Opiate aromatic pharmacophore structure-activity relationships in CTAP analogues determined by topographical bias, two-dimensional NMR, and biological activity assays

Article abstract of DOI:10.1021/jm9900218

Topographically constrained analogues of the highly μ-opioid-receptor- selective antagonist CTAP (H-D-Phe-c[Cys-Tyr-D-Trp-Arg-Thr-Pen]-Thr-NH₂, 1) were prepared by solid-phase peptide synthesis. Replacement of the D-Phe residue with conformatio

Full text of DOI:10.1021/jm9900218

J . Med. Chem. 2000, 43, 569-580
569
Op ia te Ar om a tic P h a r m a cop h or e Str u ctu r e-Activity Rela tion sh ip s in CTAP
An a logu es Deter m in ed by Top ogr a p h ica l Bia s, Tw o-Dim en sion a l NMR, a n d
Biologica l Activity Assa ys
G. Gregg Bonner,† Peg Davis,§ Dagmar Stropova,§ Sidney Edsall,§ Henry I. Yamamura,§ Frank Porreca,§ and
Victor J . Hruby*,‡
Departments of Biochemistry, Chemistry, and Pharmacology, University of Arizona, Tucson, Arizona 85721
Received J anuary 14, 1999
Topographically constrained analogues of the highly µ-opioid-receptor-selective antagonist CTAP
(H-D-Phe-c[Cys-Tyr-D-Trp-Arg-Thr-Pen]-Thr-NH2, 1) were prepared by solid-phase peptide
synthesis. Replacement of the D-Phe residue with conformationally biased â-methyl derivatives
of phenylalanine or tryptophan (2R,3R; 2R,3S; 2S,3R; 2S,3S) yielded peptides that displayed
widely varying types of biological activities. In an effort to correlate the observed biological
1
activities of these analogues with their structures, two-dimensional H NMR and molecular
modeling was performed. Unlike the parent (1), which is essentially a pure µ antagonist with
weak δ agonist activities in the MVD bioassay, the diastereomeric â-MePhe1-containing peptides
exhibited simultaneous δ agonism and µ antagonism by the (2R,3R)-containing isomer 2; µ
antagonism by the (2R,3S)-containing isomer 3; weak µ agonism by the (2S,3R)-containing
isomer 4; and δ agonism by the (2S,3S)-containing isomer 5. Incorporation of â-MeTrp isomers
into position 1 led to peptides that were µ antagonists (2R,3R), 8; (2R,3S), 9, or essentially
inactive (<10%) in the MVD and GPI assays (2S,3R), 10; (2S,3S), 11. Interestingly, in vivo
antinociceptive activity was predicted by neither MVD nor GPI bioactivity. When D-Trp was
incorporated in position 1, the result (7) is a partial, yet relatively potent µ agonist which also
displayed weak δ agonist activity. Molecular modeling based on 2D NMR revealed that low
energy conformers of peptides with similar biological activities had similar aromatic pharma-
cophore orientations and interaromatic distances. Peptides that exhibit µ antagonism have
interaromatic distances of 7.0-7.9 Å and have their amino terminal aromatic moiety pointing
in a direction opposite to the direction that the amino terminus points. Peptides with δ opioid
activity displayed an interaromatic distance of <7 Å and had their amino terminal aromatic
moiety pointing in the same direction as the amino terminus.
In tr od u ction 1
Several observations suggest opiate receptor interplay
in ligand-induced antinociception. Pretreatment with
the non-peptide δ antagonist naltrindole prevented the
development of morphine dependence and tolerance in
mice.2 This suggested that ligands with mixed µ ago-
nist/δ antagonist activity, such as H-Tyr-Tic-Phe-Phe-
NH2 (TIPP-NH2), may be useful drug leads.3 Alterna-
tively, ligands may be potent antinociceptive agents due
to synergistic agonism at both the µ and δ opiate recep-
tor. A conformationally restricted analogue of CTAP (H-
D-Tca-c[Cys-Tyr-D-Trp-Arg-Thr-Pen]-Thr-NH2)4,5 was re-
ported to have a much higher antinociceptive activity
than that predicted by its affinity for opiate receptors
in radioligand competition assay or in the MVD and GPI
assays.6 The antinociceptive activity of this ligand was
blocked by both µ and δ antagonists. Hence, it was
hypothesized that an interaction between µ and δ
receptors was responsible for its antinociceptive potency.
Finally, it may be possible to develop lead analogues
which take advantage of the antinociception resulting
from δ agonism without the side effects associated with
morphine administration due to a simultaneous µ
antagonist activity. Analogues with such simultaneous
µ antagonism/δ agonism have been derived from the µ
antagonist CTAP,7 from the δ agonist Deltorphin,8 and
from DPDPE.9,10
The development of therapeutically useful painkillers
with minimal side effects for severe pain continues to
be a major goal. Opiate agonists continue to be used to
treat pain but suffer from several liabilities including
withdrawal, severe constipation, tolerance, and depen-
dence. Because the affects of opioid ligands may be
differentially mediated by the different opiate receptors
(µ, δ, κ), or by simultaneous action at more than one
opiate receptor type, the development of high affinity
and highly selective ligands remains a major goal.
Furthermore, opiate antagonist administration has been
shown to have an effect in reducing opiate-related side-
effect liability.2 These factors, when taken together,
illustrate the need for selective, high affinity agonists
and antagonists for the opiate receptor types and, where
possible, the development of ligands that possess agonist
activity at one receptor type and antagonist activity at
another. It is hoped that these ligands, through simul-
taneous multiple opiate activity, will lead to therapeuti-
cally useful painkillers devoid of side-effect liabilities.
* To whom correspondence should be addressed. Tel: (520) 621-
6332. Fax: (520) 621-8407. E-mail HRUBY@MAIL.ARIZONA.EDU.
† Department of Biochemistry.
‡ Department of Chemistry.
§ Department of Pharmacology.
10.1021/jm9900218 CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/28/2000

570 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4
Bonner et al.
Regardless of which combination of opiate receptor
activities will ultimately prove useful in humans, it is
still necessary to determine how the structures of the
various opioid ligands lead to their observed pharma-
cological profiles. These sets of structure-activity pro-
files will be essential in predicting which ligands will
yield potent painkilling effects, and which may have
side-effect liabilities, as well as determine which fea-
tures of the ligands must be retained for biological
activity as the molecule is modified to make it more
biologically available. The elements of opioid peptides
hypothesized to be responsible for their activities (the
pharmacophores) are a pair of aromatic amino acid
residuessone in the amino terminal position, and
another in either position three or four. The importance
of these pharmacophores is evidenced by their presence
in casomorphins, dermorphins, dermenkephalins, del-
torphins, enkephalins, dynorphins, endomorphins, and
in somatostatin-derived peptides of the CTAP family as
well as other synthetic peptides such as DAMGO, TIPP,
and ICI-174,864. A free tyramine moiety is almost
universally present in opiate peptides, and a phenol
moiety is present in non-peptide ligands such as the µ
agonist oxymorphone, µ antagonist naltrexone, and the
δ antagonist naltrindole. These facts, taken together
with the fact that CTAP analogues differing only in the
side-chain group stereochemistry of the amino terminal
â-MePhe show varying pharmacological profiles,7 strongly
suggest that the relative orientation and interaromatic
distance between the two pharmacophoric aromatic
amino acid residue side-chain groups are the principle
factors which determine their pharmacological profile.
F igu r e 1. Newman projections of the three low energy
rotamers of (2S,3R)â-MePhe.
previously reported in the literature. Unusual amino
acids with â-methyl substituents have been incorporated
into several classes of opioid peptides including en-
kephalins,9,10,18,19 deltorphins,20 and the δ-selective tet-
rapeptide J OM-13.21 These â-methyl amino acids have
side-chain groups that are biased in favor of one side-
chain rotamer (Figure 1). This rotamer population bias
leads to a less flexible side-chain group and allows for
a more confident assessment of both peptide topography
and analogue conformation-activity relationships.22
This â-methylation yields an amino acid with only one
R proton and one â proton. Consequently, it is possible
to determine which amino acid side-chain rotamer is
most populated by use of COSY NMR spectroscopy and
the Karplus correlation.23 With knowledge of the side-
chain group rotamer position one can determine the
orientation of the aromatic substituent on the â carbon
and its relationship to the rest of the peptide. On the
basis of these studies and biological activity assays, we
will suggest interaromatic distances and relative ori-
entations of the µ and δ opioid receptor aromatic
pharmacophores that lead to the various types of
pharmacological activities displayed by these peptides.
Several research groups have reported investigations
which were intended, at least in part, to determine the
interaromatic distances associated with ligands of vary-
ing biological activity. The results from these studies
have led to different conclusions. One group reported
an interaromatic distance of 6.3 Å for the enkephalin-
derived analogue DPDPE,11 whereas another group12
reported a distance of 4.9-5.6 Å. These values are much
less than the >10 Å,13 11.5 Å,9 or 10.9 Å14 reported
elsewhere. The larger separations are in better agree-
ment with the X-ray crystallographic analyses which
showed the interaromatic distances in DPDPE to be 15.0
Å and 15.9 Å.15 The smaller distances are in better
agreement with rigid ligands such as naltrindole and
SIOM. Similarly conflicting results have been reported
for µ-selective peptides. Interaromatic separations of >9
Å16 and <8 Å17 have been reported.
In the present paper, we describe the syntheses,
biological activity profiles, and structural characteriza-
tion of analogues of the somatostatin-derived µ antago-
nist CTAP which have incorporated in their amino
terminal positions, D-Trp or one of the four isomers of
the rotamer topography-biasing â-methyl amino acids
â-MePhe and â-MeTrp (2R,3R; 2R,3S; 2S,3R; 2S,3S).
In particular, the work describes the synthesis and
biological activity determinations of five new analogues
(2, 3, 4, 5, and 7). Two-dimensional NMR studies of
these five analogues, as well as of four other previously
reported7 analogues (8, 9, 10, 11), were performed.
These data and NMR-derived computer-aided molecular
modeling were used to form a comprehensive structure-
activity analysis for all the analogues including others
Resu lts a n d Discu ssion
Str u ctu r e-Activity Rela tion sh ip s. As previously
reported,4,5 the cyclic parent peptide CTAP (1) displays
high affinity for µ opioid receptors and very low affinity
for δ receptors and is, therefore, quite selective. This
peptide acts predominantly as a µ antagonist in vitro
and in vivo. The replacement of the D-Phe residue in
the amino terminal position with a â-MePhe residue of
D configuration leads to analogues 2 ([(2R,3R)â-MePhe1]-
CTAP) and 3 ([(2R,3S)â-MePhe1]CTAP). Both of these
analogues are lower in affinity versus the µ receptor by
about an order of magnitude, but with 4-fold to 5-fold
higher affinity at the δ receptor than 1 (Table 1).
Consequently, they are much less selective than the
parent peptide 1. Peptides 2 and 3 also differ from 1 in
their in vitro biological activity profiles (Table 2) in that
they are less than half as effective in shifting the dose-
response curve of the known µ agonist PL-017. More-
over, both 2 and 3 differ from 1 in their MVD agonism.
Analogue 2 has about twice the MVD agonism of 1,
whereas 3 has only about half the MVD agonism of 1
under the same assay conditions. Furthermore, 1 dis-
plays slight GPI agonism, whereas both 2 and 3 do not.
Analogue 2 also displays twice as much antinociception
(36%) in the tail-flick assay as 3 (Table 3). It should be
noted that administration of these analogues led to
behavior which indicates toxicity in several cases

Opiate Aromatic Pharmacophore SAR in CTAP Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4 571
data. Analogues which contain an amino acid of
Ta ble 1. Rat Brain Opiate Receptor Binding Affinities of
CTAP Analogues
D
configuration retain high affinity for the µ receptor.
When D-Trp is substituted in position 1, the result is
peptide 7 with opioid receptor binding characteristics
similar to 1 (Table 1). This peptide, however, acts as a
fairly potent, yet partial µ agonist in the GPI assay
(Table 2) and is the most potent µ agonist CTAP
analogue yet synthesized based on the GPI biological
activity assay. It displays a 43% reduction in contraction
at 1 µM concentration and an EC50 of 37 nM (Table 2).
no.
analogue
IC50µ (nM)a
2.1 ((0.3)
IC50δ (nM)b
δ/µ
1
2
3
4
5
6
7
8
9
CTAPc
5310((280)
1190 ((330)
1320 ((240)
2530
59
63
[(2R,3R)â-MePhe1] CTAPd 20 ((5)
[(2R,3S)â-MePhe1] CTAPd 21 ((3)
[(2S,3R)â-MePhe1] CTAPd 5420 ((1020) 13200 ((2470) 2.43
[(2S,3S)â-MePhe1] CTAPd 1500 ((670) >15000
>10
1060
[D-Tic1]CTAPe
[D-Trp1]CTAP
1.2 ((0.2)
1.39 ((0.37) 3530 ((1440) 2540
1270 ((78)
[(2R,3R)â-MeTrp1] CTAP 18.7 ((4.1)
[(2R,3S)â-MeTrp1] CTAPe 12.1 ((3.8)
3640 ((795)
7480 ((1960) 618
195
10 [(2S,3R)â-MeTrp1] CTAPe 3400 ((400) 7640 ((4800) 2.2
When conformationally constrained tryptophan ana-
logues are substituted for D-Trp in the N-terminal
position, some of the characteristics of the analogues
change in a manner similar to those seen upon â-methyl
substitution of phenylalanine, but others are unlike
those seen in the phenylalanine derivatives. When D-Trp
is substituted with â-methyltryptophans of D configu-
ration (8 and 9, Table 1), the resulting analogues have
their µ receptor affinity reduced about 10-fold (Table 1).
This also is what is seen for the â-MePhe analogues.
Unlike the â-MePhe analogues, the â-MeTrp analogues
of D configuration do not have increased affinity for the
δ receptor compared to CTAP, and as a result they are
much more µ selective than their â-MePhe counterparts.
When a â-MeTrp residue of L configuration is incorpo-
rated, peptides 10 and 11 are obtained, which have
reduced affinities for both µ and δ receptors. Further-
more, the reductions in affinity for each receptor are
very similar to the â-MePhe-containing peptides. These
general similarities in receptor affinities, however, were
not predictive of biological activities. The analogues with
amino acids of D configuration did act as potent µ
antagonists, with 9 showing the greatest GPI antagonist
activity of any CTAP analogue reported thus far (Table
2). The µ antagonism of these analogues is somewhat
surprising considering that the unbiased analogue 7
exhibits no µ antagonism but instead is a potent partial
agonist at the µ receptor (Table 2). The analogues with
L amino acids in position 1 are not potent µ antagonists,
as is expected from their low affinity for the µ receptor.
In this regard, analogues 10 and 11 are similar to 4 and
5. When considering the agonist properties of the
â-MeTrp-containing analogues, the outcome is inactiv-
ity. Only one â-MeTrp-containing analogue, 8, shows
even slight agonism at either receptor (Table 2), with
agonism at the δ receptor only. It is arguable that the
â-MeTrp-containing analogues are not appreciably dif-
ferent from their â-MePhe-containing counterparts. In
two cases, the agonism is slight (δ activity of 3 and µ
activity of 4), and therefore a small total reduction
would be viewed as inactivity (in â-MeTrp analogues 9
and 10), and in another (δ activity of 2) a small total
reduction in agonism is still observable in the â-MeTrp
analogue 8. However, there is one clear difference in
biological activity seen between the two classes (â-
MeTrp1 vs â-MePhe1) even when the stereochemistry
of the N-terminal amino acid is the same. Analogue 5
is a δ agonist of modest potency. Even considering an
argument of generally reduced potency in â-MeTrp-
containing peptides, one would still expect to see some
activity in analogue 11. This activity was not observed.
When the data are taken as a whole, it is clear that the
nature of the side-chain group moiety in position 1 is
important in determining the binding affinity and
11 [(2S,3S)â-MeTrp1] CTAPe 1360 ((160) 9180 ((1600) 7.0
12 [D-Tca1] CTAPe
173 ((135)
211 ((65.5)
1.2
a
b
Versus [3H]DAMGO. Versus [3H][D-Pen2,4′-Cl-Phe4,D-Pen5]-
enkephalin. c Data taken from ref 4. Data taken from ref 7.
d
e Data taken from ref 6.
(Table 3). There was a general occurrence of seizure,
barrel rolls, circling, and in some cases, death. The
frequency, severity, and duration of these phenomena
were dose-dependent. At lower doses (6 nmol), tail-flick
latency was not observable, and at higher doses (>30
nmol) toxicity-related behavior made testing of tail-flick
latency impractical. So these data are presented with
the caveat that loss of motor coordination may be
erroneously ascribed to a change in tail-flick latency.
Substitutions with L-amino acids in the N-terminal
position (i.e., [(2S,3R)â-MePhe1]CTAP, 4, and [(2S,3S)â-
MePhe1]CTAP, 5) lead to analogues with dramatically
reduced affinities for both the µ and δ receptor. This
substitution evidently does not result in analogues that
are inactive in the MVD and GPI assays, however.
Analogue 5 exhibits MVD agonism to the same degree
as 2, but unlike 2, does not act as a GPI antagonist
(Table 2). Analogue 4 is the only â-MePhe analogue
which acts essentially only as a µ agonist (Table 2),
although a weak one. Analogue 5 has a similar anti-
nociceptive potency as 2 as well as a similar activity in
the MVD which suggests that their antinociceptive
activities are due to their agonism at δ receptors despite
low affinity for the δ receptor. Analogue 6 is in all
respects very similar to the parent peptide. This peptide
has a high affinity for the µ receptor and is very µ opiate
receptor-selective (Table 1). This analogue also shifts
the dose-response curve of PL-017 in the GPI assay
(Table 2), but not that of DPDPE in the MVD assay
(data not shown), and has little antinociceptive activity
in the tail-flick assay (data not shown). The â-MePhe1
analogues in this series lead to a variety of biological
activities. Depending on which diastereoisomer is in-
corporated in the N-terminal position, the resultant
analogue is a weak µ agonist (4), or a δ agonist (5), or
a µ antagonist (3), or a simultaneous δ agonist/µ
antagonist (2). These widely varying types of biological
activities observed underscore the importance of the
topographical arrangement of the aromatic pharma-
cophores in these analogues. Because the backbone is
a stable template,5 these various activities apparently
result from differential side-chain group orientations.
CTAP analogues with D-Trp and â-methyl-tryptophan
derivatives in the N-terminal position also were syn-
thesized and characterized in vitro and in vivo (Tables
1 and 2). Some of their characteristics were strikingly
similar to their Phe1 counterparts. The most obvious
case of this is seen in the µ receptor binding affinity

572 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4
Bonner et al.
Ta ble 2. Opioid Agonist and Antagonist Activities of CTAP Analogues
agonist activity at 1 µM
(%) inhibitiona
antagonismb
no.
analogue
MVD
17
34
8
0
35
10 ((2)
13.6 ((3)
3.7
0
0
0
25
GPI
PL-017 shift
1
2
3
4
5
6
7
8
9
CTAPc
4
0
0
13
0
55
19
[(2R,3R)â-MePhe1] CTAPd
[(2R,3S)â-MePhe1] CTAPd
[(2S,3R)â-MePhe1] CTAPd
[(2S,3S)â-MePhe1] CTAPd
[D-Tic1]CTAPc
15
1e
1
1
60 ((15)
[D-Trp1]CTAP
43f
0
1
39
68
1
2.6
1.2
[(2R,3R)â-MeTrp1] CTAP
[(2R,3S)â-MeTrp1] CTAP
[(2S,3R)â-MeTrp1] CTAP
[(2S,3S)â-MeTrp1] CTAP
[D-Tca1]CTAPe
0
0
0
18
10
11
12
a
b
Reported as the percent reduction in electrically evoked twitch height at a 1 µM concentration of the test peptide analogue. Reported
as the rightward fold-shift of the dose-response curve of the known agonist PL-017 (GPI, µ) at a test peptide analogue concentration of
1 µM. c Data taken from ref 4. Data taken from ref 7. e A PL-017 shift of 1-fold indicates no change in the dose-response curve. f Partial
d
g
agonism (Emax ) 43%; EC50 ) 37 ((10) nM). Data taken from ref 6.
Ta ble 3. Percent Tail-Flick Latency Increases of CTAP Analoguesc
minutes after injection
no.
analogue
10
20
30
45
60
2
3
4
5
7
8
9
10
11
12
[(2R,3R)â-MePhe1] CTAPb
[(2R,3S)â-MePhe1] CTAPb
[(2S,3R)â-MePhe1] CTAPb
[(2S,3S)â-MePhe1] CTAPb
[D-Trp1]CTAP
4.3 ((1.5)
5.6 ((2.3)
6.1 ((2.9)
7.6 ((3.0)
20.4 ((5.8)
30.6 ((3.6)
2.2 ((1.4)
4.0 ((2.2)
45.9 ((5.8)
22
27.6 ((5.5)
11.1 ((3.4)
13.9 ((3.1)
10.3 ((1.8)
42.4 ((6.9)
41.2 ((5.9)
-0.6 ((0.4)
5.3 ((2.3)
39.8 ((6.1)
57.2
36.1 ((5.1)
17.5 ((8.2)
20.2 ((6.6)
34.9 ((10.2)
22.9 ((4.1)
NDc
3.4 ((1.1)
ND
ND
72
24.4 ((4.2)
5.0 ((2.6)
10.8 ((7.0)
36.1 ((8.0)
25.1 ((4.1)
ND
ND
ND
ND
65
5.8 ((4.2)
1.2 ((2.5)
4.0 ((4.6)
22.8 ((6.7)
17.8 ((3.7)
ND
ND
ND
ND
68
[(2R,3R)â-MeTrp1] CTAP
[(2R,3S)â-MeTrp1] CTAP
[(2S,3R)â-MeTrp1] CTAP
[(2S,3S)â-MeTrp1] CTAP
[D-Tca1]CTAPd
a
b
Data are presented as the percent increase in tail-flick latency. See METHODS for the equation used. Data taken from ref 7. c ND
d
indicates that tail-flick latency could not be determined due to evidence of toxicity-related behavior. Data taken from ref 6.
biological activity of the peptide. In the case of amino
acids with no conformational constraint, the receptor
affinities are similar, but their activities vary widely.
Analogue 1 acts predominantly as a µ antagonist,
whereas 7 acts as a potent but partial µ agonist. In the
case of peptides with a conformational bias, 5 acts as a
δ agonist, whereas 11 is inactive. On the other hand,
when the side-chain group is cyclized to confer near
rigidity, the results are similar. Analogue 6 exhibits
activity very similar to its linear analogue 1sgenerally
µ antagonism. Similarly, the side-chain cyclized amino
acid-containing peptide 12 exhibits the same type of
activity as its linear analogue 7, namely, simultaneous
weak agonism at both the µ and δ receptors.
Because these analogues are not particularly potent
agonists, they do not necessarily provide particular
insight for the topographical features required for opiate
receptor activation. Nonetheless, these peptides do
demonstrate that subtle changes in structure can have
substantial effects on the type of activity displayed by
the peptides (e.g., agonism vs antagonism). The impor-
tant finding from this study is not that the peptides are
potent, but rather that they show widely varying types
of affinity and biological activity that are dictated by
the nature of a conservative topographical structural
modification. The parent peptide, CTAP, shows no µ
agonism and little δ agonism, whereas the stereospecific
addition of a single methyl group leads to peptides with
improved agonism for µ (4, 7) or δ (2, 5) receptors. By
understanding how these features lead to improved
agonist activity, one can design new analogues which
could be more capable of activating the opioid receptors.
Con for m a tion a l An a lysis by Tw o-Dim en sion a l
NMR Sp ectr oscop y a n d Molecu la r Mod elin g. The
conformations of analogues 2-5 and 7-11 were studied
by NMR spectroscopy. Results of these NMR experi-
ments and those performed previously on CTAP ana-
logues5,24 served for NMR-based molecular modeling of
all 12 analogues. Complete assignment of all proton
resonances was achieved using TOCSY, ROESY, and
COSY experiments (Supporting Information). The chemi-
cal shift values are in good agreement with those
reported previously for other CTAP analogues.24 Proton
chemical shifts also are in good agreement with values
predicted for them based on published “typical” values
for each residue in peptides.25 Large deviations from
typical amino acid residue proton chemical shift values
are taken as evidence for atypical magnetic environ-
ments for them and as evidence for the presence of some
chemical shift-inducing moiety, such as ring current
effects from a neighboring aromatic moiety. The two-
dimensional NMR experiments showed that these pep-
tides adopt the same peptide backbone conformation as
previously determined for other peptides in this se-
ries.5,24 An upfield chemical shift was seen for the Câ
and Cδ protons of Arg5. This fact, when taken with
ROESY cross-peaks identifying through-space interac-
tions of D-Trp4 and Arg5 residues, led to confirmation
of the presence of a âII′-type turn with D-Trp4 in the
i+1 and Arg5 in the i+2 “corner” positions. It is common

Opiate Aromatic Pharmacophore SAR in CTAP Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4 573
Ta ble 4. Final Energy Miminization Values for CTAP Analogues (kJ /mol)a
1
2
3
4
5
6
7
8
9
10
11
12
-350.31 (10) -297.50(2) -302.60(2) -292.24 -283.53
-311.24
-296.97
-295.16
-327.39
-311.21
-310.58
-282.66 -297.93 -318.84(4) -288.65 -338.86
-340.00
-339.95
-337.72(4)
-337.34(2)
-336.11
-336.04
-332.36
-332.12
-330.96
-326.45
-324.66
-318.92
-296.43
-301.57
-284.57 -283.40
-275.36 -295.77 -311.89
-271.78 -292.33 -309.93
-271.59 -291.06 -308.98
-270.75 -290.56 -306.85
-270.54 -289.69 -306.02
-284.30 -330.01
-283.27 -328.53
-281.86 -323.10
-281.86 -320.69(2)
-281.31 -315.73(2)
-293.71
-300.09(2) -280.70 -282.95
-292.52(2) -299.81
-278.35 -279.67(2) -293.43(4) -309.03
-276.46 -278.37(3) -292.55 -306.04
-276.37 -276.49
-275.77 -276.21
-291.52
-290.71
-288.84(2) -293.13
-287.98
-287.56(2) -287.21
-286.54(2) -286.33
-285.94
-285.21
-283.46
-283.05
-281.38
-281.27
-280.41
-278.44
-275.69
-275.47
-271.51
-297.81
-295.67
-291.54(2) -306.03
-290.86(2) -303.64
-289.39
-287.88
-270.45 -284.76 -305.14(5) -281.01 -311.42
-290.04(3) -275.31 -275.98
-302.81
-302.58
-296.19
-294.77
-293.32
-292.97
-292.64
-290.80
-290.44
-269.77 -281.16 -302.69
-269.73 -274.86 -302.59
-269.72 -274.15 -301.54
-268.91 -273.78 -301.24
-268.61 -273.66 -300.95
-280.30 -311.23
-279.94 -309.63
-279.91 -309.13(4)
-279.89 -308.55
-279.64 -304.55
-274.44 -275.60
-273.66 -275.49(3) -284.97
-285.68(2) -273.57 -274.93(2) -282.85
-284.81
-270.76 -274.87
-276.87
-274.26
-271.17
-265.02
-282.05(3) -270.73 -274.86
-267.69 -273.45 -300.70(4) -278.04 -304.40(2)
-281.81
-278.81
-276.44
-261.45
-270.64 -271.95
-270.41 -271.45
-270.33 -270.69
-270.30 -268.23(2)
-268.42 -265.40
-268.29 -255.88
-267.82
-267.40 -273.29 -300.67
-266.70 -273.17 -299.61
-265.73 -271.52 -278.24
-277.17 -304.05
-275.84 -299.89
-264.77 -289.90
-297.27
-289.68(2) -260.37 -271.42
-289.51
-279.27
-277.69
-257.68 -265.44
-254.80 -263.85
-250.46 -263.71
-247.69 -258.41
-236.89 -248.84
-290.02
-290.02
-267.67
-265.66
-261.37
-260.69
-259.18
a
Numbers in parentheses represent the number of conformers that converged during minimization to the same structure.
for peptides to fold into a conformation where a D-amino
acid residue adopts the i+1 position of a â-turn. Previ-
ous analyses using NMR and molecular dynamics
simulations demonstrated that the CTAP peptide back-
bone is quite stable.5,24 Therefore, a conservative amino
acid side-chain group substitution (i.e., addition of a
â-methyl group) can be made to an exocyclic amino acid
residue which influences the topography of the peptide
without drastically altering its peptide backbone con-
formation. Thus it is not surprising that the disulfide
cyclic conformation remained the same in all of the
analogues examined. Initial analogue conformations
were based on these previous NMR and molecular
dynamics simulations and were modified to reflect
Arg5 CR proton and the D-Trp4 H4 proton. Therefore,
the D-Trp4 side-chain group points away from the amino
terminus and not in the same direction as the D-Phe1
and Tyr3 side-chain groups as suggested earlier. How-
ever, there was not an abundance of informative ROE
cross-peaks. Typical ROE cross-peak patterns were
evident in each of the analogues and were dominated
by intraresidual cross-peaks and cross-peaks of the type
CRH (i) f NH (i+1) and CâH (i)f NH (i+1). The
presence of a âII′ turn was also confirmed by the
presence of an ROE cross-peak between the Arg5 amide
proton and the Thr6 amide proton. Finally, several ROE
cross-peaks were observed for interactions between Câ
protons and intraresidual (H4 and H7) tryptophyl
protons and the H2 protons of tyrosyl and phenylalanyl
residues (see Supporting Information).
3
individual atypical proton chemical shift values and J
3
HR-HN and J HR-Hâ coupling constants determined
from DQF-COSY and COSY-35 experiments, respec-
tively.
The other difference between the currently proposed
conformation and that proposed by Kazmierski et al.24
deals with the carboxy terminus. Previous analysis
indicated that the carboxy terminus is near the amino-
terminal portion of the peptide. The current analysis
suggests that the carboxy terminus is more closely
associated with the positively charged Arg5 side-chain
group. These modifications of the previously suggested
conformation for 1 led to a conformation for CTAP that
was lower in energy than the conformation previously
suggested.
CTAP analogue 2 is interesting in that it is a potent
µ antagonist with some δ agonist activity. This peptide
exhibits its multiple opioid activity by virtue of its ability
to adopt multiple low energy conformations (Figure 2).
One low energy conformation resembles the stereotypi-
cal µ antagonist conformationsone in which the aro-
matic group of the amino-terminal residue is over the
Cys2 â protons. In this case the interaromatic distance
is 7.5 Å. In another low energy conformation this residue
is closer to the Tyr3 side-chain group. Here the two
aromatic moieties are separated by <7.0 Å (Figure 2).
Analogues of CTAP which have low energy conforma-
tions with this interaromatic distance demonstrate δ
agonism as a rule. Analogue 3 is very similar to 2 in
Analysis of the parent peptide 1 by molecular dynam-
ics using NMR-derived restraints led to the identifica-
tion of 13 distinct low energy conformations within 50
kJ /mol of the lowest energy conformer (see Table 4 for
final energy values for each of the new analogues).
While the conformations are folded slightly differently,
each forms an arrangement where the D-Phe1 f Tyr3
interaromatic distance is approximately 7.4 Å. In this
arrangement, the aromatic ring of the D-Phe1 side-chain
group points over the Cys2 â carbon. This feature is
present in all CTAP analogues reported here that act
as potent µ antagonists, and thus this is suggested to
be in the “stereotypical” µ antagonist conformation. The
conformation identified for 1 agrees in large part with
the conformation proposed earlier.5,24 A few differences
have been identified, however. In the current study, it
is clear that the amino acid side-chain groups of D-Trp4
and Arg5 are closer than that proposed by Kazmierski
et al.24 The affect of the D-Trp4 aromatic ring is evident
by the upfield shifting of the Arg5 â and γ protons in
comparison with typical values. Furthermore, ROE
cross-peaks between the Arg5 CR proton and the Trp4
H4 proton demonstrate a close proximity between the

574 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4
Bonner et al.
F igu r e 2. Stereoscopic view of three low energy conformers of 2. The conformations shown are responsible for µ antagonism, δ
agonism, and inactivity (see text).
structure. They differ only in the stereochemical ar-
rangement of a â-methyl group in the amino terminal
residue. This small change leads to a significant differ-
ence in biological activity. Unlike 2, 3 is essentially a
pure antagonist (<10% agonist inhibition at 1 µM) with
very weak (8%, Table 2) agonist activity. In this
analogue, the interaromatic distance is >7 Å for both
of its low energy conformationssone of which is the
typical µ antagonist arrangement. This suggests that
the second low energy conformation is inactive and
identifies the active conformation for δ agonism in
analogue 2. This means that the low energy conforma-
tion of 2 in which the phenyl ring of the amino terminal
residue is pointing in the same direction as the Tyr3
side-chain group phenol moiety is the conformation
responsible for δ agonism. Analogue 4 is the only
â-MePhe-containing peptide without both δ agonist
activity and antagonist activity at the µ receptor (Table
2). This bioactivity can be explained by its low energy
conformers. The amino terminal phenyl moiety is not
positioned over the Cys2 â protons, but rather is over
the Cys2-Tyr3 peptide bond in two of the three low
energy conformations. In the third conformation, this
aromatic moiety is in the region normally occupied by
the Tyr3 side-chain group. In this case, the Tyr3 side-
chain group adopts a different rotamer position, pointing
generally toward the Arg5 side-chain, and not toward
the amino terminus, as is the case in the other ana-
logues in this series. The energy minimized structure
of analogue 5 has both position 1 and position 3
aromatic rings pointing toward the amino terminus.
This arrangement is similar to that seen for one of the
low energy conformations of analogue 2, and it is
similarly a δ agonist. In contrast with analogue 2,
however, the interaromatic distance in this peptide is
4.6 Å compared to a nearly 7 Å separation in 2. While
analogue 5 has relatively high agonist activity in the
MVD, it has low affinity for the δ receptor. This may
be due to a high efficacy of the analogue. Alternatively,
it may be due to the activation of other pathways, the
components being in the MVD but not in the purified δ
receptor homogenate or which exist in different abun-
dances, which lead to the observed effects. Other
functional studies, such as the GTPγ35S binding assay,
might be able to distinguish between these possibilities.
When taken together, these four isomers of [â-MePhe1]-
CTAP reveal the criteria for opiate receptor activation
in the CTAP series. It is clear that the µ antagonist
activity is a result of a Tyr3 side chain pointing toward
the amino terminus and a â-MePhe1 side-chain group
pointing over the Cys2 â protons and an interaromatic
distance of g7 Å. The δ agonist conformation consists
of both â-MePhe1 and Tyr3 side-chain aromatic moieties
pointing toward the amino terminus. In this case a wide
range of interaromatic distances (4.6-7 Å) will suffice.
This near overlap in interaromatic distances for µ
antagonism and δ agonism (around 7 Å) may be
responsible for the creation of multiply biologically
active analogues such as 2. These interaromatic dis-
tances identified as typical for µ antagonism are in
excellent agreement with distances determined for
peptides in the Dermorphin series (7.9-8.0 Å).17 This
7.9 Å separation is seen in each of the low energy
conformations of 6, which adopts the typical conforma-
tion for µ antagonism and acts essentially as a pure µ
antagonist (e10% twitch-height reduction at 1 µM) with
very weak if any δ agonist activity. In regards to µ opiate
receptor affinity, selectivity, and µ opiate receptor
antagonist potency, 6 is one of the best analogues
characterized to date.
Analogue 7 is unique among the reported CTAP series
peptides. This peptide is the only known high affinity
partial agonist of the µ opiate receptor. Unlike the
â-MePhe-containing peptides, this analogue is not con-
formationally biased toward one rotamer position for the
amino terminal side-chain group. This lack of confor-
mational constraint is seen in its 7.5 Hz 3J HR-Hâ
coupling constant, which indicates that this side-chain
group is freely rotating about its R-â bond. This is
unlike the 3J HR-Hâ coupling constants of amino
terminal â-methyl amino acids, which are greater than
11 Hz. Due to the flexibility of this analogue, it is
difficult to ascribe a particular conformation to this
peptide or to gauge accurately which conformation leads
to µ agonism. However, the low energy conformation
suggests that µ agonism may be the result of an
arrangement in which both the D-Trp1 and Tyr3 side-
chain group aromatic moieties point away from the
amino terminus.

Opiate Aromatic Pharmacophore SAR in CTAP Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4 575
F igu r e 3. Stereoscopic view of the low energy conformation of 9. This conformation shows the stereotypical arrangement of the
aromatic pharmacophores that leads to µ antagonism.
The â-MeTrp-containing analogue 8 is very similar
to its â-MePhe-containing counterpart 2. They are both
potent µ antagonists and exhibit measurable δ agonism.
Like 2, 8 adopts the stereotypical µ antagonist arrange-
ment and has a >7 Å interaromatic separation (7.7 Å).
Unlike 2, however, the second low energy conformation
of 8 does not have a <7 Å interaromatic distance (12.3
Å). This suggests that there may be more than one
arrangement of the aromatic moieties of the first and
third amino acid residue that is able to activate the δ
receptor. In contrast, 9 is a pure µ antagonist, similar
to its â-MePhe counterpart 3, although 3 has some slight
δ agonist activity. This peptide adopts the characteristic
µ antagonist conformation (Figure 3). This conformation
is especially stable in this analogue, as evidenced by the
dramatic upfield shift of the Cys2 â protons. This peptide
is the most potent µ antagonist synthesized to date, as
determined by its ability to shift rightwardly the dose-
response curve of the known µ agonist PL-017. Energy
minimization of this peptide also offered evidence that
the NMR constraints used in refining the structure were
not changing an otherwise unsatisfactory initial con-
formation into an artificially low local energy minima.
The structure obtained after energy minimization using
NMR constraints converged with the energy minimized
structure predicted prior to the NMR experiments. The
interaromatic separation for this analogue is 7.3 Å.
Analogues 10 and 11 are both essentially inactive at
both the µ and δ opiate receptors, although 11 has some
slight µ antagonist activity. Both analogues adopt low
energy structures with aromatic separation and relative
orientation similar to the conformations suggested to
be responsible for the δ agonism seen in analogues 2
and 5. The difference between 10 and 11, when com-
pared to 2 and 5, is that the amino terminus is on the
other face of the peptide ring structure. It is most likely
that the unfavorable interaction between this terminal
amine group and the δ receptor leads to these analogues’
inactivity. Analogues 9, 10, and 11 also are interesting
in that they exhibit no δ agonism, yet have some (albeit
micromolar) affinity for the δ receptor. This may be
explained by very low efficacy. Testing these analogues
at even higher concentrations may reveal some very
weak agonist activity. Analogue 12 has a cyclized amino
acid in position 1, and hence it is unable to rotate freely
around its R-â bond. This analogue is unusual in that
it has similar weak binding affinity to µ and δ receptors
as well as weak agonism of both receptors. The low
energy conformation of this analogue explains its δ
agonism. This peptide adopts a conformation similar to
the one suggested as being responsible for the δ agonism
of analogues 2 and 5. The µ agonism of this peptide is
surprising given its low energy conformation. However,
a loss of interaction of the Tyr3 side-chain group
hydroxyl with the amino terminus would convert the
conformation to the µ agonist conformation suggested
for 7. The side-chain cyclization in essence stabilizes the
δ agonist topography at the expense of µ agonist
arrangement (compare µ and δ agonist activities of 12
to 7). In the low energy conformation, the interaromatic
distance is 6.3 Å.
These 12 CTAP analogues show a remarkable variety
of types of biological activities including partial µ
agonism (7), µ antagonism (1, 3, 6, 8, 9), simultaneous
µ and δ agonism (12), simultaneous µ antagonism and
δ agonism (2), δ agonism (5), and inactivity (4, 10, 11).
By comparing and contrasting the predicted low energy
structures and the observed biological activities, it was
possible to derive common characteristics responsible
for agonist activity at the δ and µ receptor and antago-
nist activity at the µ receptor.
Con clu sion s
Incorporation of aromatic â-methyl-containing amino
acids into the amino terminal position of CTAP led to
analogues of widely varying types of biological activity
profiles. It is clear that an amino terminal residue of D
configuration is a requirement for high affinity binding
of the µ opiate receptor. Opiate receptor activation does
not have this requirement. For µ opiate receptor activa-
tion, as evidenced by GPI agonism, analogue 4 was
shown to have modest agonism, despite its >5 µM IC50
for the receptor. The results are more striking when δ
agonism, as seen in the MVD assay, is surveyed. The
best example of this is analogue 5, which reduces MVD
twitch height by nearly 35%, despite a 15 µM IC50
binding affinity for the receptor. The nature of the
aromatic group in the amino terminal position (phenyl
vs indole) is very important for receptor activation.
Analogues 1 and 7 have practically identical receptor
binding affinities and similar MVD (δ) agonism. They
differ dramatically in their ability to activate the µ
receptor. At a 37 nM concentration, the D-Trp-containing
peptide 7 exhibits >20% GPI twitch-height reduction,

576 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4
Bonner et al.
ods.26 The solid support resin (1% DVB cross-linked pMBHA)
and l-Pen(S-4′-methylbenzyl) were purchased from Peptides
International (Louisville, KY). NR-Boc-protected D-Trp, D-Phe,
L-Cys(S-4′-methylbenzyl), L-Tyr (O-2′,6′-dichlorobenzyl), l-Arg-
(Nγ-tosyl), and L-Thr(O-benzyl) were purchased from Bachem
(Torrance, CA). Protected â-MeTrp and â-MePhe analogues
suitable for solid-phase peptide synthesis were made according
to procedures developed in this laboratory.27-29 Tic and Tca
were synthesized by published procedures.30 These cyclized
amino acids, â-methyl amino acids, and L-Pen(S-4′-methyl-
benzyl) were NR-t-Boc protected at their R-amine group by
literature methods.31
whereas 1 is essentially inactive at 1 µM concentration.
Another example includes a comparison of analogues 5
and 11. Whereas 5 is the most potent δ agonist known
in this series, 11 exhibits no detectable δ agonism up
to a 1 µM concentration. Finally, a comparison of the
cyclic amino acids R (D-Tic1 and D-Tca1) reveals a cyclic
phenylalanine (D-Tic1)-containing peptide 6, which has
a little δ agonism and barely detectable µ agonism,
whereas the cyclic tryptophan-containing (D-Tca1) pep-
tide 12 is, in the CTAP series, the most potent simul-
taneous δ agonist/µ agonist known. Taken as two groups
(peptides with â-MePhe in position 1 and peptides with
â-MeTrp in position 1), the differences in activities
become more obviousspeptides with â-MePhe include
a simultaneous δ agonist/µ antagonist (2), a µ antagonist
(3), a µ agonist (4), and a δ agonist (5), whereas peptides
containing â-MeTrp display essentially only µ antago-
nism (8, 9) or lack of activity (10, 11).
The conformational characterization of these peptides
was carried out in an effort to determine how peptides
that are so similar in structure can elicit such varying
biological responses. It was shown that the incorpora-
tion of the â-methyl group does have the intended affect.
Namely, this methyl group, when placed on the â
carbon, will bias the rotamer population through steric
effects toward a particular rotamer. This is evidenced
by the high 3J HR-Hâ coupling constants (>11 Hz)
determined in the COSY-35 experiments. On the other
hand, the coupling constant of the unbiased amino acid
D-Trp1 in analogue 7 was much lower (7.5 Hz), and
indicated that it experienced free rotation about its R-â
bond. The preferred conformation stabilizes certain
interaromatic distances. By inspection of molecules with
certain shared biological activities (e.g., δ agonism),
suggestions can be made as to the structural features
that lead to the activity. On the basis of the low energy
conformations of these analogues and their known
biological activities, aromatic pharmacophore arrange-
ments are discovered for opiate receptor activity. Acti-
vation of the δ opiate receptor requires aromatic groups
that are relatively close together (<7 Å) and both point
in the same direction as the amino terminus. Antago-
nism of the µ receptor occurs when the aromatic groups
are separated by >7 Å and the aromatic side-chain
group of the amino terminal residue is pointing away
from the amino terminus and toward the cysteine in the
second position. Because there is a near overlap of the
interaromatic distances around 7 Å and because some
analogues adopt multiple low energy conformations, it
is possible that some analogues will exhibit multiple
activities when their interaromatic separation is near
7 Å. These structure-activity profiles suggest that it
should be possible to make molecules that are simul-
taneously δ agonists and µ antagonists that take
advantage of these pharmacophore arrangements and
at the same time improve on other aspects of the drug
such as bioavailability, efficacy, potency, selectivity, and
affinity. These molecules could be potentially therapeu-
tically useful in killing pain through δ agonism, yet
avoid the unwanted side-effects that are typically as-
sociated with µ agonism.
A 1 mmol scale was generally used, starting with 2.5 g of
pMBHA resin (0.4 mequiv/g substitution, 100-200 mesh). The
resin was swollen overnight in DMF and then washed with
DCM, then DIEA, and then washed again in DCM. The
carboxy terminal protected amino acid (NR-Boc-L-Thr(O-ben-
zyl)) and subsequent NR-Boc-protected amino acids were
coupled sequentially using an automated Milligen 9500 solid-
phase peptide synthesizer, using DIC/HOBt as the coupling
reagent for all amino acids except the amino terminal amino
acids. The N-terminal amino acids were coupled using BOP/
HOBt. The completeness of each coupling reaction was moni-
tored by ninhydrin testing. After amino acid coupling the resin
was washed with DMF and DCM, and then the NR-Boc group
was removed with 48.5% TFA in DCM with 2.5% anisole as
scavenger. The deblocked resin was then neutralized with
DIEA in DCM. Subsequent washing with DCM made the resin
ready for the next round of protected amino acid coupling. The
amino acids were used in excess (with respect to the resin)
during coupling (3-fold excess for DIC/HOBt coupling and 1.2-
fold excess for BOP/HOBt coupling). After the N-terminal
protected amino acid was coupled, the peptide-resin was NR-
Boc deblocked with TFA, washed several times with DMF, and
then several times with DCM. The resin was then dried under
a nitrogen stream and dried overnight in vacuo. The peptide
was then cleaved from the resin with concomitant amino acid
side-chain group deprotection with anhydrous liquid HF (10
mL of HF per gram of dry peptide-resin) in the presence of
p-cresol/p-thiocresol (1:1, v/v) as scavengers (1 mL of scavenger
mixture per gram of dry peptide-resin). The HF cleavage
reaction was performed at 0 °C for 1 h. The HF was then
removed in vacuo and the resin washed in ether and extracted
with glacial acetic acid. The extracted peptide was then frozen
and lyophilized. This crude peptide was purified by RP-HPLC
with a Vydac C-18 column using a gradient of 10-40%
acetonitrile in 0.1% aqueous TFA. Fractions which contained
pure peptide were pooled and lyophilized.
Gen er a l Meth od of Oxid a tion /Cycliza tion . The peptide
was cyclized by dissolving 100 mg of the purified linear peptide
in 600 mL of H2O/acetonitrile (2:1, v/v) and adding an excess
of K3Fe(CN)6. The solution was maintained at a pH of 8.5 by
dropwise addition of dilute NH4OH and was stirred for 2 h.
The excess ferro- and ferri-cyanides were removed with
Amberlite IRA-68 resin. The resin was then filtered and
extracted with glacial acetic acid. The filtrate and the extract
were then pooled and concentrated by rotary evaporation to
give the crude cyclic peptide. This cyclic peptide was then
purified by RP-HPLC using the method described above for
the purification of the crude linear peptide. The purity of these
peptides was assessed by TLC in four solvent systems and by
analytical HPLC. In each case only one spot was observed by
TLC, and by integration of HPLC chromatograms it was
determined that each peptide was >98% pure. The identity of
each peptide was confirmed by AAA, mass spectral [M + H]+
molecular ion fragmentation pattern, and by two-dimensional
NMR. Amino acid analyses were performed at the University
of Arizona Biotechnology Core Facility using an Applied
Biosystems model 420A amino acid analyzer with automatic
hydrolysis (vapor phase at 160 °C for 100 mins using 6 N HCl)
or with prior hydrolysis (6 N HCl, 110 °C, 24 h) and precolumn
PTA-AA analysis. FAB-MS spectra were obtained from the
College of Pharmacy at the University of Arizona. The analyti-
cal data for each analogue are given in Table 5.
Exp er im en ta l Section
Gen er a l Meth od s for P ep tid e Syn th esis. All peptides
were synthesized in a stepwise fashion by solid-phase meth-

Opiate Aromatic Pharmacophore SAR in CTAP Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4 577
Ta ble 5. Analytical Data of CTAP Analoguesa
amino acid analysis
TLCb Rf Values
HPLCc
K
FAB-MS (MH+)
no.
analogue
Tyr(1)
Arg(1)
Thr(2)
A
B
C
D
calcd
obs
2
3
4
5
7
8
9
10
11
[(2R,3R)â-MePhe1] CTAPd
[(2R,3S)â-MePhe1] CTAPd
[(2S,3R)â-MePhe1] CTAPd
[(2S,3S)â-MePhe1] CTAPb
[D-Trp1]CTAP
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.09
1.08
1.02
1.08
1.09
1.04
1.03
1.03
1.03
1.86
2.14
2.14
2.00
1.80
1.96
1.82
2.05
1.87
0.746
0.711
0.711
0.724
0.742
0.766
0.733
0.738
0.716
0.760
0.744
0.744
0.744
0.768
0.746
0.746
0.737
0.737
0.427
0.427
0.393
0.385
0.455
0.475
0.475
0.442
0.442
0.585
0.585
0.560
0.551
0.600
0.608
0.616
0.558
0.558
2.32
2.13
1.39
1.48
1.78
3.39
2.21
2.11
1.98
1118
1118
1118
1118
1143
1157
1157
1157
1157
1118
1118
1118
1118
1143
1157
1157
1157
1157
[(2R,3R)â-MeTrp1] CTAP
[(2R,3S)â-MeTrp1] CTAP
[(2S,3R)â-MeTrp1] CTAP
[(2S,3S)â-MeTrp1] CTAP
a
b
Structures of all peptides were confirmed by one- and two-dimensional 1H NMR. See General Methods. c Vydac 218TP104, 4.6 mm
d
x 25 cm, Isocratic, 75% acetonitrile, 25% 0.1% TFA in H2O; flow rate 1 mL/min; detection 230, 254, and 280 nm. Data taken from ref
7.
Am in o Acid a n d P ep tid e Syn th eses. The four diastere-
oisomers of NR-Boc-â-MeTrp(1′-Mes)27 and the four diastere-
oisomers of NR-Boc-â-MePhe28,29 were synthesized according
to published methods developed in this laboratory. D-Phe and
D-Trp were cyclized using the Pictet-Spengler reaction30 to
generate D-Tic and D-Tca, respectively. D-Tic, D-Tca, the four
NR-Boc-â-MePhe isomers, the four NR-Boc-â-MeTrp(1′-Mes)
isomers, and L-Pen(S-4′-methylbenzyl) were NR-t-Boc protected
at their R-amine group using a published method.31
of purified linear peptide, 120 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 53
mg. See Table 5 for the analytical data of this peptide.
[(2S,3S)â-MeP h e1]CTAP (H-(2S,3S)â-MeP h e-c[Cys-Tyr -
D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (5). This peptide was syn-
thesized, cyclized, and purified according to the procedure
described above, starting from 2.5 g of pMBHA resin. Yield of
peptide-resin, 3.6 g; yield of crude linear peptide, 340 mg; yield
of purified linear peptide, 210 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 87
mg. See Table 5 for the analytical data of this peptide.
[D-Tic1]CTAP (H-D-Tic-c[Cys-Tyr -D-Tr p -Ar g-Th r -P en ]-
Th r -NH2) (6). This peptide was synthesized, cyclized, and
purified according to the procedure described above, starting
from 2.5 g of pMBHA resin. Yield of peptide-resin, 3.6 g; yield
of crude linear peptide, 340 mg; yield of purified linear peptide,
210 mg; yield of purified cyclic disulfide peptide from 100 mg
of purified linear peptide, 87 mg. See Table 5 for the analytical
data of this peptide.
[D-Tr p 1]CTAP (H-D-Tr p -c[Cys-Tyr -D-Tr p -Ar g-Th r -P en ]-
Th r -NH2) (7). This peptide was synthesized, cyclized, and
purified according to the procedure described above, starting
from 2.5 g of pMBHA resin. Yield of peptide-resin, 4.17 g; yield
of crude linear peptide, 794.2 mg; yield of purified linear
peptide from 500 mg of crude linear peptide, 284 mg; yield of
purified cyclic disulfide peptide from 200 mg of purified linear
peptide, 154.3 mg. See Table 5 for the analytical data of this
peptide.
[(2R,3R)â-MeTr p1]CTAP (H-(2R,3R)â-MeTr p-c[Cys-Tyr -
D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (8). This peptide was syn-
thesized, cyclized, and purified according to the procedure
described above, starting from 2.5 g of pMBHA resin. Yield of
peptide-resin, 3.39 g; yield of crude linear peptide, 430.1 mg;
yield of purified linear peptide, 143.3 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 78.8
mg. See Table 5 for the analytical data of this peptide.
[(2R,3S)â-MeTr p 1]CTAP (H -(2R,3SR)â-MeTr p -c[Cys-
Tyr -D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (9). This peptide was
synthesized, cyclized, and purified according to the procedure
described above, starting from 2.5 g of pMBHA resin. Yield of
peptide-resin, 3.19 g; yield of crude linear peptide, 390.7 mg;
yield of purified linear peptide, 127.9 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 72.4
mg. See Table 5 for the analytical data of this peptide.
[(2S,3R)â-MeTr p 1]CTAP (H-(2S,3R)â-MeTr p -c[Cys-Tyr -
D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (10). This peptide was syn-
thesized, cyclized, and purified according to the procedure
described above, starting from 2.5 g of pMBHA resin. Yield of
peptide-resin, 3.41 g; yield of crude linear peptide, 411.1 mg;
yield of purified linear peptide, 129.9 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 68.4
mg. See Table 5 for the analytical data of this peptide.
[(2S,3S)â-MeTr p 1]CTAP (H-(2S,3S)â-MeTr p -c[Cys-Tyr -
D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (11). This peptide was syn-
thesized, cyclized, and purified according to the procedure
described above, starting from 2.5 g of pMBHA resin. Yield of
peptide-resin, 3.44 g; yield of crude linear peptide, 386.4 mg;
CTAP (H -D-P h e-c[Cys-Tyr -D-Tr p -Ar g-Th r -P en ]-Th r -
NH2) (1). This analogue was obtained by stepwise elongation
of peptide-resin by the method outlined in the section on
general peptide synthesis methods. Starting from 5 g of
pMBHA resin, the following amino acids were added to the
growing peptide chain: NR-Boc-L-Thr(O-benzyl), NR-Boc-L-Pen-
(S-4′-methylbenzyl), NR-Boc-L-Thr(O-benzyl), NR-Boc-L-Arg(Nγ-
tosyl), NR-Boc-D-Trp, NR-Boc-L-Tyr(O-2′,6′-dichlorobenzyl), NR-
Boc-L-Cys(S-4′-methylbenzyl), and NR-Boc-D-Phe. After the
final amino acid was coupled, the terminal t-Boc group was
removed with TFA and the resin washed several times with
DMF, then several times with DCM, dried under a nitrogen
stream, and dried additionally overnight in vacuo. Dry diethyl
ether (50 mL) was added to the vessel containing the cleaved
peptide and stirred for 20 min. The mixture was filtered and
subject to several rounds of washing in diethyl ether (4 × 50
mL), vacuum-drying, and mechanical grinding with a spatula
to break up clumping material. This product was then dried
by vacuum filtration. The peptide was then extracted from the
resin with glacial acetic acid (4 × 50 mL). The pooled extracts
were then frozen and lyophilized (yield, 1.7 g). The crude linear
peptide (500 mg) was then purified by RP-HPLC as described
above to yield 293.3 mg of pure linear peptide. The pure linear
peptide was then cyclized by the method described above and
purified by RP-HPLC as described above to yield 188 mg of
the pure cyclic disulfide peptide. See Table 5 for the analytical
data of this peptide.
[(2R,3R)â-MeP h e1]CTAP (H -(2R,3R)â-MeP h e-c[Cys-
Tyr -D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (2). This peptide was
synthesized, cyclized, and purified according to the procedure
described above, starting from 1.75 g of pMBHA resin. Yield
of peptide-resin, 2.3 g; yield of crude linear peptide, 190 mg;
yield of purified linear peptide, 110 mg; yield of purified cyclic
disulfide peptide, 73 mg. See Table 5 for the analytical data
of this peptide.
[(2R,3S)â-MeP h e1]CTAP (H-(2R,3S)â-MeP h e-c[Cys-Tyr -
D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (3). This peptide was syn-
thesized, cyclized, and purified according to the procedure
described above, starting from 2.5 g of pMBHA resin. Yield of
peptide-resin, 3.3 g; yield of crude linear peptide, 310 mg; yield
of purified linear peptide, 220 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 68
mg. See Table 5 for the analytical data of this peptide.
[(2S,3R)â-MeP h e1]CTAP (H-(2S,3R)â-MeP h e-c[Cys-Tyr -
D-Tr p -Ar g-Th r -P en ]-Th r -NH2) (4). This peptide was syn-
thesized, cyclized, and purified according to the procedure
described above, starting from 1.5 g of pMBHA resin. Yield of
peptide-resin, 2.2 g; yield of crude linear peptide, 170 mg; yield

578 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4
Bonner et al.
yield of purified linear peptide, 137.2 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 65.5
mg. See Table 5 for the analytical data of this peptide.
site fits were made using the F-ratio test using a p value of
0.05 as the cutoff for significance. Data best fitted by a one-
site model were reanalyzed using the logistic equation. Data
obtained from independent measurements are presented as
the arithmetic mean ( SEM.
In Vivo Assa y Meth od s: Su bjects. Male ICR mice weigh-
ing 20-30 g were used throughout these studies. They were
housed in groups of four in Plexiglas boxes, maintained in a
light- and temperature-controlled environment with food and
water provided ad libidum until antinociception testing was
performed. All testing was in accordance with the recom-
mendations and policies of the International Association for
the Study of Pain (IASP) and National Institutes for Health
(NIH) and the University of Arizona Guidelines for the care
and use of laboratory animals.
An tin ocicep tion Assa y. Antinociception determinations
were made using the warm water tail-flick assay as described
previously.7 Water of 50 °C was used, and a 15 s cutoff time
was employed. Antinociception was quantified from the test
and control latencies to reflexive tail-flick by the relation-
ship: % antinociception ) (test - control)/(15 - control) × 100.
Dr u g Ad m in istr a tion . The CTAP analogues were admin-
istered intracerebroventricularly to male ICR mice which were
lightly ether-anesthetized. The drug was delivered by the
method of Haley and McCormick.34 Briefly, the skin covering
the skull was split with a scalpel and the drug delivered via
Hamilton microsyringe with a 25-gauge needle with a 2.5 mm
depth-guard plastic collar attached. The drug was injected 2
mm caudal to bregma and 2 mm lateral to midline to ensure
delivery to the lateral ventricle. Doses from 6 to 100 nmol per
mouse were analyzed for antinociception and gross behavioral
effects. Percent analgesia was determined for each dose at 10,
20, 30, 45, and 60 min postinjection.
[D-Tca 1]CTAP (H-D-Tca -c[Cys-Tyr -D-Tr p -Ar g-Th r -P en ]-
Th r -NH2) (12). This peptide was synthesized, cyclized, and
purified according to the procedure described above, starting
from 4 g of pMBHA resin. Yield of peptide-resin, 5.9 g; yield
of crude linear peptide from 4 g of peptide-resin, 590 mg; yield
of purified linear peptide, 228 mg; yield of purified cyclic
disulfide peptide from 100 mg of purified linear peptide, 44
mg. See Table 5 for the analytical data of this peptide.
MVD a n d GP I Bioa ssa ys. Electrically induced smooth
muscle contraction of mouse vas deferens (MVD) and strips
of guinea pig ileum (GPI) longitudinal muscle-myenteric plexus
was used for the bioassays. The smooth muscle assays were
performed as described previously.32,33 The GPI assay was
performed by taking strips of longitudinal muscle-myenteric
plexus from nonterminal ileum from male Hartley guinea pigs
weighing 250-500 g. Tissues were suspended in organ baths
(20 mL capacity) containing Krebs-bicarbonate buffer that was
continuously bubbled with 95% O2 and 5% CO2 and main-
tained at 37 °C. The tissues were attached to isometric force
transducers calibrated to 1 g of resting tension, and after 15
min equilibration period without tension they were stretched
to 1 g of resting tension. The tissues were then transmurally
stimulated between platinum electrodes at 0.1 Hz, 0.4 ms
pulses, and supramaximal voltage. Contractions were recorded
on a Grass 70 multichannel recorder. The effect of agonists
on electrically evoked twitch tension was measured after
incubation for 3 min. Antagonists were added to the bath 2
min before the addition of the agonist. Drugs were added in
14-60 µL volumes. Changes in contraction height after drug
exposure were expressed as a percentage of the height average
for a minute preceding delivery of the agonist divided by the
contraction height at maximal inhibition after exposure to the
dose of agonist. Agonist activities are reported as the percent
inhibition of electrically evoked twitch height at a 1 µM
concentration. Antagonism is represented by the dose-
response curve rightward fold-shift after the addition of a 1
µM concentration of the antagonist. Standard errors are
determined.
The MVD assay required the removal of vasa deferentia
from male ICR mice (25-40 g) and placement in organ baths
at 37 °C, which are oxygenated as above, in magnesium-free
Krebs buffer. The tissues were hung without tension for 15
min, then placed under 0.5 g resting tension. The tissues were
transmurally stimulated at 0.1 Hz, with 2 ms pulses and
supramaximal voltage. Drug studies were performed as de-
scribed for the GPI assays.
Ra d ioliga n d Bin d in g Meth od s. Receptor binding assays
were performed essentially as previously described.7 Mem-
branes were prepared by taking whole brains from adult
Sprague-Dawley rats (250-300 g) which were sacrificed by
decapitation. The brain was rapidly removed and homogenized
at 0 °C in 20 volumes of 50 mM Tris-HCl buffer at pH 7.4
using a glass-Teflon homogenizer. The membrane fraction
resulting from a 15 min, 4 °C centrifugation at 48 000g was
resuspended in 20 volumes Tris-HCl buffer and incubated at
25 °C to dissociate receptor from bound endogenous ligands.
The incubated homogenate was then recentrifuged and resus-
pended in buffer.
Sta tistics. The values given in each of the in vitro and in
vivo assays are given as the mean ( SEM.35
NMR Exp er im en ts. NMR samples were prepared by
dissolving 5 mg of the dry, lyophilized, pure cyclic peptide in
500 µL of a 90% H2O/10% D2O sodium acetate (50 mM) buffer
containing sodium azide (1 mM) and were corrected to pH 4.5
with acetic acid-d3 (final peptide concentration ∼ 10 mM). For
COSY-35 spectra, peptides were prepared identically, except
the buffer contained no H2O (100% D2O). Spectra were
acquired without spinning at 303 K on a Bruker AM-500
spectrometer equipped with an Aspect 3000 computer and
using a 5 mm water suppression probe. Resonance assign-
ments were made by analysis of the two-dimensional DQF-
COSY, TOCSY, and ROESY spectra. All 2D spectra were
acquired in the phase-sensitive mode. Each spectrum was
acquired using 2048 complex points in t2 and an F2 spectral
width of 6250 Hz. Next, 750 real points were acquired in t1
with an F1 spectral width of 6250 Hz. The initial sampling
delay was one dwell time, and the final t1 delay was 60 ms.
At least 32 scans were collected per experiment. Solvent signal
was suppressed by a phase-locked presaturation for 1.5 s at
low power. Normal TOCSY and ROESY pulse sequences were
followed by a Hahn-echo period to improve solvent suppres-
sion.36 TOCSY spectra were acquired using the clean DIPSI-
2RC TOCSY sequence.37 ROESY spectra were acquired with
a CW spin-lock field strength of 6670 Hz using a mixing time
of 200 ms.
The generation of full 2D matrices was achieved by zero-
filling to 2048 real points in t1 and right-shifting by one point
in t1, multiplying by a sine-bell window function (ROESY) or
a skewed sine-bell window function (TOCSY), and transform-
ing into a 2048 × 1024 matrix. All 2D matrixes were created
and analyzed using the FELIX95 computer program (Biosym
Technologies, Inc., San Diego, CA) on a Silicon graphics
workstation. The complete assignments and chemical shift
data for the analogues examined (2, 3, 4, 5, 7, 8, 9, 10, and
11) are given in the Supporting Information.
For all inhibition studies, rat brain plasma membranes were
incubated at 25 °C for 3 h in a total volume of 1 mL of 50 mM
Tris-HCl, 1.0 mg/mL BSA, 30 µM bestatin, 50 µg/mL bacitra-
cin, 10 µM captopril, and 0.1 mM phenylmethylsulfonyl
fluoride buffer at pH 7.4 containing [3H][4′-Cl-Phe4]DPDPE
(0.75 nM) or [3H]DAMGO (1.0 nM), and at least 10 concentra-
tions of each analogue. Each analogue was tested at least three
times. Specific binding for the µ and δ receptors was defined
as the difference in the amounts of radioligands bound in the
absence and presence of 10 µM naltrexone.
Molecu la r Mod elin g. All theoretical conformational analy-
ses were performed using the MacroModel molecular modeling
package38 (versions 4.0 and 4.5). MacroModel was executed
from a Silicon Graphics IRIS workstation running the SGI
Binding data were analyzed by a nonlinear least-squares
regression analysis program (Inplot 4.03, GraphPad, San
Diego, CA). Statistical comparisons between one-site and two-

Opiate Aromatic Pharmacophore SAR in CTAP Analogues
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4 579
(6) Horan, J . H.; Wild, K. D.; Kazmierski, W. M.; Ferguson, R.;
Hruby, V. J .; Weber, S. J .; Davis, T. P.; Fang, L.; Knapp, R. J .;
Yamamura, H. I.; Kramer, T. H.; Burks, T. F.; Bowen, W. D.;
Takemori, A. E.; Porreca, F. Unexpected Antinociceptive Potency
of Cyclic [D-Tca1]CTAP: Potential for a Novel Mechanism of
Action. Eur. J . Pharmacol. 1993, 233, 53-62.
(7) Bonner, G. G.; Davis, P.; Stropova, D.; Ferguson, R.; Yamamura,
H. I.; Porreca, F.; Hruby, V. J . Opioid Peptides: Simultaneous
δ Agonism and µ Antagonism in Somatostatin Analogues.
Peptides 1997, 18, 93-100.
(8) Sasaki, Y.; Chiba, T. Novel Deltorphin Heptapeptide Analogues
with Potent Delta Agonist, Delta Antagonist, or Mixed Mu
Antagonist/Delta Agonist Properties. J . Med. Chem. 1995, 38,
3995-3999.
(9) Qian, X.; Ko¨ve´r, K. E.; Shenderovich, M. D.; Lou, B.-S.; Misicka,
A.; Zalewska, T.; Horvath, R.; Davis, P.; Bilsky, E. J .; Porreca,
F.; Yamamura, H. I.; Hruby, V. J . Newly Discovered Stereo-
chemical Requirements in Side Chain Conformation of δ Opioid
Agonists for Recognizing Opioid δ Receptors. J . Med. Chem.
1994, 37, 1746-1757.
(10) Qian, X.; Shenderovich, M. D.; Ko¨ve´r, K. E.; Davis, P.; Horva´th,
R.; Zalewska, T.; Yamamura, H. I.; Porreca, F.; Hruby, V. J .
Probing the Stereochemical Requirements for Receptor Recogni-
tion of δ Opioid Agonists Through Topographic Modifications
in Position 1. J . Am. Chem. Soc. 1996, 118, 7280-7290.
(11) Chew, C.; Villar, H. O.; Loew, G. H. Characterization of the
Bioactive Form and Molecular Determinants of Recognition of
Cyclic Enkephalin Peptides at the δ Opiate Receptor. Biopoly-
mers 1993, 33, 647-657.
(12) Froimowitz, M. Conformational Search in Enkephalin Analogues
Containing a Disulfide Bond. Biopolymers 1990, 30, 1011-1025.
(13) Ohno, Y.; Segawa, M.; Ohishi, H.; Doi, M.; Kitamura, K.; Ishida,
T.; Inoue, M.; Iwashita, T. Conformation of Deltorphin-II in
Membrane Environments Studied by Two-Dimensional NMR
Spectroscopy and Molecular Dynamics Calculations. Eur. J .
Biochem. 1993, 212, 185-191.
(14) Nikiforovich, G. V.; Hruby, V. J .; Prakash, O.; Gehrig, C. A.
Topographical Requirements for δ Selective Opioid Peptides.
Biopolymers 1991, 31, 941-955.
(15) Flippen-Anderson, J . L.; Hruby, V. J .; Collins, N.; George, C.;
Cudney, B. X-ray Structure of [D-Pen2, D-Pen5]enkephalin, a
Highly Potent, δ Opioid Receptor-Selective Compound: Com-
parisons with Proposed Solution Conformations. J . Am. Chem.
Soc. 1994, 116, 7523-7531.
operating system AIX 4.0.5. The AMBER force field was used
to calculate energy values using a distant-dependent dielectric
constant (ꢀ ) 4 D) and using the united-atom protocol and
N-protonated forms. The starting conformation for these CTAP
analogues was taken from the extensive NMR and molecular
dynamics analyses reported previously for the parent pep-
tide,5,24 1, which showed the peptide backbone of this peptide
to be conformationally stable. The conformations of these
peptides were refined to reflect vicinal coupling constants
determined from COSY spectra using the Karplus correla-
tion,23 ROE cross-peaks determined from ROESY spectra using
correlation between signal strength and interatomic distance
as reported in the literature,39 and to reflect the changes in
proton chemical shift values from typical values25 seen in
peptides that are due to anisotropic affects. A conformational
grid search of 30° was then used for exocyclic rotatable bonds
and conformers energy minimized. Conformations that were
within 50 kJ /mol of the minimum-energy structure were
grouped into families based on dihedral angle similarity and
the lowest energy member of each family extensively mini-
mized. Table 4 gives the results of these studies.
Ack n ow led gm en t. The authors thank Dr. Neil
J acobson for his help in obtaining NMR data and
analysis. The authors thank Drs. Lakmal Boteju and
Guigen Li for the syntheses of the four NR-Boc-â-
MeTrp(NinMes) and NR-Boc-â-MePhe analogues, respec-
tively. This research was supported by a grant from the
National Institute on Drug Abuse, NIDA grant number
DA-06284. The contents of this paper are solely the
responsibility of the authors and do not necessarily
represent the official views of the National Institute on
Drug Abuse.
Su p p or tin g In for m a tion Ava ila ble: CTAP analogues,
proton chemical shift assignment tables, and NOE/ROE data.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(16) Loew, G.; Keys, C.; Luke, B.; Polgar, W.; Toll, L. Structure-
Activity Studies of Morphiceptin Analogues: Receptor Binding
and Molecular Determinants of µ Affinity and Selectivity. Mol.
Pharmacol. 1986, 29, 546-553.
(17) Mierke, D. F.; Schiller, P. W.; Goodman, M. A Conformational
Comparison of Two Stereoisomeric Cyclic Dermorphin Analogues
Employing NMR and Computer Simulations. Biopolymers 1990,
29, 943-952.
(18) Hruby, V. J .; Toth, G.; Gehrig, C. A.; Kao, L. F.; Knapp, R.; Lui,
G. K.; Yamamura, H. I.; Kramer, T. H.; Davis, P.; Burks, T. F.
Topographically Designed Analogues of [D-Pen2, D-Pen5]enke-
phalin. J . Med. Chem. 1991, 34, 1823-1830.
(19) Toth, G.; Russell, K. C.; Landis, G.; Kramer, T. F.; Fang, L.;
Knapp, R.; Davis, P.; Burks, T. F.; Yamamura, H. I.; Hruby, V.
J . Ring Substituted and Other Conformationally Constrained
Tyrosine Analogues of [D-Pen2, D-Pen5]enkephalin with Delta
Opioid Receptor Selectivity. J . Med. Chem. 1992, 26, 2384-2391.
(20) Peter, A.; Toth, G.; Olajos, E.; Tourwe, D. Chromatographic
Behavior of Opioid Peptides Containing â-Methylphenylalanine
Isomers. J . Chromatogr. 1995, A 705, 267-273.
(21) Mosberg, H. I.; Omnaas, J . R.; Lomize, A. Heyl, D. L.; Nordan,
I.; Mousigian, C.; Davis, P.; Porreca, F. Development of a Model
for the Delta Opioid Receptor Pharmacophore. 2. Conformation-
ally, Restricted Phe3 Replacements in the Cyclic Delta Receptor
Selective Tetrapeptide Tyr-c[D-Cys-Phe-D-Pen]OH (J OM-13). J .
Med. Chem. 1994, 37, 4384-4391.
Refer en ces
(1) Symbols and abbreviations are in accordance with the recom-
mendations of the IUPAC-IUB J oint Commission on Biochemi-
cal Nomenclature: Nomenclature and Symbols for Amino Acids
and Peptides. Biochem. J . 1984, 219, 345-373. Other abbrevia-
tions include the following: DCM, dichloromethane; DMF,
dimethylformamide; HF, hydrogen fluoride; DIEA, diisopropyl-
ethylamine; MVD, mouse vas deferens; GPI, guinea pig ileum;
SEM, standard error of the mean; FAB-MS, fast atom bombard-
ment-mass spectroscopy; RP-HPLC, reversed-phase high per-
formance liquid chromatography; AAA, amino acid analysis;
BSA, bovine serum albumin; DIC, diisopropylcarbodiimide; TFA,
trifluoroacetic acid; Tic, tetrahydroisoquinoline carboxylic acid;
TCA, tetrahydo-â-carboline carboxylic acid; Boc, tertiary-butyl-
oxycarbonyl; Pen, penicillamine; DVB, divinylbenzene; pMBHA,
para-methylbenzhydrylamine; NMR, Nuclear Magnetic Reso-
nance; DPDPE, c[D-Pen2,D-Pen5]enkephalin; PLO-17, H-Tyr-Pro-
NMePhe-D-Pro-NH2.
(2) Abdelhamid, E. E.; Sultana, M.; Portoghese, P. S.; Takemori,
A. E. Selective Blockage of Delta Opioid Receptors Prevents the
Development of Morphine Tolerance and Dependence in Mice.
J . Pharmacol. Exp. Ther. 1991, 258, 299-303.
(3) Schiller, P. W.; Nguyen, T. M.-D.; Weltrowska, G.; Wilkes, B.
C.; Marsden, B. J .; Lemieux, C.; Chung, N. N. Differential
Stereochemical Requirements of µ vs. δ Opioid Receptors for
Ligand Binding and Signal Transduction: Development of a
Class of Potent and Highly δ-Selective Peptide Antagonists. Proc.
Natl. Acad. Sci. U.S.A. 1992, 89, 11871-11875.
(22) Hruby, V. J .; Li, G.; Haskell-Luevano, C.; Shenderovich, M.
Design of Peptides, Proteins and Peptidomimetics in Chi Space.
Biopolymers 1997, 43, 219-266.
(23) Karplus, M. Contact Electron-Spin Coupling of Nuclear Magnetic
Moments. J . Phys. Chem. 1959, 30, 11-15.
(24) Kazmierski, W. M.; Ferguson, R. D.; Lipkowski, A. W.; Hruby,
V. J . A Topographical Model of µ-Opioid and Brain Somatostatin
Receptor Selective Ligands. Int. J . Pept. Protein Res. 1995, 46,
265-278.
(4) Kazmierski, W. M.; Wire, W. S.; Lui, G. K.; Knapp, R. J .; Shook,
J . E.; Burks, T. F.; Yamamura, H. I.; Hruby, V. J . Design and
Synthesis of Somatostatin Analogues with Topographic Proper-
ties that Lead to Highly Potent and Specific µ Opiate Receptor
Antagonists with Greatly Reduced Binding at Somatostatin
Receptors. J . Med. Chem. 1988, 31, 2170-2177.
(25) Bundi, A.; Wu¨thrich, K. 1H NMR Parameters of the Common
Amino Acid Residues Measured in Aqueous Solution of the
Linear Peptides H-Gly-Gly-X-L-Ala-OH. Biopolymers 1979, 18,
285-297.
(26) Upson, D. A.; Hruby, V. J . Synthesis of Specifically Deuterated
S-Benzylcysteines and of Oxytocin and Related Diastereomers
Deutereated in the Half-cystine Positions. J . Org. Chem. 1976,
41, 1353-1358.
(5) Kazmierski, W. M.; Yamamura, H. I.; Hruby, V. J . Topographic
Design of Peptide Neurotransmitters and Hormones on Stable
Backbone Templates: Relation of Conformation and Dynamics
to Bioactivity. J . Am. Chem. Soc. 1991, 113, 2275-2283.

580 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 4
Bonner et al.
(27) Boteju, L. W.; Wegner, K.; Qian, X.; Hruby, V. J . Asymmetric
Synthesis of Unusual Amino Acids: Synthesis of Optically Pure
Isomers of N-Indole-(2-mesitylenesulfonyl)-â-methyltryptophan.
Tetrahedron 1994, 50, 2391-2404.
(28) Dharanipragada, R.; Nicolas, E.; To¨th, G.; Hruby, V. J . Asym-
metric Synthesis of Unusual Amino Acids: Synthesis of Optically
Pure Isomers of Beta-Methylphenylalanine. Tetrahedron Lett.
1989, 30, 6841-6844.
(29) Li, G.; J arosinski, M. A.; Hruby, V. J . Diastereospecific Tandem
Michael-like Addition/Electrophilic Bromination: A One-pot
Tandem Asymmetric Synthesis of Precursors of Unusual Amino
Acids. Tetrahedron Lett. 1993, 34, 2561-2564.
(30) Pictet, A.; Spengler, T. Uberdie Bildung von Isochinolin-Deri-
vaten durch Einworking von Methylal auf Phenyl-athylamine
und Tyrosin. Chem. Ber. 1911, 44, 2030-2036.
(34) Haley, T. J .; McCormick, W. G. Pharmacological Effects Pro-
duced by Intracerebral Injections of Drugs in the Conscious
Mouse. Br. J . Pharmacol. Chem. 1957, 12, 12-15.
(35) Tallarida, R. J .; Murray, R. B. Manual of Pharmacological
Calculations with Computer Programs, 2nd ed.; Springer-Verlag:
New York, 1987, pp 26-31.
(36) Rance, M.; Sorensen, O. W.; Bodenhausen, G.; Wagner, G.; Ernst,
R. R.; Wu¨thrich, K. Improved Spectral Resolution in COSY 1H
NMR Spectra of Proteins via Double Quantum Filtering. Bio-
chem. Biophys. Res. Commun. 1983, 117, 479-485.
(37) Cavanaugh, J .; Rance, M. Suppression of Cross-Relaxation
Effects in TOCSY Spectra via
a Modified DIPSI-2 Mixing
Sequence. J . Magn. Res. 1992, 96, 670-678.
(38) Mohamadi, F.; Richards, N. G. J .; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W.
C. Macromodel - An Integrated Software System for Modeling
Organic and Bioorganic Molecules Using Molecular Mechanics.
J . Comput. Chem. 1990, 11, 440-467.
(31) Tarbell, D. S.; Yamamoto, Y.; Pope, B. M. New Method to
Prepare N-tert-butoxycarbonyl Derivatives and the Correspond-
ing Sulfur Analogues from Di-tert-butyl Dicarbonate or Di-tert-
butyl Dithiolodicarbonates and Amino Acids. Proc. Natl. Acad.
Sci. U.S.A. 1972, 69, 730-732.
(32) Hughes, J .; Kosterlitz, W.; Leslie, F. M. Effect of Morphine on
Adrenergic Transmission in the Mouse Vas Deferens: Assess-
ment of Agonist and Antagonist Potencies of Narcotic Analgesics.
Br. J . Pharmacol. 1975, 53, 371-381.
(33) Kosterlitz, H. W.; Lydon, R. J .; Wat, A. J . The Effects of
Adrenaline, Noradrenaline, and Isoprenaline on Inhibitory
alpha- and beta-Adrenoceptors in the Longitudinal Muscle of
the Guinea Pig Ileum. Br. J . Pharmacol. 1970, 39, 398-413.
(39) Polinsky, A.; Cooney, M. G.; Toy-Palmer, A.; Osapay, G.;
Goodman, M. Synthesis and Conformational Properties of the
Lanthionine Bridged Opioid Peptide [D-Ala2, Ala5]enkephalin as
Determined by NMR and Computer Simulations. J . Med. Chem.
1992, 35, 4185-4194.
J M9900218