution of m-benzyloxybenzyl bromide[7]
(3.18 g, 11.49 mmol, 2 equiv.) in dry THF (30 mL) was added and
the mixture was left stirring for 1.5 h at Ϫ78 °C. (If no DMPU
was added, the reaction took 3 days to complete.) The cooling was
stopped, and the reaction mixture was poured into water (50 mL)
and washed with Et2O (3 ϫ 40 mL). The water fraction was acidi-
fied to pH 2 with 1  HClaq and extracted with Et2O (3 ϫ 50 mL).
The combined Et2O fractions were then dried (MgSO4) and fil-
tered, and the solvents were evaporated. Excess DMPU could be
removed by redissolving the residue in water and extracting exten-
sively with Et2O. The compound was crystallized from EtOAc or
Et2O.
Z-Gly-(1R,3S)-1-(m-benzyloxybenzyl)Tic-OMe (8): The HCl salt of
methyl (1R,3S)-1-(m-benzyloxybenzyl)-1,2,3,4-tetrahydroisoquino-
line-3-carboxylate (7, 0.92 g, 2.18 mmol) was washed with a satu-
rated aqueous NaHCO3 solution (100 mL) and extracted with
EtOAc (3 ϫ 50 mL). The collected organic fractions were dried
(MgSO4) and filtered, and the solvents were evaporated. The resi-
due was dissolved in dichloromethane (5 mL), and Z-Gly-OH
(0.68 g, 3.27 mmol) and EDC (0.63 g, 3.27 mmol) were added. The
mixture was stirred overnight at room temp. and protected from
daylight, as described.[16] After a reaction time of 1 night, ad-
ditional Z-Gly-OH (0.09 g, 0.44 mmol) and EDC (0.08 g,
0.44 mmol) were added. The reaction was allowed to continue for
another 3 to 4 days. When all starting material 7 had been con-
sumed (TLC), the solvents were evaporated. The residue was dis-
solved in chloroform (40 mL) and extracted with HClaq (1 , 3 ϫ
40 mL), H2O (40 mL), saturated NaHCO3 solution (40 mL), and
H2O (40 mL). The organic phase was then dried and filtered, and
the solvents were evaporated. Purification was carried out by flash
chromatography, with eluents varying from cyclohexane through
cyclohexane/EtOAc (4:1) to cyclohexane/EtOAc (2:1). Yields varied
from 1.08 to 1.22 g or 86 to 97%, light yellow oil; Rf (cyclohexane/
EtOAc, 4:1) 0.06; Rf (cyclohexane/EtOAc 2:1) 0.20. HPLC: Rt ϭ
30.5 min. MS (ESϩ): 388, 579 [M ϩ Hϩ]. EMM (m/z): [M ϩ Hϩ]
calcd. for C35H35N2O6, 579.2495; found, 579.2502. 1H NMR
(CDCl3) (500 MHz): compound 8 exists as a mixture of two confor-
mational isomers in a ratio of 7:3. The NMR spectroscopic data
on the minor isomer (if resolved) are given between square brack-
ets. δ ϭ 2.41 (dd, 1 H, J1 ϭ 4.5, J2 ϭ 16.0, Hβ), 3.00 (m, 3 H, HβЈ,
C-CH2-Ar), 3.48 [3.53] (s, 3 H, COOCH3), 3.80 [4.06] (dd, 1 H,
J1 ϭ 3.4, J2 ϭ 16.0, Hα of Gly), 4.18Ϫ4.25 (m, 1 H, HαЈ of Gly),
4.51 (m, 1 H, Hα of Tic), 4.87 [4.93] (s, 2 H, O-CH2-Ph), 5.15 [5.13]
(s, 2 H, Ph-CH2-OCO), 5.44 [4.96] (dd, 1 H, J1 ϭ 3.6, J2 ϭ 7.2, N-
CH-Ar of Tic), 5.99 [5.77] (1 H, b.s., NH), 6.30Ϫ6.60 (m, 2 H, H
arom.), 6.70Ϫ6.85 (m, 2 H, H arom.), 7.00Ϫ7.40 (m, 14 H, H
Yields varied from 1.05 to 1.58 g or 40 to 62% (with DMPU, 1 h)
or 2.10 to 2.21 g or 80 to 84% (without DMPU, 3 days), white
solid; m.p. 189.2Ϫ189.8 °C. [α]D ϭ Ϫ3.5 (c ϭ 2, CH2Cl2). Rf
(EtOAc/petroleum ether, 4:1 ϩ1% AcOH) 0.43. HPLC: Rt
ϭ
28.4 min. MS (ESϩ): 122, 269, 290, 310, 440, 458 [M ϩ Hϩ]. EMM
(m/z): [M ϩ Hϩ] calcd. for C29H32NO4, 458.2331; found, 458.2345.
1H NMR ([D6]DMSO): δ ϭ 1.26 (s, 9 H, tBu), 2.26 and 2.75Ϫ3.12
(m, 4 H, C-CH2-Ar, Hβ and HβЈ), 4.88Ϫ4.95 (m, 3 H, N-CH and
O-CH2-Ph), 5.40 (m, 1 H, Hα), 6.15Ϫ6.24 (m, 2 H, H arom.),
6.76Ϫ6.83 (m, 2 H, H arom.), 7.06Ϫ7.17 (m, 4 H, H arom.), 7.33
(s, 5 H, H arom.) ppm. 13C NMR (CDCl3): δ ϭ 28.1 (3 CH3), 31.9
(CH2), 42.0 (CH2), 56.0 (CH), 57.9 (CH), 67.0 (C, tBu), 70.0 (CH2,
OBn), 112.9 (CH, arom.), 116.9 (CH, arom.), 122.5 (CH, arom.),
126.2 (CH, arom.), 127.0 (CH, arom.), 127.4 (3 CH, arom.), 127.6
(2 CH, arom.), 128.3 (3 CH, arom.), 133.4 (C, arom.), 137.0 (C,
arom.), 139.0 (C, arom.), 157.7 (C, arom.), 173.2 (CϭO, carboxylic
acid), 178.9 (CϭO, pivaloyl) ppm.
Methyl (1R,3S)-1-(m-Benzyloxybenzyl)-1,2,3,4-tetrahydroisoquino-
line-3-carboxylate Hydrochloride (7): A solution of (1R,3S)-1- arom.) ppm. 13C NMR (CDCl3): δ ϭ 31.4 [30.2] (CH2,β), 42.8
(m-benzyloxybenzyl)-2-pivaloyl-1,2,3,4-tetrahydroisoquinoline-3-
carboxylic acid (5, 2.0 g, 4.38 mmol) in dry MeOH (120 mL) was
[44.7] (CH2), 43.5 (CH2), 52.7 [52.2] (CHα), 55.2 [54.9] (CH3), 58.2
[59.0] (CH), 66.9 (CH2), 69.6 [69.8] (CH2, OBn), 113.7 [114.3] (CH,
3304
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3300Ϫ3307

1,2,3,4-Tetrahydroisoquinoline-3-carboxylic Acid Derivatives as Opioid Peptide Mimetics
FULL PAPER
arom.), 116.0 [116.2] (CH, arom.), 122.7 [122.2] (CH, arom.), diketopiperazine formation was observed. The workup was per-
126.9Ϫ129.5 (15 CH, arom.), 131.9 [131.8] (C, arom.), 136.9 [136.4] formed as described in procedure 1. A mixture of epimers (major/
(C, arom.), 137.1Ϫ138.7 (3 C, arom.), 156.2 (C-O, arom.), 158.3 minor ϭ 9:1) was obtained; Rf (EtOAc/MeOH, 9:1) 0.12; Rf
[158.7] (CϭO),168.8 [169.0] (CϭO), 170.9 [171.1] (CϭO, methyl (EtOAc/MeOH, 9:1 ϩ AcOH) 0.37. HPLC: Rt ϭ Ϫ28.9 min.
ester) ppm.
(major isomer), 29.0 min (minor isomer). MS (ESϪ): 455, 563 [M
Ϫ Hϩ]. EMM (m/z): [M ϩ Hϩ] calcd. for C34H33N2O6, 565.2338;
found, 565.2347
c[Gly-(1R,3S)-1-(m-Hydroxybenzyl)Tic] (9): Methyl Z-Gly-(1R,3S)-
1-(m-benzyloxybenzyl)-1,2,3,4-tetrahydroisoquinoline-3-carb-
oxylate (8, 0.20 g, 0.35 mmol) was dissolved in EtOH/water (3:1,
40 mL), and Pd/C (10%, 0.10 g) was added. The mixture was placed
under 4 atm hydrogen pressure at room temp. for 4 h. After reac-
tion the Pd catalyst was filtered through a layer of celite and rinsed
with EtOH. The filtrate was evaporated, some water was added,
and the mixture was lyophilized. Yield varied from 1.00 to 1.08 g
or 85 to 95%, beige solid; m.p. 111.3Ϫ113.7 °C. Rf (acetonitrile/
MeOH/water, 4:1:1) 0.78; Rf (EtOAc/MeOH, 9:1) 0.39. HPLC:
Rt ϭ 16.9 min. MS (ESϩ): 113, 149, 167, 215, 323 [M ϩ Hϩ]. EMM
Z-Gly-(1R,3S)-1-(m-benzyloxybenzyl)Tic-Phe-OBn (14): A solution
of Z-Gly-(1R)-1-(m-benzyloxybenzyl)Tic (13, 0.52 g, 0.92 mmol),
Phe-OBn·pTos-OH (0.59 g, 1.38 mmol), and TBTU (0.33 g,
1.01 mmol) in THF/acetonitrile (1:1, 8 mL) was cooled to Ϫ20 °C.
NMM (1.0 mL, 9.20 mmol) was added and the mixture was stirred
for 15 min at Ϫ20 °C. The mixture was then allowed to warm to
room temp. and further stirred overnight. After 24 h all starting
material had been consumed. The solvent was evaporated. The resi-
due was dissolved in EtOAc (40 mL) and the solution was washed
with HClaq (1 , 3 ϫ 30 mL), water (30 mL), saturated NaHCO3
(3 ϫ 30 mL), water (30 mL), and brine (30 mL). After drying
(MgSO4), filtration, and evaporation, the residue was crystallized
from EtOH, which afforded the major single (1R,3S) isomer of Z-
Gly-1-(m-benzyloxybenzyl)Tic-Phe-OBn 14. Yield 0.52 g or 70%,
white needles; m.p. 128.7Ϫ129.1 °C. Rf (EtOAc/cyclohexane, 1:1)
0.36. HPLC: Rt ϭ 33.9 min. MS (ESϩ): 547, 802 [M ϩ Hϩ]. EMM
(m/z): [M
ϩ
Hϩ] calcd. for C19H19N2O3, 323.1395; found,
323.1399. 1H NMR ([D6]DMSO): δ ϭ 2.98Ϫ3.11 (m, 4 H, CH2,
Hβ and HβЈ), 3.64 (d, 1 H, Hα of Gly, J ϭ 17.6), 3.90 (d, 1 H,
HαЈ of Gly, J ϭ 17.6), 4.26 (dd, 1 H, Hα, J1 ϭ 5.3, J2ϭ10.8), 5.75
(1 H, pseudo-t, J ϭ 6.9, N-CH), 6.59 (m, 3 H, H arom.), 7.02 (t,
1 H, H arom., J ϭ 7.5), 7.21 (m, 4 H, H arom.), 8.12 (br. s, 1 H,
NH), 9.21 (br. s, 1 H, OH) ppm. 13C NMR ([D6]DMSO): δ ϭ 32.2
(CH2), 40.2 (CH2), 43.8 (CH2), 51.3 (CH), 52.5 (CH), 113.4 (CH,
arom.), 116.2 (CH, arom.), 119.9 (CH, arom.), 126.2 (CH, arom.),
126.7 (CH, arom.), 127.1 (CH, arom.), 128.8 (CH, arom.), 128.9
(CH, arom.), 132.3 (C, arom.), 135.7 (C, arom.), 139.0 (C, arom.),
157.1 (C-O, arom.), 162.6 (CϭO), 166.2 (CϭO) ppm.
(m/z): [M
ϩ
Hϩ] calcd. for C50H48N3O7, 802.3492; found,
1
802.3469. H NMR (CDCl3): compound 14 exists as a mixture of
two conformational isomers in a ratio of 3:1. The NMR spectro-
scopic data on the minor isomer (if resolved) are given between
square brackets. δ ϭ 2.51 (dd, 1 H, J1 ϭ 5.2, J2 ϭ 14.0, Hβ of
Phe), 2.64 (dd, 1 H, J1 ϭ 5.5, J2 ϭ 15.3, Hβ of Tic or C-CH-Ar
Z-Gly-(1R)-1-(m-benzyloxybenzyl)Tic-OH (13). Procedure 1: Z-
Gly-(1R,3S)-1-(m-benzyloxybenzyl)Tic-OMe (8, 0.30 g, 0.52 mmol) of Bn), 2.70-2.93 (m, 2 H, Hβ of Phe and Hβ of Tic or C-CHϪAr
was dissolved in EtOH/water (2:1, 54 mL). The temperature was of Bn), 3.09Ϫ3.19 (m, 2 H, HβЈ and C-CHЈ-Ar of Bn), 3.75 (m, 1
lowered to 0 °C, and KOH (0.57 mL of a 1  aqueous solution,
0.57 mmol) was added dropwise. The mixture was stirred at 0 °C
for 1 h and was then allowed to warm to room temp. After 4 to 6
H, Hα of Gly), 4.06 [4.11] (m, 1 H, HαЈ of Gly), 4.45 [4.15] (m, 1 H,
Hα/N-CH-Ar of Tic), 4.63 [4.70] (m, 1 H, Hα of Phe), 4.89Ϫ5.06 (4
H, m. 2CH2), 5.12 [5.33] (m, 2 H, CH2), 5.30 (m, 1 H, Hα/N-CH-
days of stirring, all starting material 8 had been consumed (TLC) Ar of Tic), 5.50 [5.62] (br. s, 1 H, NH of Gly), 5.86 [6.0] (d, J ϭ
and the reaction was quenched with a 10% aqueous acetic acid
solution to an acidic pH. The ethanol was evaporated and the water
phase was made basic again with a 6% aqueous NaHCO3 solution.
The solution was extracted with diethyl ether (30 mL). The water
phase was acidified to a pH of 2 with concentrated HCl and again
extracted with dichloromethane (3 ϫ 40 mL). The organic phase
was dried (MgSO4) and filtered, and the solvents were evaporated.
Yield varied from 2.46 to 2.75 or 84 to 94%, oil (mixture of epimers,
major/minor ϭ 9:1).
8.1 Hz, 1 H, NH of Phe), 6.38Ϫ7.40 (m, 28 H, H arom.) ppm. 13
C
NMR (CDCl3): δ ϭ 32.0 [30.9] (CH2), 37.8 [38.3] (CH2), 42.9 [43.6]
(CH2), 44.0 [45.1] (CH2), 53.3 (CH), 57.4 [56.8] (CH), 58.7 [59.7]
(CH), 67.4 [67.6] (CH2), 67.9 (CH2), 70.2 [70.4] (CH2), 114.3 [114.6]
(CH, arom.), 116.5 (CH, arom.), 123.1 [122.7] (CH, arom.),
127.2Ϫ130.2 (25 CH, arom.), 132.4 (C, arom.), 135.2 (2C, arom.),
135.4 (C, arom.), 137.0 (C, arom.), 137.6 (C, arom.), 139.2 (C,
arom.), 156.8 (C-O, arom.), 159.0 [159.3] (CϭO), 169.9 (CϭO),
170.9 (CϭO), 171.1 [171.4] (CϭO) ppm.
Procedure 2:[9] A solution of Z-Gly-(1R,3S)-1-(m-benzyloxyben-
zyl)Tic-OMe (8, 0.59 g, 1.02 mmol) in dry diethyl ether (60 mL)
was placed under nitrogen gas atmosphere, and KOSiMe3 (0.20 g,
1.53 mmol) was added. The mixture was stirred for 30 min at room
temp. All starting material 8 had been consumed, but some diketo-
piperazine (DKP) formation had occurred (4 to 18%). Workup was
performed as described in procedure 1 and additional column chro-
matography was needed for purification. Eluents were varied from
cyclohexane/EtOAc (1:1) ϩ1% AcOH until all DKP was washed
off, to EtOAc ϩ1% AcOH. Yield: 0.32 g or 55%, oil (mixture of
epimers, major/minor ϭ 9:1).
Gly-(1R,3S)-1-(m-hydroxybenzyl)Tic-Phe-OH (15): Z-Gly-(1R,3S)-
1-(m-benzyloxybenzyl)Tic-Phe-OBn (14, 0.16 g, 0.20 mmol) was
dissolved in a mixture of dioxane/water (3:2, 50 mL), and acetic
acid (0.011 mL, 0.20 mmol) was added, followed by Pd/C catalyst
(10%, 0.08 g). This mixture was shaken in a Parr apparatus under
4 atm of hydrogen pressure at room temp. for 2 h. The catalyst was
filtered through a layer of Celite and washed with dioxane and
water, and all dioxane was evaporated off. The aqueous solution
was then lyophilized. A small amount of diketopiperazine 9 was
formed. The tripeptide mimetic 15 was purified by washing of the
fraction over a small silica gel column with acetonitrile to remove
the diketopiperazine 9 and then with acetonitrile/MeOH/acetic acid
(9:1:1) and (5:1:1) to recover the tripeptide 15. Yield 0.04 g or 45%,
Procedure 3:[10] A solution of Z-Gly-(1R,3S)-1-(m-benzyloxyben-
zyl)Tic-OMe (8, 0.08 g, 0.13 mmol) in THF (3 mL) was cooled to
0 °C and nBu4NOHaq solution (40%, 0.213 mL, 0.33 mmol) was beige powder; Rf (EtOAc/MeOH/AcOH 9:1:1) 0.13. HPLC: Rt ϭ
added dropwise, whereupon formation of a bright yellow color was
19.0 min. MS (ESϩ): 166, 323, 388, 488 [M ϩ Hϩ]. EMM (m/z):
observed. The mixture was stirred and the reaction was complete [M ϩ Hϩ] calcd. for C28H30N3O5, 488.2185; found, 488.2181.
after 50 min. All starting material 8 had been consumed and less
NMR gave very broad peaks that were not identifiable.
Eur. J. Org. Chem. 2003, 3300Ϫ3307
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3305

´
E. Mannekens, M. Crisma, S. Van Cauwenberghe, D. Tourwe
FULL PAPER
(1R,3S)-1-(m-Benzyloxybenzyl)-2-pivaloyl-Tic-Phe-OBn (6):
1
A
621.2973. H NMR (CDCl3) (500 MHz): compound 11 exists as a
solution of (1R,3S)-1-(m-benzyloxybenzyl)-2-pivaloyl-Tic (5,
0.50 g, 1.09 mmol), Phe-OBn·pTosOH (0.70 g, 1.63 mmol), and
mixture of two conformational isomers in a 6:4 ratio. The NMR
spectroscopic data on the minor isomer (if resolved) are given be-
TBTU (0.39 g, 1.20 mmol) in dry THF (5 mL) was brought to Ϫ20 tween square brackets. δ ϭ 1.40 [1.39] (s, 9 H, tBu), 2.36 [3.00] (m,
°C under nitrogen atmosphere. NMM (1.2 mL, 10.90 mmol) was
added to the mixture and stirring at Ϫ20 °C was continued for
4 H, Hβ, HβЈ, C-CH2-Ar of Bn), 3.75 (m, 1 H, Hα of Gly), 4.07
(m, 1 H, HαЈ of Gly), 4.48 [4.03] (m, 1 H, Hα/N-CH-Ar of Tic),
15 min. After the mixture had warmed to room temp. it was stirred 4.77Ϫ4.85 (m, 4 H, O-CH2-Ph, CO-O-CH2-Ph), 5.35 [4.88] (d, 1
for another 20 h. The solvent was evaporated and the residue was
dissolved in chloroform (40 mL). The solution was washed with
H, J1 ϭ 3.4, J2 ϭ 7.3, Hα/ NϪCHϪAr of Tic), 5.50 (br. s, 1 H,
NH of Gly), 6.24Ϫ6.50 (m, 2 H, H arom.), 6.64Ϫ7.35 (m, 14 H,
HClaq (1 , 2 ϫ 40 mL), aqueous NaHCO3 solution (6%, 2 ϫ H arom.) ppm. 13C NMR (CDCl3): δ ϭ 27.5 [30.8] (CH2), 29.0 (3
40 mL), and brine (40 mL). After drying, filtration, and evapor-
ation, the residue was crystallized from EtOH. Yield 0.53 g or 70%,
white needles; m.p. 128.2Ϫ128.7 °C. Rf (EtOAc/cyclohexane, 1:1)
CH3), 32.2 [30.9] (CH2), 43.5 [45.4] (CH2), 55.9 [55.7] (CH), 58.9
[59.6] (CH), 68.0 [67.3] (CH2), 70.3 [70.5] (CH2), 80.3 (C, tBu.),
114.4 [114.6] (CH, arom.), 116.7 (CH, arom.), 123.4 [122.9] (CH,
0.55; Rf (EtOAc/cyclohexane, 1:4) 0.15. HPLC: Rt ϭ 36.2 min. MS arom.), 127.4Ϫ130.1 (15 CH, arom.), 132.4 (C, arom.), 135.5 (C,
(ESϩ): 288, 440, 695 [M ϩ Hϩ]. EMM (m/z): [M ϩ Naϩ] calcd. arom.), 136.2 (C, arom.), 137.7 (C, arom.), 139.4 (C, arom.), 156.3
for C45H46N2O5Na, 717.3304; found, 717.3319Ϫ1H NMR
(C-O, arom.), 159.0 [159.3] (CϭO), 169.8 [170.1] (CϭO), 171.0
(CDCl3): δ ϭ 1.27 (s, 9 H, tBu), 2.40Ϫ2.60 (m, 2 H, Hβ and HβЈ [171.2] (CϭO) ppm.
of Phe), 2.75 (m, 2 H, Hβ and CϪCHϪAr of Bn), 3.15 (dd, 1 H,
Boc-Gly-(1R,3S)-1-(m-hydroxybenzyl)Tic-OH (12): A solution of
Boc-Gly-(1R,3S)-1-(m-benzyloxybenzyl)-Tic-OBn (11, 0.36 g,
0.58 mmol) in EtOH/water (3:1, 40 mL) with Pd/C catalyst (10%,
0.10 g) was shaken in a Parr apparatus under 4 atm hydrogen press-
ure at room temp. for 1 h. The catalyst was filtered through a layer
of Celite and washed with ethanol and water, and all ethanol was
evaporated off. The aqueous solution was then lyophilized. Yield
0.23 g or 92%, white powder; m.p. the compound decomposed at
183.6 °C. Rf (EtOAc/MeOH/AcOH, 9:1:1) 0.75; Rf (nBuOH/
MeOH/AcOH, 4:1:1) 0.71. HPLC: Rt ϭ 21.2 min. MS (ESϩ): 130,
164, 186, 341, 385, 441 [M ϩ Hϩ]. EMM (m/z): [M Ϫ Hϩ] calcd.
for C24H27N2O6, 439.1869; found, 439.1883. 1H NMR ([D6]DMSO
ϩ2 drops D2O): compound 12 exists as a mixture of two confor-
mational isomers in a 4:1 ratio. The NMR spectroscopic data on
the minor isomer (if resolved) are given between square brackets.
δ ϭ 1.38 (s, 9 H, tBu), 2.48Ϫ3.09 (m, 4 H, Hβ, HβЈ, C-CH2-Ar of
Bn), 3.97 (m, 2 H, Hα and HαЈ of Gly), 4.42 [4.70] (m, 1 H, Hα of
Tic), 5.19 [4.05] (m, 1 H, N-NH-Ar of Tic), 6.15Ϫ6.33 (m, 2 H, H
arom.), 6.50Ϫ6.70 (m, 2 H, H arom), 6.87Ϫ7.12 (m, 4 H, H arom.)
ppm. 13C NMR ([D6]DMSO ϩ2 drops D2O): δ ϭ 30.2 (3 CH3),
31.5 [30.0] (CH2), 42.8 (2 CH2), 55.9 [54.9] (CH), 57.7 (CH), 78.5
(C, tBu.), 113.4 [113.8] (CH, arom.), 116.8 (CH, arom.), 120.8 (CH,
arom.), 126.0 (CH, arom.), 126.9 (CH, arom.), 127.2 (CH, arom.),
127.7 (CH, arom.), 128.9 (CH, arom.), 133.8 [132.9] (C, arom.),
136.3 (C, arom.), 139.6 (C, arom.), 156.1 (C-O, arom.), 156.9
[157.2] (CϭO), 169.8 [169.5] (CϭO), 173.4 [172.9] (CϭO) ppm.
J1 ϭ 3.1, J2 ϭ 12.5, HβЈ of Tic), 3.25 (dd, 1 H, J1 ϭ 2.8, J2
ϭ
15.3, C-CHЈ-Ar of Bn), 4.42 (m, 1 H, Hα of Phe), 4.89 (s, 2 H,
CO-O-CH2), 4.99 (m, 3 H, N-CH-Ar of Tic and O-CH2-Ph), 5.34
(m, 1 H, Hα of Tic), 6.03 (br. s, 1 H, NH), 6.36Ϫ7.38 (m, 23 H, H
arom.) ppm. 13C NMR (CDCl3): δ ϭ 29.0 (3CH3, tBu), 33.1 (CH2),
38.6 (CH2), 40.8 (C, tBu), 43.2 (CH2), 54.0 (CH), 59.0 (CH), 59.7
(CH), 67.7 (CH2), 70.4 (CH2), 114.0 (CH, arom.), 116.8 (CH,
arom.), 123.4 (CH, arom.), 127.6Ϫ129.5 (20 CH, arom.) 133.7 (C,
arom.), 135.6 (C, arom.), 135.9 (C, arom.), 136.2 (C, arom.), 137.7
(C, arom.), 139.6 (C, arom.), 159.0 (C-O, arom.), 171.0 (CϭO),
172.5 (CϭO), 178.9 (CϭO, piv.) ppm.
Benzyl (1R,3S)-1-(m-Benzyloxybenzyl)-1,2,3,4-tetrahydroisoquino-
line-3-carboxylate (10): (1R,3S)-1-(m-Benzyloxybenzyl)-2-pivaloyl-
1,2,3,4-tetrahydroisoquinoline-3-carboxylic
acid (5,
0.46 g,
1.0 mmol), benzyl alcohol (0.53 mL, 5.1 mmol), and p-toluenesul-
fonic acid monohydrate (0.23 g, 1.2 mmol) were heated at reflux
in benzene solution (10 mL) for 5 h. The reaction solvents were
evaporated. Water (30 mL) was added to the residue, and the solu-
tion was basified with a saturated NaHCO3 solution to pH 9. The
aqueous suspension was then extracted with diethyl ether (3 ϫ
40 mL). The organic fractions were combined, dried (MgSO4) and
filtered, and the solvents were evaporated. A Kugelrohr distillation
was performed (p ϭ 0.15 mm, T ϭ 50Ϫ70 °C) to distill all benzyl
alcohol. The residue was pure 10. Yield 0.47 g or quantitative, oil;
Rf (EtOAc/cyclohexane, 1:1) 0.58. HPLC: Rt ϭ 26.9 min. MS
(ESϩ): 145, 464 [M ϩ Hϩ]. EMM (m/z): [M ϩ Hϩ] calcd. for
C31H30NO3, 464.2225; found, 464.2210. The product was immedi-
ately used in the next reaction.
Gly-(1R,3S)-1-(m-hydroxybenzyl)Tic TFA (2·TFA): A solution of
Boc-Gly-(1R,3S)-1-(m-hydroxybenzyl)-Tic-OH
(12,
0.10 g,
0.23 mmol) in TFA containing 5% of water (5 mL) was stirred for
15 min at room temp. The solvent was evaporated, acetic acid
(10 mL) was added, and the mixture was lyophilized. Yield 0.09 g
or quantitative, sticky beige compound. The compound cyclizes
spontaneously to its diketopiperazine derivative 9 on standing; no
NMR spectroscopic data could therefore be determined; Rf
(EtOAc/MeOH/AcOH, 9:1:1) 0.23. HPLC: Rt ϭ 15.5 min. MS
(ESϩ): 341 [M ϩ Hϩ].
Boc-Gly-(1R,3S)-1-(m-Benzyloxybenzyl)Tic-OBn (11): A mixture
of benzyl (1R,3S)-1-(m-benzyloxybenzyl)-1,2,3,4-tetrahydroisoqui-
noline-3-carboxylate (10, 0.46 g, 1.0 mmol), Boc-Gly (0.26 g,
1.5 mmol), and EDC (0.29 g, 1.5 mmol) in dichloromethane (4 mL)
was stirred at room temp., with protection from daylight. After 1
night stirring, additional Boc-Gly (0.04 g) and 0.2 equiv. EDC
(0.04 g) were added and the mixture was stirred for 2 more days.
The solvent was evaporated and the residue was dissolved in
chloroform (40 mL). This solution was then washed with HClaq (1
, 40 mL), aqueous NaHCO3 (6%, 40 mL) and brine (40 mL),
dried (MgSO4), and filtered, and the solvents were evaporated.
Purification was carried out by flash chromatography with EtOAc/
X-Ray Structure Determination of (1R, 3S)-1-(m-Benzyloxybenzyl)-
2-pivaloyl-Tic-(S)-Phe-OBn (6): Single crystals of 6 were grown
from ethyl acetate/petroleum ether by vapor diffusion. Molecular
formula: C45H46N2O5. Molecular mass: 694.84. Unit cell dimen-
˚
˚
˚
sions: a ϭ 5.6350(10) A, b ϭ 12.315(2) A, c ϭ 15.010(2) A, α ϭ
cyclohexane (1:6) as eluent. Yield 0.37 g or 60%, oil; Rf (EtOAc/ 112.37(3)°, β ϭ 96.50(3)°, γ ϭ 96.54(3)°. Z ϭ 1; dcalc ϭ 1.224 Mg/
cyclohexane 1:1) 0.57; Rf (EtOAc/cyclohexane 1:4) 0.16. HPLC:
m3; crystal system: triclinic; space group: P1. Crystal size: 0.45 ϫ
Rt ϭ 34.3 min. MS (ESϩ): 163, 179, 279, 335, 621 [M ϩ Hϩ]. EMM 0.30 ϫ 0.20 mm. Absorption coefficient: 0.630 mmϪ1. Data collec-
(m/z): [M
ϩ
Hϩ] calcd. for C38H41N2O6, 621.2964; found, tion was performed on a Philips PW1100 four-circle diffractometer
3306
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 3300Ϫ3307

1,2,3,4-Tetrahydroisoquinoline-3-carboxylic Acid Derivatives as Opioid Peptide Mimetics
FULL PAPER
[2]
B. C. Wilkes, P. W. Schiller, Biopolymers 1994, 34, 1213Ϫ1219.
P. A. Temussi, S. Salvadori, P. Amodeo, C. Bianchi, R. Guer-
rini, R. Tomatis, L. H. Lazarus, D. Picone, T. Tancredi, Bio-
chem. Bioph. Res. Co. 1994, 198, 933Ϫ939.
with use of graphite-monochromated Cu-Kα radiation (λ ϭ
1.54178 A) in the θ-2θ scan mode up to θ ϭ 60°. Limiting indices:
Ϫ6 Յ h Յ 6, Ϫ13 Յ k Յ 12, 0 Յ l Յ 16. A total of 2810 reflections
were collected, 2801 of which were symmetry-independent and
2653 of which had I Ն 2σ(I).
[3]
˚
[4]
[5]
A. Coop, A. E. Jacobson, Bioorg. Med. Chem. Lett. 1999, 9,
357Ϫ362.
I. M. P. Huber, D. Seebach, Helv. Chim. Acta 1987, 70,
1944Ϫ1954.
K. Hayashi, Y. Ozaki, K. I. Nunami, N. Yoneda, Chem. Pharm.
Bull. 1983, 31, 312Ϫ314.
A. Vanoeveren, J. F. G. A. Jansen, B. L. Feringa, J. Org. Chem.
1994, 59, 5999Ϫ6007.
D. Seebach, J. J. Lohmann, M. A. Syfrig, M. Yoshifuji, Tetra-
hedron 1983, 39, 1963Ϫ1974.
E. D. Laganis, B. L. Chenard, Tetrahedron Lett. 1984, 25,
5831Ϫ5834.
A. F. Abdel-Magid, J. H. Cohen, C. A. Maryanoff, R. D. Shah,
F. J. Villani, F. Zhang, Tetrahedron Lett. 1998, 39, 3391Ϫ3394.
F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G.
Orpen, R. Taylor, J. Chem. Soc., Perkin Trans. 2 1987, S1ϪS19.
C. J. Creighton, T. T. Romoff, J. H. Bu, M. Goodman, J. Am.
Chem. Soc. 1999, 121, 6786Ϫ6791.
Z. Darula, K. E. Köver, K. Monory, A. Borsodi, A. Mako,
A. Ronai, D. Tourwe, A. Peter, G. Toth, J. Med. Chem. 2000,
43, 1359Ϫ1366.
The structure was solved by direct methods with the aid of the
SHELXS 97 program,[17] and refined by full-matrix block least-
squares on F2, by using all data, with all non-H atoms anisotropic,
by application of the SHELXL 97 program,[18] allowing the posi-
tional parameters and the anisotropic displacement parameters of
the non-H atoms to refine at alternate cycles. All phenyl rings were
constrained to the idealized geometry. H-atoms were placed at ide-
alized positions and refined as riding, with Uiso set equal to 1.2 (or
1.5 for methyl groups) times the Ueq of the parent atom. Refine-
ment converged to R1 ϭ 0.0397 and wR2 ϭ 0.1135. Data/restraints/
parameters ratio: 2801:3:412.
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
CCDC-200237 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge at
www.ccdc.cam.ac.uk/conts/retrieving.html [or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, UK; Fax: (internat.) ϩ44-(0)1223-336-033; E-mail:
deposit@ccdc.cam.ac.uk].
´
´
[14]
[15]
[16]
[17]
´
F. Gosselin, D. Tourwe, M. Ceusters, T. Meert, L. Heylen, M.
Jurzak, W. D. Lubell, J. Pept. Res. 2001, 57, 337Ϫ344.
A. N. Tyler, E. Clayton, B. N. Green, Anal. Chem. 1996, 68,
3561Ϫ3569.
A. Madrigal, M. Grande, C. Avendano, J. Org. Chem. 1998,
63, 2724Ϫ2727.
G. M. Sheldrick, SHELXS 97. Program for the Solution of
Crystal Structures. University of Göttingen, Göttingen, Ger-
many, 1997.
G. M. Sheldrick, SHELXL 97. Program for the Refinement of
Crystal Structures, University of Göttingen, Göttingen, Ger-
many, 1997.
Acknowledgments
The authors acknowledge the help of A. Borsodi and S. Benyhe
(Biological Research Center, Szeged, Hungary) for the receptor
binding assay, M. Jurzak (J&J, PRI, Janssen Pharmaceutica,
Belgium) for the signal transduction assay, and T. Meert (J&J, PRI,
Janssen Pharmaceutica, Belgium) for the in vivo experiments.
[18]
[1]
M. E. Fundytus, P. W. Schiller; M. Shapiro, G. Weltrowska, T.
J. Coderre, Eur. J. Pharmacol. 1995, 286, 105Ϫ108.
Received March 13, 2003
Eur. J. Org. Chem. 2003, 3300Ϫ3307
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3307