Replacement of glycine with dicarbonyl and related moieties in analogues of the C-terminal pentapeptide of cholecystokinin: CCK2 agonists displaying a novel binding mode

Article abstract of DOI:10.1021/jm0000416

Recent advances in the field of cholecystokinin have indicated the possible occurrence of multiple affinity states of the CCK₂ receptor. Besides, numerous pharmacological experiments performed 'in vitro' and 'in vivo' support the eventuality of different pharmacological profiles associated to CCK₂ ligands. Indeed, some agonists are essentially anxiogenic and uneffective in memory tests, whereas others are not anxiogenic and appear as able to reinforce memory. The reference compound for the latter profile is the CCK-8 analogue BC 264 (Boc-Tyr(SO₃H)-gNle-mGly-Trp-(NMe)Nle-Asp-Phe-NH₂). However, although tetrapeptide ligands based on CCK-4 (Trp-Met-Asp-Phe-NH₂) are known to possess sufficient structural features for CCK₂ recognition, none shares the properties of BC 264. Hence we have developed new short peptidic or pseud-opeptidic derivatives containing the C-terminal tetrapeptide of BC 264. Our results indicate that some compounds characterized by the presence of two carbonyl groups at the N-terminus, as in 2b (HO₂C-CH₂-CONH-Trp-(NMe)Nle-Asp-Phe-NH₂), are likely to show a BC 264-like profile, bind to the CCK₂ receptor in a specific way, and display remarkable affinities (2b: 0.28 nM on guinea-pig cortex membrane preparations). This original binding mode is discussed and further enlightened by NMR and molecular modeling studies.

Full text of DOI:10.1021/jm0000416

3614
J . Med. Chem. 2000, 43, 3614-3623
Rep la cem en t of Glycin e w ith Dica r bon yl a n d Rela ted Moieties in An a logu es of
th e C-Ter m in a l P en ta p ep tid e of Ch olecystok in in : CCK2 Agon ists Disp la yin g a
Novel Bin d in g Mod e
Bruno Bellier,†,‡ Marie-Emmanuelle Million,†,‡ Sophie DaNascimento,‡ Herve´ Meudal,‡ Safia Kellou,§
Bernard Maigret,§ and Christiane Garbay*,‡
De´partement de Pharmacochimie Mole´culaire et Structurale, U266 INSERM, UMR 8600 CNRS, UFR des Sciences
Pharmaceutiques et Biologiques, 4, Avenue de l’Observatoire, 75270 Paris Cedex 06, France, and Laboratoire de Chimie
The´orique, S.R.S.M.C. ‚ UMR 7565 CNRS, Universite´ Nancy 1 - Universite´ Henri Poincare´, B.P. 239 - Domaine Scientifique
Victor Grimard, 54506 Vandoeuvre-les-nancy Cedex, France.
Received J anuary 31, 2000
Recent advances in the field of cholecystokinin have indicated the possible occurrence of multiple
affinity states of the CCK2 receptor. Besides, numerous pharmacological experiments performed
“in vitro” and “in vivo” support the eventuality of different pharmacological profiles associated
to CCK2 ligands. Indeed, some agonists are essentially anxiogenic and uneffective in memory
tests, whereas others are not anxiogenic and appear as able to reinforce memory. The reference
compound for the latter profile is the CCK-8 analogue BC 264 (Boc-Tyr(SO3H)-gNle-mGly-
Trp-(NMe)Nle-Asp-Phe-NH2). However, although tetrapeptide ligands based on CCK-4 (Trp-
Met-Asp-Phe-NH2) are known to possess sufficient structural features for CCK2 recognition,
none shares the properties of BC 264. Hence we have developed new short peptidic or pseudo-
peptidic derivatives containing the C-terminal tetrapeptide of BC 264. Our results indicate
that some compounds characterized by the presence of two carbonyl groups at the N-terminus,
as in 2b (HO2C-CH2-CONH-Trp-(NMe)Nle-Asp-Phe-NH2), are likely to show a BC 264-like
profile, bind to the CCK2 receptor in a specific way, and display remarkable affinities (2b: 0.28
nM on guinea-pig cortex membrane preparations). This original binding mode is discussed
and further enlightened by NMR and molecular modeling studies.
In tr od u ction
Trp-Nle-Asp-Phe-NH2)11 and the modified heptapeptide
BC 264 (Boc-Tyr(SO3H)-gNle-mGly-Trp-(NMe)Nle-Asp-
Phe-NH2), which is resistant to enzymatic degrada-
tion.12 These two compounds exert somewhat different
pharmacological properties, although both of them are
selective CCK2 agonists.
The neurotransmitter and neuromodulator peptide
cholecystokinin (CCK) is known to intervene in various
physiological processes, in both central and peripheral
nervous systems. Its biological actions are mediated by
two G-protein coupled receptors designated CCK1 and
CCK2.1 The latter is involved in phenomenons as diverse
as anxiety,2 analgesia,3 memory processes,4 or psychi-
atric disorders.5 The actions of CCK are predominantly
due to the sulfated octapeptide CCK-8, Asp-Tyr(SO3H)-
Met-Gly-Trp-Met-Asp-Phe-NH2, and they may result
from interactions with other neurotransmitters such as
dopamine6 or enkephalins7 and, as shown more recently,
serotonine.8
On the basis of binding experiments using CCK2
selective agonists9a-c or antagonists,9d,e several authors
have suggested the occurrence of CCK2 receptor sub-
types or subsites. Nevertheless, the cloning of CCK2
receptors has not proven the existence of structurally
different entities yet, as far as proteic primary sequence
is concerned.10
First, behavioral experiments have proved BC 197 to
be anxiogenic for rats in the (+)-maze test13 and to have
no effect on the locomotor activity,15a whereas BC 264
does not exert any anxiogenic effects at doses as high
as 10 mg/kg13 and increases the activity of the animals
in both the actimeter and the open-field tests.15a
Besides, in the more elaborated memory test of Dellu
et al.,14 BC 264 increases the memory of rats while BC
197 seems to be amnesia-inducing.15b In the Y-maze
test, which measures the attention and/or the working
memory of the animals, BC 264 promotes the attention,
while BC 197 remains uneffective.15a Moreover, these
compounds have also opposite actions on the grooming
behavior of animals, which is enhanced by BC 264 and
inhibited by BC 197.15c Last, “in vitro”15d and “in vivo”
measurements of dopamine release in the anterior
nucleus accumbens have once again confirmed a dif-
ferential action of BC 197 and BC 264, the latter being
able in particular to increase the extracellular levels of
dopamine in the nucleus accumbens of awake animals
while BC 197 is not.15e Thus, the differential effects of
BC 197 and BC 264 have been associated with phar-
macological profiles termed respectively “CCK-B1” and
Pharmacological studies conducted in our laboratory
have focused on two peptidic probes to further explore
CCK2 receptor potential heterogeneity: the cyclic oc-
tapeptide BC 197 (Boc-c[D-Asp-Tyr(SO3H)-Nle-D-Lys]-
* Address correspondence to Professor Ch. Garbay. Tel: (33) 01-
43-25-50-45. Fax: (33) 01-43-26-69-18. E-mail: Garbay@pharmacie.univ-
paris5.fr.
† Both authors contributed equally to this work.
‡ De´partement de Pharmacochimie Mole´culaire et Structurale.
§ Laboratoire de Chimie The´orique.
10.1021/jm0000416 CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/19/2000

CCK2 Agonists Displaying a Novel Binding Mode
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20 3615
“CCK-B2”, which will now, according to the IUPHAR
nomenclature, be renamed “CCK2A” and “CCK2B”.
Subsequent amidation of 2 with benzylamine in the
presence of BOP led to the corresponding amide deriva-
tive. Direct amidation of 2 with ammonia did not
provide the monoamide of malonic acid. Alternatively,
alcoholysis of ethyl cyanoacetate, following procedure
B, yielded the desired product 6.19b Last, monobenzyl
succinate 4 was prepared by the nucleophilic ring
opening of succinic anhydride by benzyl alcohol.19c
Thereafter, introduction of supplementary substitu-
ents at the N-terminus of I was again achieved by
standard techniques: acid esters 2-4, and 8, or com-
mercial sulfoacetic acid (7), monoethyl fumarate (9),
monoethyl adipate (10), methoxyacetic acid (11), Z-
glycine (12), Z-alanine (13), Z-D-alanine (14), Z-aspartic
acid â-benzyl ester (15), Z-D-aspartic acid â-benzyl ester
(16), Boc-â-alanine (17), acetoacetic acid (18), Boc-L or
D-aspartic acid R or â benzyl ester (19-22), or acid
amides 5 (prepared following procedure C) and 6 were
coupled to I in the presence of BOP and DIEA to give
protected pentapeptides 2a -22a . Catalytic hydrogena-
tion of 2a -7a and 12a -22a or saponification of 8a -
10a gave the final deprotected compounds 2b-22b
(Scheme 1).
All these products were purified by semipreparative
HPLC. Compound 8b (methylmalonyl-Trp-(NMe)Nle-
Asp-Phe-NH2) was the mixture of two epimers; only one
could be isolated as a pure entity, while the other was
obtained in a 40% diastereomeric excess. Compound 17b
(Boc-â-Ala-Trp-(NMe)Nle-Asp-Phe-NH2) was further
deprotected in a TFA/CH2Cl2 mixture with anisole as
scavenger, to yield 17c (Scheme 1).
Compounds 23a -26a were synthesized by acylation
of I with ethyl oxalyl chloride, methyl chloroformate,
propionyl chloride, and isobutyryl chloride, respectively.
Deprotection of the aspartate side chain by saponifica-
tion (23a ) or hydrogenolysis (24a -26a ) led to 23b-26b.
Nucleophilic opening of maleic anhydride by I19d in a
mixture of DMF and triethylamine gave 27a , whose
deprotection finally afforded 27b (Scheme 1).
Derivatives 28b (Ac-Trp-(NMe)Nle-Asp-Phe-NH2) and
29b (malonyl-(NMe)Trp-(NMe)Nle-Asp-Phe-NH2) were
obtained from the intermediate tripeptide II (NMe)Nle-
Asp(OBzl)-Phe-NH2, which was condensed with Ac-Trp-
OH, or Boc-NMe-Trp-OH giving aspartate-protected 28a
and 29a , respectively. Compound 29a was deprotected
with TFA, coupled to malonic acid monobenzyl ester (2),
and the resulting pentapeptide was submitted to H2 on
Pd/C to give 29b, while hydrogenolysis of 28a gave 28b
(Scheme 2).
N-Acylation of I and II by either pathway and
deprotection of the esters and carbamates were always
carried out with good yields and did not necessitate any
intermediate purification but only extractions and/or
washings. It is important to note that all (NMe)Nle-
containing peptides were characterized by the occur-
rence of two NMR signals for numerous protons, due to
the equilibrium between two rotamers around the Trp-
(NMe)Nle bond.
To support the assumption of two dissociated profiles
for CCK2 agonists, it was essential to search for new
compounds belonging to other chemical series and
meeting the characteristics of either profile. Since
CCK-4 (Trp-Met-Asp-Phe-NH2) is the minimal feature
for CCK2 recognition, it was then focused on tetrapep-
tide-like compounds, hoping they would be separable
between “CCK2A” and “CCK2B” agonists. Thus, cyclic
CCK-4 analogues RB 360 ([N-cycloamido)-RMe(R)-Trp-
[(2S)-2-amino-9-(cycloamido)carbonyl)nonanoyl]]-Asp-
Phe-NH2) and RB 380 ([cyclo-S-S-[(5-thiopentanoyl)-
RMe(R)Trp-Cys]]-Asp-Phe-NH2)16 were designed to mimic
the preferential folded conformation of the linear tet-
rapeptide Boc-Trp-(NMe)Nle-Asp-Phe-NH2. These com-
pounds were shown to be anxiogenic and devoid of effect
in the Y-maze test and on the locomotor activity. This
strongly suggested that these two compounds had a BC
197-like profile, that is to say “CCK2A”. Until the
beginning of the research reported here, no CCK-4
analogue had revealed a “CCK2B” profile so far, when
this profile is clearly the most interesting. The search
for new “CCK2B” compounds appeared then to be crucial,
as well for a better understanding of the potential
pharmacological heterogeneity of CCK2 receptors, as for
developing tools chemically simpler than BC 264.
Keeping in mind that CCK-4 analogues may furnish
a convenient basis for designing new CCK2 agonists, the
C-terminal tetrapeptide of BC 264 (Trp-(NMe)Nle-Asp-
Phe-NH2) was chosen as the starting point for the quest
of “CCK2B” molecules. Unfortunately, NMR and com-
puter-aided studies on BC 264 did not identify a
preferential conformation of this compound. Therefore,
the only possible rational guideline was to add step by
step the N-terminal structural features of BC 264 in
order to determine the key pattern for “CCK2B” selectiv-
ity.17 As a matter of fact, previous investigations had
essentially been directed on the CCK-8 series, with a
systematic analysis of the role played by each amino
acid in the octapeptidic sequence. CCK-4 derivatives
have also been widely studied with the aim of obtaining
small and restrained CCK2 ligands. But the influence
of the N-terminus of a “CCK-5-like” structure remained
to be focused on, for no particular interest had been
taken in pentapeptide CCK analogues. In this paper,
we describe several types of modifications (constrained
or not) performed on the N-terminal part of the CCK-4
analogue Trp-(NMe)Nle-Asp-Phe-NH2 and try to pro-
pose a differential mode of receptor recognition for these
compounds in order to analyze their binding character-
istics.
Resu lts
Ch em istr y. The starting material TFA‚Trp-(NMe)-
Nle-Asp(OBzl)-Phe-NH2 (I) was prepared in large
amounts by standard liquid phase peptide synthesis
essentially as previously described.18 Some agents to be
used to modify the N-terminus of I had to be prepared.
Thus, malonic acid monobenzyl ester (2) or monomethyl
ester (3) or methylmalonic acid monoethyl ester (8) were
obtained by selective monosaponification of the corre-
sponding diesters.19a Compound 8 was obtained as a
racemic mixture, which did not need to be separated.
Discu ssion
The aim of this study was to explore the influence of
the N-terminal extension of CCK-4-like CCK2 agonists
represented by the derivative Trp-(NMe)Nle-Asp-Phe-
NH2 (1c). The compounds synthesized were tested for

3616 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20
Bellier et al.
Sch em e 1
Sch em e 2
their ability to displace [3H]pCCK-8 from preparations
of CHO cells stably transfected with the rat CCK2
receptor and, for the most interesting ones, from guinea-
pig pancreatic membranes to assay their CCK1 affinity,
and their affinities were compared to the those of the
reference agonists 1b (Boc-Trp-(NMe)Nle-Asp-Phe-NH2,
Ki ) 2.6 nM) and BC 264 (Ki ) 0.31 nM). The results
are reported in Tables 2 and 3, while the affinities of
standard CCK ligands are recalled in Table 1. Finally,
the compounds endowed with the best affinities were
evaluated for their CCK2 agonists profile by measuring
their ability to stimulate inositol phosphate production
in transfected CHO cells, as described previously.16
At first sight, the enlargement of 1c was designed to
obtain C-terminal fragments of BC 264 and thus to
determine the elements responsible for the specific
activity of the parent compound. This resulted in the
preparation of pentapeptidic derivatives characterized

CCK2 Agonists Displaying a Novel Binding Mode
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20 3617
Ta ble 1. Affinities of Standard Cholecystokinin Agonists
a
Ki values (expressed in nM) represent the mean of three separate experiments, each performed in triplicate for determining CCK2
b
and CCK1 affinities on guinea-pig cortex and pancreatic memebranes respectively. Ki values are the mean of three separate experiments,
each performed in triplicate on the rat brain CCK2 receptor expressed in CHO cells. c Results (expressed in nM) are the mean of three
separate measurements of inositol phosphate production induced by the compounds each in triplicate. From ref 24. e S configuration at
d
the * carbon.
by the N-terminal feature X-CO-CH2-CO-NH-Trp, cor-
responding to the retro-inverso bond characteristic of
the sequence of BC 264. In the following sections, the
N-terminus of the discussed compounds (R-NH-Trp-
(NMe)Nle-Asp-Phe-NH2) will be recalled if necessary for
perfect clarity, for instance 2b (R ) HO2C-CH2-CO). The
binding properties of these compounds are presented in
Table 2, which brings out the remarkable subnanomolar
affinities of 2b (R ) HO2C-CH2-CO) and 7b (R ) HO3S-
CH2-CO) (0.75 nM and 0.54 nM, respectively). Thus, an
acidic N-terminus seems favorable for receptor recogni-
tion, but it must not be compulsory, as proved by
compounds 3b (R ) H3CO2C-CH2-CO), 5b (R ) PhCH2-
NH-CO-CH2-CO), and 6b (R ) H2N-CO-CH2-CO), which
retain nanomolar affinities. These first results high-
lighted that a hydrophilic N-terminal capping of CCK-4
induced an increase in affinity, as did a hydrophobic
capping. This indicated the occurrence of two different
binding processes to the receptor followed by 2b and 1b,
respectively. Therefore, several new derivatives were
prepared in order to further characterize the new
hydrophilic binding mode.
Op tim iza tion of th e Len gth of th e N-Ter m in a l
Acid Su bstitu en t. Considering the successful addition
of a malonate residue, mimicking the retro-inverso
glycine in BC 264, we further attempted to optimize the
spatial location of the N-terminal negative charge. The
distance between the two carbonyl groups was length-
ened to two and five methylenes, with compounds 8b
(R ) HO2C-(CH2)2-CO) and 10b (R ) HO2C-(CH2)5-CO),
respectively. Addition of one methylene bridge to 2b
already decreases the affinity by a 4-fold factor, while
a longer extension is really dramatic. On the contrary,
shortening this distance by replacing the malonate by
an oxalate in 23b (R ) HO2C-CO) also considerably
reduced the affinity, probably because of both an
increased rigidity and an unsufficient distance to the
tryptophan at this level. This suggests relatively strict
requirements for the recognition of a N-terminal car-
boxylate by the receptor.
Con for m a tion a lly Restr icted Der iva tives. An-
other mean of studying the apparently important role
of the N-terminal acidic moiety was to introduce con-
formational restraints at this level. The N-methylation
of the tryptophan residue in 2b was a way in this
purpose. The corresponding di-N-methylated pentapep-
tide 29b (HO2C-CH2-CO-(NMe)Trp-(NMe)Nle-Phe-NH2)
suffered a dramatic loss in affinity (Ki ) 3860 nM),
probably indicating either a steric hindrance or the loss
of an hydrogen bond essential for receptor recognition.
Ron and co-workers20 had already studied the effects
of various backbone N-methylations in the CCK-8 series.
Single N-methylation of Trp did not seem to avoid
binding to CCK2 receptors, but in our case, double
N-methylation probably results in too strong steric
constraints and eventually in an increased number of
inactive conformers. Replacement of the malonate by
either a fumarate (9b) or a maleate (21b) constrained
the carboxylate moiety with only a little increase in the
distance to the tryptophan amide nitrogen. Curiously,
this N-acylation of I resulted in both cases in two
separable products, each of these being itself a couple
of N-cis/N-trans rotamers around the Trp-(NMe)Nle
bond as revealed by the presence of two NMR signals
for the N-methyl group. NMR and mass spectra of these
compounds strongly suggested that they also be rota-
mers of a single molecule, presumably because of a slow
isomerization of the conjugated N-terminal system.
Actually, the CH-CO bonds are here partial double
bonds and may present sufficiently high rotation bar-
riers to give rise to separable products. However, none
of these showed a good affinity for the CCK2 receptor,
but conversely, all had micromolar potencies. This
probably indicated the blockade of the acid moiety in
an unappropriate position, or electronic and/or steric
unfavorable effects of the N-terminal unsaturated chain.

3618 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20
Bellier et al.
Ta ble 2. Affinities of N-Terminal Modified RNH-Trp-(NMe)Nle-Asp-Phe-NH2 Derivatives of CCK-4
d
a-c See footnotes a-c in Table 1. Each value corresponds to a putative rotamer (see text for details). n.d.: not determined.

CCK2 Agonists Displaying a Novel Binding Mode
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20 3619
Ta ble 3. Influence of a a-substitution of the N-terminal Residue in CCK-5 Derivatives RNH-Trp-(NMe)Nle-Asp-Phe-NH2
a,b
See footnotes a and b in Table 1. c Diastereoisomers could not be unambiguously characterized. n.d.: not determined.
Sch em e 3
Alk yla ted N-Ter m in a l Der iva tives. To fully un-
derstand the influence of a carbonyl functionality at a
γ position from the Trp NH (Scheme 3), a second series
of derivatives of I was prepared, with alkyl or amine
N-terminal substituents. In the alkyl series, compound
28b (R ) H3C-CO) curiously exerted only a weak affinity
(228 nM) whereas Ac-CCK-4 is a potent CCK2 agonist.
This is all the more surprising as the substitution of
Met by (NMe)Nle in the CCK-4 series generally in-
creases the affinity for the CCK2 receptor, as exemplified
by the comparison of 1b (Ki ) 2.6 nM) with Boc-CCK-4
(Ki ) 9.4 nM). However, increasing the alkyl chain
length improves the affinity, as proved by the sequence
28b (R ) acetyl, Ki ) 228 nM); 25b (R ) propionyl, Ki
) 4.35 nM); 1b (R ) tert-butoxycarbonyl, Ki ) 2.6 nM).
Conversely, the affinity of positively charged derivatives

3620 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20
Bellier et al.
of I decreases when the distance from the NH Trp to
the N-terminal charge is increased: this is shown by the
sequence 1c (R ) H, Ki ) 2.6 nM); 12b (R ) Gly, Ki )
18 nM); 17c (R ) âAla, Ki ) 95 nM). Compound 12b
(Gly-Trp-(NMe)Nle-Asp-Phe-NH2) had already been
described with an affinity for rat cortex of 3 nM.
Surprising as it may be, the 6-fold difference found here
in CHO cells is not fully inconsistent, since variations
between biological models have often been observed, as
exemplified by the 4-fold difference in affinity of the
reference antagonist L-365,260 for rat cortex (3.9 nM)
and CHO cells (15 nM).
receptor. In the second case, such a substitution could
lead to steric repulsion between the methyl and a
neighboring group in the receptor. The 20-fold (at least)
difference in affinity between the two isomers of 4b
speaks rather for the second hypothesis, since the
spatial position of the features of the ligand is generally
crucial for receptor recognition. The structural hypoth-
esis presented in these sections are schematized in
Scheme 3.
Asp a r ta te Der iva tives. Finally, an attempt was
made to combine the favorable hydrophobic features
found in 17b (R ) Boc-âAla, Ki ) 1.35 nM) to the
hydrophilic malonates, with the introduction of an
aspartyl residue at the N-terminus. The amide func-
tionality was left unprotected in 15b (R ) L-Asp, Ki )
4.8 nM) and 16b (R ) D-Asp, Ki ) 11.9 nM) (which can
be considered as hybrids of 8b (R ) HO2C-(CH2)2-CO,
Ki ) 3 nM) and 12b (R ) H2N-CH2-CO, Ki ) 18 nM))
whose affinities were little different from those of the
related compounds.
These two series of results strongly tend to indicate
the existence of an interaction of N-terminal hydropho-
bic moieties with a hydrophobic pocket at this level of
the receptor. Thus, the conformational restraint induced
by (NMe)Nle would prevent recognition of this pocket
by the acetyl moiety, while propionyl and Boc are long
enough to counteract this effect. This would explain why
only 28b (R ) CH3CO) is far less efficient than its Met-
containing analogue. On the contrary, a positively
charged N-terminus is only tolerated when it does not
produce too much unfavorable hydrophilic/hydrophobic
repulsion, that is to say when the amine moiety may
be far enough from this putative hydrophobic pocket.
However, this interpretation may seem inconsistent
with the ability of hydrophilic, acid-N-terminal deriva-
tives such as 2b (R ) HO2C-CH2CO) to bind with the
highest affinities to the receptor, unless these com-
pounds have a specific binding mode. Such an hypoth-
esis necessitated further inquiries on the role of the
γ-carbonyl or eventually of the sequence CO-CH2-CO,
which is specific of BC 264 in the octapeptide series and
hence a major point of interest.
Protection of the amine with a Boc group was hoped
to significantly improve the affinity of these pentapep-
tides, as in 17b (R ) Boc-âAla, Ki ) 1.35 nM) versus
17c (R ) âAla, Ki ) 95 nM). Unfortunately, the four
Boc-aspartate-containing compounds 19b-22b had all
comparable affinities (Table 3), suggesting that the
favorable effects of a N-terminal hydrophobic group and
of a second carbonyl at the same level cannot be added,
presumably because the binding of either feature to the
corresponding sites of the receptor furthers the other
off its own binding pocket. The negligible effect of the
configuration of the chiral carbons on the affinity further
supports this hypothesis, since it proves that neither
compound is able to simultaneously recognize two (or
more) essential regions of the receptor.
N-Ter m in a l Br a n ch ed An a logu es. Therefore, com-
pound 4b (R ) HO2C-CH(CH3)CO) was prepared in
order to study the influence of a substitution at the â
position (toward the NH Trp). Astonishingly, the intro-
duction of a methyl substituent produced a considerable
(more than 200-fold for the better diastereomer, 4000-
fold for the other) loss of affinity as compared to 2b.
This was all the more surprising as several potent CCK2
ligands are substituted at this position, which is the case
for BC 197. In particular, substitution of Gly29 by Ala
or D-Ala in the CCK-8 series did not induce such a
decrease in affinity, but only a 2- and 5-fold loss,
respectively.22a,b Consequently, the compounds 13b and
14b were prepared, with Ala and D-Ala replacing the
N-terminal methylmalonate. Both pentapeptides re-
tained a good affinity (20 and 15 nM, respectively) for
CCK2 receptors, and they did not show any significant
difference with their glycine-containing analogue 12b
(see Table 3). Taken together, these results again
indicate a very original behavior of malonate-containing
derivatives of 1b: in fact, the â position is crucial only
for the latter whereas it seems of little importance for
fully peptidic derivatives.
Ad d ition a l Im p lica tion s for th e Recep tor Bin d -
in g Requ ir em en ts. The properties of the aspartate
derivatives compared to that of, on one hand, the alkyl
analogues (1b (R ) Boc)-28b (R ) CH3CO)-25b (R )
CH3CH2CO)) and, on the other hand, the malonate ones
(2b (R ) HO2C-CH2CO), 8b (R ) HO2C-(CH2)2-CO), and
4b (R ) HO2C-CH(CH3)CO), in particular) deserve a
more accurate discussion. It seems surprising that
aspartate derivatives, which can be seen also as branched
succinates, are tolerated by the receptor while the
substitution of the â position by a methyl (compound
8b) was forbidden. This must mean that the binding
feature crucial for the high affinity of aspartate deriva-
tives 19b-22b is the Boc group and that the resulting
spatial arrangement does not lead to the same orienta-
tion of the carboxylate as in 2b and thus avoids the
steric hindrance at the â position seen in 4b. Schemati-
cally, as outlined in Scheme 4, since it appears as
impossible to recognize both features of the receptor,
the ligand makes the choice of the hydrophobic pocket
as a priority over the feature binding of the malonate
(or more generally hydrogen bond-accepting) deriva-
tives. In fact, the thermodynamic contribution of the
hydrophobic interactions at this level of the receptor
might override that of the malonate recognition, but the
steric constraints involved in the hydrophilic recognition
may also force the ligand to adopt another fit.
Thus, two hypothesis may be inferred here about the
role of the N-terminal carbonyl group. On one hand, it
could participate in the structuration of the ligand in a
particular way; on the other hand, it could recognize a
particular feature of the receptor. In the first case,
substituting the â position by a methyl could prevent
the convenient folding and thus any recognition of the
This suggests that the “default” binding mode of the
ligands to the CCK2 receptor is that described for

CCK2 Agonists Displaying a Novel Binding Mode
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20 3621
F igu r e 1. Conformational analysis of compounds 1b, 2b, and RB 360 (cyclic analogue of 1b). It clearly appears that extended
conformations are more important in the case of 2b and folded ones in the case of 1b. Only folded conformations are found in the
case of the cyclic RB 360.
Sch em e 4
characterized the folded structure of 1b, proves that 2b
adopts a different, presumably unfolded conformation.
Very interestingly, this is also the case for BC 264,
whose dominant conformations are likely to be unfolded.
Computer-aided molecular studies were also performed
on 1b and 2b, involving energetic minimization followed
by conformational analysis. Interestingly, the super-
position of the backbones of the resulting conformers
suggested the occurrence of two families of conforma-
tions, extended and folded (Figure 1). The folded con-
formations seemed to be predominant in 1b while the
extended ones would be more populated in 2b, in
accordance with NMR data. As expected, the analysis
of the cyclic derivative of 1b, RB 360 (whose formula
and properties are recalled in Table 1), showed only the
folded structures, which confirms the validity of our
approach.
Con clu sion
The search for new CCK2 agonists possessing the
desirable pharmacological properties of the well-de-
scribed compound BC 264 has led us to an extensive
structure-affinity study of the role of the N-terminal
residue in CCK-derived pentapeptides. Introduction of
derivatives of malonic acid has led to highly selective
CCK2 agonists, eventually endowed with subnanomolar
affinities. Most interestingly, some of the new com-
pounds have been shown to bind to the receptor in an
unusual way, which can be thought to be that of BC
264. This resemblance is supported by NMR and
computer-aided structural studies, confirming a similar
behavior of malonate-containing pentapeptides and BC
264. The results presented above support the proposal
of two binding modes to CCK2 receptors: one chosen
by CCK-4-derived ligands possessing an hydrophilic
N-terminus, the other being that of hydrophobic ana-
logues. Among the two binding modes characterized,
hydrophobic interactions seem to be dominant in the
recognition of all formerly known ligands and should
have priority on hydrophilic recognition. This hydro-
philic mode probably characterizes BC 264 and mal-
compound 1b, which could explain that BC 264 had
remained an exception for such a long time.
Str u ctu r a l Stu d ies. Following the structure-affin-
ity results suggesting a differential binding mode for
the malonate derivatives derived from 1b (Boc-Trp-
(NMe)Nle-Asp-Phe-NH2), structural investigations were
conducted on 2b (HO2CCH2CONH-Trp-(NMe)Nle-Asp-
Phe-NH2) to further characterize a putative specificity
of this lead compound. Unfortunately, 2D NMR experi-
ments, performed under the conditions that had given
NOE data for 1b,23 did not give any precise information
about the position of the carboxylate which seems to be
crucial in the affinity of 2b for CCK2 receptors. However,
our results again support the hypothesis of a differential
behavior for 1b and 2b recognition: indeed, the absence
in the NMR spectra of 2b of NOE cross-peaks, which

3622 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20
Bellier et al.
dried over Na2SO4, and evaporated to dryness, yielding 2.66 g
(78%) of a yellow oil, crystallizing to a white solid. 1H NMR
(DMSO + TFA) δ 3.43 (2 H, s, COCH2CO), 5.12 (2 H, s, CH2-
Ph), 7.37 (5 H, m, Ar).
2. Malonic acid monomethyl ester and methylmalonic acid
monoethyl ester were prepared in the same manner from
dimethyl malonate and diethyl methylmalonate using metha-
nol and ethanol as solvents, respectively, instead of benzyl
alcohol.
onate-containing derivatives such as 2b. Structural
hypotheses are described, which suggest how this can
occur. In particular, NMR and computer-aided analysis
support the assumption of a greater flexibility of 2b and
its analogues, maybe authorizing a particular adapta-
tion process to the agonist binding pocket of the CCK2
receptor. Taken together, this yields highly valuable
information about the CCK ligand-receptor interac-
tions.
Thus, the new compounds with dicarbonyl moieties
on the N-terminal of Trp-(NMe)Nle-Asp-Phe-NH2, pre-
sented here, are greatly expected to exert the same
favorable pharmacological effects as BC 264. Their
pharmacological profile is now under study; preliminary
results indicated that compound 2b is devoid of anxio-
genic effects in the (+)-maze and able to increase the
spontaneous alternation of rats in the Y-maze at a dose
of 0.3 mg/kg. These results, analogous to those obtained
with BC 264, suggest that 2b exerts a “CCK2B” phar-
macology. Further studies on this compound and its
analogues are now requested, together with extended
structural investigation, to fully understand this un-
usual behavior.
P r ocedu r e B. Malon ic Acid Mon oam ide, Sodiu m Salt.19b
A total of 7.5 mL (70.5 mmol) of ethyl cyanoacetate was
dissolved in 4.125 mL of ethanol (70.5 mmol, 1 equiv). The
mixture was put under a slow stream of hydrogen chloride
during 16 h, with formation of a white precipitate. This
precipitate was filtered off, washed with ethanol and dried in
a vacuum to yield 4.583 g (33%) of a solid, which was pyrolyzed
at 115 °C in an oil bath. A total of 500 mg of the resulting
product was redissolved in acetone, filtered, and evaporated
to dryness to yield an oil, which crystallized quickly to a beige
solid (284 mg, 85%), identified as malonic acid monoethyl ester,
monoamide. The latter product was saponified with sodium
hydroxide to yield quantitatively the desired malonic acid
monoamide, sodium salt.
P r oced u r e C. Ma lon ic Acid , Su bstitu ted Mon oa m id e.
A solution of malonic acid monobenzyl ester (370 mg, 1.905
mmol), in 5 mL of DMF, was treated at 0 °C with benzylamine
(210 µL, 1.905 mmol, 1 equiv), BOP (1.011 g, 2.286 mmol, 1.2
equiv), and diisopropyl ethylamine (1.17 mL, 6.671 mmol, 3.5
equiv), and the mixture was allowed to stand overnight at room
temperature. After evaporation of the solvents, the residue was
redissolved in ethyl acetate, washed with aqueous citric acid
10%, water, sodium hydrogenocarbonate 10%, water, brine.
The resulting organic phase was dried over sodium sulfate and
evaporated to dryness giving 450 mg (90%) of malonic acid
benzyl ester, benzylamide. The latter was dissolved in 5 mL
of methanol, which was added to a suspension of 100 mg of
Pd 10% on charcoal in 5 mL of methanol. The mixture was
stirred overnight under hydrogen, then filtrated on Celite and
evaporated to dryness, giving 210 mg of the expected substi-
tuted monoamide of malonic acid (76%).
Gen er a l P r oced u r e for th e P r ep a r a tion of P en ta p ep -
tid es fr om I: Ma lon yl-Tr p -(NMe)Nle-Asp -P h e-NH2 (2b).
A total of 750 mg (0.94 mmol) of TFA‚H-Trp-(NMe)Nle-Asp-
(OBzl)Phe-NH2 (I) was dissolved in 15 mL of anhydrous DMF
with 182 mg (0.94 mmol, 1 equiv) of malonic acid, monobenzyl
ester, prepared according to procedure A. To this cooled (0 °C)
stirred solution, were added 457 mg (1.03 mmol, 1.1 equiv) of
BOP and 654 µL (3.76 mmol, 4 equiv) of DIEA.
The mixture was allowed to stand overnight at room
temperature, then concentrated. The residue was redissolved
in 50 mL of ethyl acetate and washed successively with water
(50 mL), 10% aqueous citric acid (3 × 50 mL), water (50 mL),
10% aqueous sodium hydrogenocarbonate (3 × 50 mL), water
(2 × 50 mL), and brine. The resulting organic phase was dried
over sodium sulfate and evaporated to dryness, yielding 811
mg (99%) of the intermediate malonyl-Trp-(NMe)Nle-Asp-
(OBzl)-Phe-NH2.
Exp er im en ta l Section
Ch em istr y. Amino acids and derivatives were purchased
from Bachem (Switzerland), BOP was from Novabiochem, and
other reagents from Aldrich (France) or Acros (Belgium).
Solvents were from SDS (France).
Commercial chemicals were used without further purifica-
tion. Anhydrous solvents were used when necessary after
drying over 4 Å molecular sieves. TLC were performed on
Merck plates precoated with F254 silica gel, 0.2 mm. Plates
were developed with UV, iodine vapor, TFA/ninhydrin, and
Ehrlich’s reagent. Column chromatography was performed
using Merck silica gel (230-400 mesh). Final compounds were
purified by semipreparative HPLC on a Waters apparatus
(column Vydac C18, 300 Å, 10 × 250 mm), using the following
solvent system {A: H2O + 0.1% TFA, B: CH3CN/H2O/TFA 70/
30/0.09 v/v/v}, flow rate: 2 mL/min, detection 210 nm.
The structures of all compounds were confirmed by 1H NMR
spectroscopy in DMSO-d6, with concentrations ranging 3-5
mM. Chemical shifts were measured in ppm with TMS as
internal standard. The purity of the final compounds was
checked by HPLC (Shimadzu apparatus, Vydac C18 column,
300 Å, 4.6 × 150 mm) with the same solvent system as that
for semipreparative purifications, flow rate 1 mL/min, UV
detection 214 nM. Compounds I and II were prepared as
previously described.18
Mass spectra were performed by Quad Service (Poissy,
France) by electrospray techniques using acetonitrile/water/
acetic acid (50/50/0.1, v/v/v) as solvent.
A solution of 800 mg of the latter (0.92 mmol) in 10 mL of
ethanol was added to an activated suspension of Pd/C (55 mg,
60 mg/mmol) in 5 mL of methanol. The mixture was stirred
under an H2 atmosphere during 24 h, then filtrated on Celite
and evaporated. The residue was purified by HPLC and
lyophilized to a white powder.
Abbr evia tion s: Boc, tert-butyloxycarbonyl; BOP, benzot-
riazol-1-yloxy-tris(dimethylamino) phosphonium-hexafluoro-
phosphate; CCK, cholecystokinin; CHO cells, Chinese hamster
ovary cells; DIEA, diisopropyl ethylamine; NOE, nuclear
Overhauser effect; pCCK-8, propionyl-CCK-8, TFA, trifluoro-
acetic acid; Z, benzyloxycarbonyl.
Biologica l Assa ys. Binding experiments and inositol phos-
phate assays were performed under the same conditions as
described previously.16
P r oced u r e A. P r ep a r a tion of th e Mon oester s of Ma l-
on ic Acid . 1. Ma lon ic Acid , Mon oben zyl Ester .19a
A
solution of 5 g of dibenzylmalonate (17.59 mmol) in 50 mL of
benzyl alcohol cooled to 0 °C were treated with 1.01 g of
potassium hydroxide (18.03 mmol, 1.025 equiv) dissolved in
200 mL of benzyl alcohol under vigorous stirring. The reaction
was complete after 2 days, with formation of a white precipi-
tate. This precipitate was filtered off, washed with Et2O, and
redissolved in 100 mL of water. The aqueous solution was
acidified to pH 2 and extracted with 2 × 70 mL of ether. The
resulting organic phases were washed with water and brine,
Molecu la r Design . The calculations were performed on
Silicon Graphics workstations using the Insight/Discover
software package (BIOSYM Technologies Inc., San Diego). The
initial structures were built using standard parameters for
amino acids supplied with the software and were submitted
to a first energy minimization cycle (200 steps using the
steepest descents algorithm followed by 1000 steps using the
conjugate gradient algorithm).

CCK2 Agonists Displaying a Novel Binding Mode
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 20 3623
(12) Charpentier, B.; Durieux, C.; Pe´laprat, D.; Dor, A.; Reibaud, M.;
Blanchard, J . C.; Roques, B. P. Enzyme-resistant CCK analogues
with high affinities for central receptors. Peptides 1988, 9, 835-
841.
(13) Derrien, M.; McCort-Tranchepain, I.; Ducos, B.; Roques, B. P.;
Durieux, C. Heterogeneity of CCK-B receptors involved in
animal models of anxiety. Pharmacol. Biochem. Behav. 1994,
49, 133-141.
The conformational analysis was then performed using high
temperature (1000 K) simulated annealing. The structures
obtained thereafter were refined by using the minimization
protocol described initially. This was repeated 100 times, and
the 100 minimized structures were clustered using the RMS
superposition method, according to conformational similarities.
NMR Stu d ies. NMR experiments were conducted under the
same conditions and with the same solvent systems as
described in ref 23.
(14) Dellu, F.; Mayo, W.; Cherkaoui, J .; LeMoal, M.; Simon, H. A two
trial memory task with automated recording: study in yound
and aged rats. Brain Res. 1992, 588, 132-139.
Ack n ow led gm en t. The authors gratefully acknowl-
edge C. Dupuis for her assistance in drafting this
manuscript, S. De Bouard for large-scale synthesis of
I, G. Pieffet for his excellent technical assistance in
analyzing the molecular modeling data, and M. Vidal
for numerous helpful critical discussions.
(15) (a) Ladurelle, N.; Keller, G.; Blommaert, A.; Roques, B. P.;
Dauge´, V. The CCK-B agonist, BC 264, increases dopamine in
the nucleus accumbens and facilitates motivation and attention
after intraperitoneal injection in rats. Eur. J . Neurosci. 1997,
9, 1804-1814. (b) Le´na, I.; Simon, H.; Roques, B. P.; Dauge´, V.
Opposing effects of two CCK-B agonists on the retrieval phase
of a two-trial memory task after systemic injection in the rat.
Neuropharmacology 1999, 38 (4), 543-553. (c) Ladurelle, N.;
Sebret, A.; Garbay, C.; Roques, B. P.; Dauge´, V. Opposite effects
of CCK-B agonists in grooming behaviour in rats: further
evidence for two CCK-B subsites. Br. J . Pharmacol. 1998, 124,
1091-1098. (d) Le´na, I.; Roques, B. P.; Durieux, C. Dual
modulation of dopamine release from anterior nucleus accum-
bens through cholecystokinin-B receptor subsites. J . Neurochem.
1997, 68, 162-168. (e) Ladurelle, N.; Keller, G.; Roques, B. P.;
Dauge´, V. Effects of CCK-8 and of CCK-B selective agonist BC
264 on extracellular dopamine content in the anterior and
posterior nucleus accumbens: A microdialysis study in freely
moving rats. Brain Res. 1993, 628, 254-262.
(16) Blommaert, A.; Dhotel, H.; Ducos, B.; Durieux, C.; Goudreau,
N.; Bado, A.; Garbay, C.; Roques, B. P. Structure-based design
of new constrained cyclic agonists of the cholecystokinin CCK-B
receptor. J . Med. Chem. 1997, 40, 647-658.
(17) Million, M.-E.; Le´na, I.; DaNascimento, S.; Noble, F.; Dauge´, V.;
Garbay, C.; Roques, B. P. Development of new potent agonists
able to interact with two postulated subsites of the cholecysto-
kinin CCK-B receptor. Lett. Pept. Sci. 1997, 4, 407-410.
(18) Corringer, P. J .; Weng, J . H.; Ducos, B.; Durieux, C.; Boudeau,
P.; Bo¨hme, A.; Roques, B. P. CCK-B agonist or antagonist
activities of structurally hindered and peptidase-resistant Boc-
CCK4 derivatives. J . Med. Chem. 1993, 36, 166-172.
(19) (a) Curtius, T.; Sieber, W. Umwandlung von alkylierten Mal-
onsau¨ren in R-Amino-sau¨ren, II. Mitteilung. Ber. Dtsch. Chem.
Ges. 1922, 55, 1543. (b) Pinner, A. Ueber Imidoa¨ther. Ber. Dtsch.
Chem. Ges. 1895, 28, 473-488. (c) Flavin, M. Preparation of
O-succinyl-L-homoserine. Methods Enzymol. 1971, 17B, 418. (d)
Liwschitz, Y.; Zilkha, A. Syntheses of â-DL-aspartyl peptides with
maleic anhydride. J . Am. Chem. Soc. 1955, 77, 1265-1266.
(20) Ron, D.; Gilon, C.; Hanani, M.; Vromen, A.; Selinger, Z.; Chorev,
M. N-methylated analogues of Ac[Nle28,31]CCK(26-33): Synthesis,
activity and receptor selectivity. J . Med. Chem. 1992, 35, 2806-
2811.
(21) Hruby, V. J .; Fang, S.; Knapp, R.; Kazmierski, W.; Lui, G. K.;
Yamamura, H. I. Cholecystokinin analogues with high affinity
and selectivity for brain membrane receptors. Int. J . Pept.
Protein Res. 1990, 35, 566-573.
(22) (a) Knight, M.; Barone, P.; Tamminga, C. A.; Steardo, L.; Chase,
T. N. Cholecystokinin octapeptide analogues stable to brain
proteolysis. Peptides 1985, 6, 631-634. (b) Holladay, M. W.;
Bennett, M. J .; Tufano, M. D.; Lin, C. W.; Asin, K. E.; Witte, D.
G.; Miller, T. R.; Bianchi, B. R.; Nikkel, A. L.; Bednarz, L.;
Nadzan, A. M. Synthesis and biological activity of CCK hep-
tapeptide analogues. Effects of conformational constraints and
standard modifications on receptor subtype selectivity, functional
activity in vitro and appetite suppression in vivo. J . Med. Chem.
1992, 35, 2919-2928.
Su p p or tin g In for m a tion Ava ila ble: Analytical data of
the final compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
Refer en ces
(1) Reviewed in (a) Noble, F.; Wank, S. A.; Crawley, J . N.; Bradwejn,
J .; Seroogy, K. B.; Hamon, M.; Roques, B. P. Structure, distribu-
tion and functions of CCK receptors using the rational nomen-
clature defined by the International Union of Pharmacology.
Pharmacol. Rev. 1999, 51 (4), 745-781. (b) Crawley, J . N.;
Corwin, R. L. Biological actions of cholecystokinin. Peptides
1994, 15, 731-755.
(2) Reviewed in Van Megen, H.; Westenberg, H.; den Boer, J .; Kahn,
R. Cholecystokinin in anxiety. Eur. Neuropsychopharmacol.
1996, 6, 263-280.
(3) Faris, P. L.; Komisaruk, B. R.; Watkins, L. R.; Mayer, D. J .
Evidence for the neuropeptide cholecystokinin as an antagonist
of opiate analgesia. Science 1983, 219, 310-312.
(4) Itoh, S.; Lal, H. Influences of cholecystokinin and analogues on
memory processes. Drug Dev. Res. 1996, 21, 257-276.
(5) Reviewed in Dauge´, V.; Roques, B. P. Reward versus anxiety:
The opposite pharmacology of opioids and CCK. In cholecysto-
kinin and Anxiety from Neuron to Behaviour, Molecular Biology
Intelligence Unit; Bradwejn, J ., Vasar, E., Eds.; Springer Press
R. G. Landes Co.: Austin, TX, 1995; Chapter 8, pp 151-171.
(6) Crawley, J . N. Subtype-selective cholecystokinin receptor an-
tagonists block cholecystokinin modulation of dopamine-medi-
ated behaviours in the rat mesolimbic pathway. J . Neurosci.
1992, 12, 3380-3391.
(7) Maldonado, R.; Derrien, M.; Noble, F.; Roques, B. P. Association
of the peptidase inhibitor RB 101 and
a selective CCK-B
antagonist strongly enhances antinociceptive responses. Neurol.
Rep. 1993, 4, 947-950.
(8) Rex, A.; Fink, H. Effects of cholecystokinin-receptor agonists on
cortical 5-HT release in guinea pigs on the X-maze. Peptides
1998, 19 (3), 519-526.
(9) (a) Durieux, C.; Coppey, M.; Zajac, J . M.; Roques, B. P. Occur-
rence of two cholecystokinin binding sites in guinea-pig brain
cortex. Biochem. Biophys. Res. Commun. 1986, 137, 1167-1173.
(b) Knapp, R. J .; Vaughn, L. H.; Fang, S.; Bogert, C. L.;
Yamamura, M. S.; Hruby, V. J .; Yamamura, H. I. A new highly
selective CCK-B receptor radioligand ([3H][N-methyl-Nle28,31]-
CCK26-33): Evidence for CCK-B receptor heterogeneity. J . Phar-
macol. Exp. Ther. 1990, 255, 1278-1286. (c) Huang, S.; Fortune,
K. P.; Wank, S. A.; Kopin, A. S.; Gardner, J . D. Multiple affinity
states of different cholecystokinin receptors. J . Biol. Chem. 1994,
269, 26121-26126. (d) Harper, E. A.; Roberts, S. P.; Shankley,
N. R.; Black, J . W. Analysis of variation in L-365,260 competition
curves in radioligand binding assays. Br. J . Pharmacol. 1996,
118, 1717-1726. (e) Bellier, B.; McCort-Tranchepain, I.; Ducos,
B.; Danascimento, S.; Meudal, H.; Noble, F.; Garbay, C.; Roques,
B. P. Synthesis and biological properties of new constrained
CCK-B antagonists: discrimination of two affinity states of the
CCK-B receptor on transfected CHO cells. J . Med. Chem. 1997,
40, 3947-3956.
(23) Goudreau, N.; Weng, J . H.; Roques, B. P. Conformational
analysis of CCK-B agonists using 1H NMR studies and re-
strained molecular dynamics: Comparison of biologically active
Boc-Trp-(NMe)Nle-Asp-Phe-NH2 and inactive Boc-Trp-(NMe)-
Phe-Asp-Phe-NH2. Biopolymers 1994, 34, 155-169.
(24) Harhammer, R.; Scha¨fer, U.; Henklein, P.; Ott, T.; Repke, H.
CCK8-related C-terminal tetrapeptides: Affinities for central
CCK-B and peripheral CCK-A receptors. Eur. J . Pharmacol.
1991, 209, 263-266.
(10) Leinre, S.; Schoneberg, T.; Meyer, R.; Martin, H.; Kaufmann,
R. Pharmacological characterization of CCK-B receptors in
human brain: No evidence for receptor heterogeneity. Neurosci.
Lett. 1996, 217, 45-49.
(11) Charpentier, B.; Dor, A.; Roy, P.; England, P.; Pham, I.; Durieux,
C.; Roques, B. P. Synthesis and binding affinities of cyclic and
related linear analogues of CCK8 selective for central receptors.
J . Med. Chem. 1989, 32, 1184-1190.
J M0000416