Synthesis of Tic-D-Phe Ψ[CH2-CH2] isostere and its use in the development of melanocortin receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2005.12.005

The first synthesis of Tic-d-Phe Ψ[CH₂-CH₂] isostere is described, which features diastereoselective alkylation of the tricyclic lactam 14. The use of this novel dipeptide isostere in the development of melanocortin agonists has been

Full text of DOI:10.1016/j.bmcl.2005.12.005

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1721–1725
Synthesis of Tic-D-Phe W[CH2–CH2] isostere and its use
in the development of melanocortin receptor agonists
Xinrong Tian,* Xuemei Chen, Lixian Gan, Jeffery C. Hayes, Adrian G. Switzer,
Mark G. Solinsky, Frank H. Ebetino, John A. Wos, Beth B. Pinney, Julie A. Farmer,
Doreen Crossdoersen and Russell J. Sheldon
Procter & Gamble Pharmaceuticals, Health Care Research Center, 8700 Mason Montgomery Rd., Mason, OH 45040, USA
Received 4 October 2005; revised 30 November 2005; accepted 1 December 2005
Available online 20 December 2005
Abstract—The first synthesis of Tic-D-Phe W[CH2–CH2] isostere is described, which features diastereoselective alkylation of the tri-
cyclic lactam 14. The use of this novel dipeptide isostere in the development of melanocortin agonists has been demonstrated by the
synthesis of peptidomimetic 7 and non-peptidic ligand 27. Both compounds displayed significant binding and agonist potency at the
MC4R.
Ó 2005 Elsevier Ltd. All rights reserved.
The melanocortin receptors (MCRs) belong to a family
of 7-transmembrane G-protein-coupled receptors and
five subtypes have been identified and cloned (MC1R-
MC5R). These receptors are activated by the peptide li-
gands: a,b,c-melanocyte stimulating hormones (MSH)
and adrenocorticotropin (ACTH), which are derived
from a common precursor protein, proopiomelanocor-
tin (POMC), by post translational cleavage.1 A common
structural feature present in all endogenous ligands
(MSH and ACTH) of the MCRs is the His-Phe-Arg-
Trp sequence, which has been identified as the minimal
peptide fragment necessary for activating the receptors.2
Further structural modifications toward a-MSH showed
that inverting Phe7 to the D-configuration resulted in an
analog, NDP-MSH, with enhanced potency and enzy-
matic stability.3 The tetrapeptide His-D-Phe-Arg-Trp4
and tripeptide D-Phe-Arg-Trp5 have been thus employed
as intriguing templates for SAR studies and for MCR
ligand design. A key approach within our research
program centered on replacement of each amino acid
residue of tetrapeptide His-D-Phe-Arg-Trp and led to
the identification of a number of structurally diverse
analogs displaying significant binding and agonist
activity at MC4R as well as moderate selectivity relative
to MC1R, such as Ac-Tyr-D-Phe-Arg-2-Nal-NHCH3 (1)
and Tic-D-Phe-Arg-2-Nal-NHCH3 (2) (Figs. 1 and 2).
A major drawback to the use of peptides in therapeutic
applications is the susceptibility to rapid degradation
in vivo by numerous peptidases. One well-established
strategy to enhance metabolic stability of peptides to
proteolytic enzymes is to substitute a peptide amide
bond with a non-hydrolyzable amide surrogate.6 In the
course of our efforts to develop peptidomimetic MCR
agonists, we have explored the replacement of two adja-
cent amino acids of tetrapeptide leads derived from His-
D-Phe-Arg-Tyr with a variety of W[CH2–CH2] dipeptide
isosteres. We anticipated that systemic substitution of
each amide bond with a CH2–CH2 unit would provide
insight into the relative importance of each amide bond
with regard to geometrical, electrostatic, and hydrogen
bond properties toward the binding and functional
potency of the tetrapeptide. This type of information,
in turn, could be incorporated into the design of pepti-
domimetics with significantly reduced peptidic charac-
ters by removing those amide bonds that are not
necessary for biological activity. Our initial results ob-
tained with replacement of the amide bond between
Tyr and D-Phe of tetrapeptide 1 were quite encouraging.
For instance, the corresponding Ôcarba’ analog of 1 (3)
not only displayed comparable affinity and agonist
potency (EC50) at MC4R but also was 7-fold more selec-
tive for MC4R relative to MC3R, as compared to the
Keywords: Tic-D-Phe W[CH2–CH2] isostere; Melanocortin agonists.
*
Corresponding author. Tel.: +1 513 622 1444; fax: +1 513 622
1195; e-mail: tian.x@pg.com
Present address: Department of Chemistry, University of Minnesota,
Minneapolis, MN 55455, USA.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.005

1722
X. Tian et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1721–1725
R2
O
NHCH3
O
NHCH3
O
O
O
H
N
NH
N
H
N
H
N
H
NH
O
NH
O
O
R1
HO
O
NH
NH
NH
NH
H2N
H2N
5 R1=H, R2=F
6 R1=Cl, R2=F
3 R1=OH, R2=H
4 R1=OH, R2=F
1
Figure 1. Tetrapeptide 1 and Ôcarba’ analogs.
the feasibility of applying these methods to prepare such
isosteres that contain a constrained amino acid, such as
Tic (a constrained Phe analog), has not been demon-
strated. Given the excellent diastereoselectivity associat-
ed with the route featuring the alkyation of 5-substituted
d-lactam for the synthesis of Ôcarba’ W[CH2–CH2] dipep-
tide unit,8 we sought to explore the possibility of extend-
ing this approach to synthesis of the Tic-D-Phe W[CH2–
CH2] dipeptide isostere required for the development of
new peptidomimetic MCR agonists. The key question
would be whether the alkylation of a constrained tricy-
clic lactam system (14, Scheme 1) in which conforma-
tional mobility of the C-5 substituent is highly
restricted would proceed with a similar level of 1,4-
asymmetric induction to that observed with the flexible
5-substituted d-lactams.8
O
NHCH3
O
H
N
x
y
N
H
NH
O
OH
N
O
HN
BOC
8
H2N
NH
2 x-y=CO-NH
7 x-y=CH2-CH2
Figure 2. Tetrapeptide 2 and the Ôcarba’ target analog.
parent peptide (Table 1, Figs. 1 and 2). This SAR find-
ing promoted us to investigate additional pseudo-tetra-
peptides derived by using other D-Phe-containing
W[CH2–CH2] dipeptide isosteres and some of them are
listed in Figure 1. It is interesting to find that replace-
ment of the phenyl ring of D-Phe side chain of Ôcarba’
analog 3 with 4-fluorophenyl group led to a potent
and selective MC4R agonist (4) which had a Ki of
5 nM, and was >4400-fold and >1200-fold selective for
MC4R over MC1R and MC3R (Table 1), respectively.
Removal of 4-hydroxyl group from the side chain of
Tyr residue or replacement of 4-hydroxyphenyl with 4-
chlorophenyl of 4 resulted in analogs with Ki values of
37 nM (5) and 214 nM (6), respectively, at MC4R and
good selectivity for MC4R over MC1R and MC3R.
The D-Phe-containing W[CH2–CH2] isosteres discussed
above were efficiently synthesized by means of the alkyl-
ation of 5-substituted d-lactam.8 These data provided
preliminary evidence that CH2–CH2 unit is a viable sur-
rogate for the amide bond between two amino acids at
the left side (N-terminus) of the tetrapeptide leads.
The precursor for the key tricyclic lactam 14 was pre-
pared by a two-carbon homologation using Meldrum’s
acid as reported.8 (S)-2-(2-(tert-Butoxycarbonyl)-
1,2,3,4-tetrahydroisoquinolin-3-yl)acetic acid 9 was cou-
pled with Meldrum’s acid to give 3-keto ester 11, which
was then deoxygenated by treatment with sodium boro-
hydride in the presence of acetic acid to afford 12. Cycli-
zation of 12 to lactam 14 was achieved using a two-step
reaction sequence. Following removal of the BOC group
of 12 with TFA, the salt of amine 13 was heated at reflux
in toluene in the presence of triethylamine to furnish the
tricyclic 14 in good yield. With 14 in hand, we then
investigated the facial selectivity of the alkylation of this
tricyclic lactam to introduce the benzyl group to the car-
bon next to the carbonyl group. A variety of bases and
solvents were examined to define the conditions that
would favor the formation of compound 15 possessing
the desired D-Phe configuration. In contrast with the
case of 5-substituted d-lactams,8 a mixture of two diaste-
reomers 15 and 16 was consistently obtained under all
conditions examined. The use of NaHMDS in THF/
DME (1:1) (0.2 mM) for enolate formation, optimized
The Ôcarba’ analog 7 became one of our next targets due
to the higher affinity and potency at the MC4R dis-
played by parent peptide 2 as compared to 1 (Fig. 2).
While only a few approaches for synthesis of Ôcarba’
W[CH2–CH2] dipeptide isosteres have been reported,8,9
Table 1. Binding affinity and agonist potency for 1–6a
Compound
MC1R Kib (nM)
MC3R Kib (nM)
MC4R Kib (nM)
MC4R EC50 (Emax, %)b (nM)
1
2
3
4
5
6
4520 1257
2276 677
1727 67
1632 77
104 10
44 5 (84)
10 2 (81)
38 5 (90)
10 3 (91)
20 1 (102)
193 45 (100)
24
214 28
2
8562 4113
22,267 1133
24,689 8171
6198 812
5
37
1
6
23,400
12,288 5600
0
55,474 13,257
18,016 13,499
214 49
a The analogs were screened against the human MC1R, MC3R, and MC4R as previously reported.7
b The data represent means of the at least three experiments SEM.

X. Tian et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1721–1725
1723
Scheme 1. Synthesis of protected Tic-D-Phe dipeptide isostere 8. Reagents and conditions: (a) DMAP, EDCI, CH2Cl2, 0 °C to room temperature,
4 h, 94%; (b) NaBH4, acetic acid, CH2Cl2, 10h, 95%; (c) TFA, H2O, CH2Cl2; (d) Et3N, toluene, 120 °C, 1 h, 73% yield over two steps; (e) LDA
(1.2 equiv), THF, ꢀ78 °C, 1 h; BnBr (1.2 equiv), ꢀ78 °C, 4 h, 58%; (f) 4 N HCl, reflux, 24 h; (g) (BOC)2O, Et3N, CH3OH, reflux, 4 h, 81% yield over
two steps.
conditions reported for the alkylation of 5-substituted
lactams,8 produced a mixture of 15:16 in a 1.2:1 ratio
in 57% yield. Ultimately, the following procedure was
found to be optimal in our hands to give a mixture of
15:16 in a 1.6:1 ratio in 58% yield. A solution of 14
was treated with LDA (1.2 equiv) at ꢀ78 °C for 1 h
and benzyl bromide (1.2 equiv) was added. The reaction
mixture was stirred for 4 h at ꢀ78 °C before it was
quenched with aqueous NH4Cl solution and extracted
with ethyl acetate. The extract was dried over MgSO4,
filtered, and concentrated. The residue was then subject-
ed to column chromatography (hexanes/ethyl acetate,
2:1).10
H-5 and H-3a, and between H-5 and H-4a. A strong
NOE observed between H-2 and H-3b and between
H-2 and H-4b, suggests an R configuration at the C-2.
For isomer 16, there is a much stronger NOE effect
between H-5 and H-4 a than H-4 b in NOESY spectra
run at both 0 °C and ꢀ30 °C. The NOESY spectrum
acquired at ꢀ30 °C clearly shows a strong NOE effect
between H-4a and H-2, and a medium NOE between
H-4 b and one of two C-10 protons. These correlation
data are suggestive of S configuration at the C-2 of 16
(Fig. 3).
The moderate diastereoselectivity achieved with the
alkylation of lactam 14 is probably because the high
degree of conformational rigidity enjoyed by the central
ring precludes the possibility for the C-5 substituent to
adopt an axial conformation, which is the key factor
contributing to the excellent selectivity seen with the
flexible 5-substituted d-lactams.8 Diastereomer 15
was subjected to acidic hydrolysis with 4 N HCl at
The stereochemistry at the newly formed chiral center
next to the carbonyl group for diastereomers 15 and
1
16 was established by the analysis of H, 2D-COSY,
and 2D-NOESY NMR data. As revealed by NOESY
experiments with 15, there is a strong NOE between
Figure 4. Pyrrolidine-based D-Phe-2-Nal dipeptide mimics and potent
peptidomimetic MCR agonists.
Figure 3. NOESY correlations of isomers 15 and 16.
O
NH
O
NHCH3
O
O
H
N
c
a
H2N
7
N
H
N
H
8 +
N
O
BOC
18
HN
H2N
HN
NO2
R
N
H2N
N
19 R=NO2
20 R=H
b
Scheme 2. Synthesis of peptidomimetic 7. Reagents and conditions: (a) EDCI, HOBt, NMM, DMF, 24 h, 49%; (b) BaSO4/Pd, acetic acid, CH3OH,
18 h; (c) TFA, CH2Cl2, 4 h, 62% yield over two steps.

1724
X. Tian et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1721–1725
Table 2. Binding affinity and agonist potency for 7, 22, 23, and 27a
Compound
MC1R Kib (nM)
MC3R Kib (nM)
23
MC4R Kib (nM)
MC4R EC50 (Emax, %)b (nM)
1 (84)
22
23
7
9
30
3
8
9
10
13
2
2
8
468 191
2411 79
382 53
22 5 (89)
160 36 (77)
142 47 (76)
3992 1322
71 29
220 14
76
27
9
a The analogs were screened against the human MC1R, MC3R, and MC4R as previously reported.7
b The data represent means of at least three experiments SEM.
reflux to afford D-Phe-Tic Ôcarba’ isostere 17 in a
quantitative yield, which was subsequently converted
to the BOC derivative 8. Isostere 8 was then applied to
synthesis of the desired Ôcarba’ analog 7, which was
accomplished by coupling 8 with Arg(NO2)-2-Nal-
NHCH3 dipeptide (18) followed by reductive deprotec-
tion of the guanidine moiety and BOC cleavage with
TFA to give 7 (Scheme 2).
with guanidine-containing pyrrolidine 24 under EDCI
activation afforded 25. Hydrogenative removal of the ni-
tro group from the guanidine moiety followed by the
BOC cleavage with TFA produced 27 (Scheme 3).
The binding affinity at three MCRs and agonist potency
for MC4R of 7 and 27 are listed in Table 2. The Ôcarba’
analog 7 possessed a Ki of 220 nM and an EC50 of
160 nM at the MC4R with moderate selectivity for the
MC4R over the MC1R and MC3R, but it was ꢁ16-fold
less potent at the MC4R than the tetrapeptide 2. Non-
peptidic analog 27, however, displayed a ꢁ3-fold better
affinity than 7 at MC4R and significant affinity at
MC1R and MC3R as well and was only 6-fold less po-
tent than the corresponding parent dipeptide 23. This re-
sult is significant in terms of the fact that 27 represents
the first analog developed by combining the SAR learn-
ing from two different approaches we have explored,
replacement of the amide bond between two amino acids
at the left side of tetrapeptides and design of conforma-
tionally constrained dipeptide mimics for two amino
acids at the right side of tetrapeptides. We have previ-
ously demonstrated that it is possible to further enhance
the potency of Ôcarba’ analogs by introducing substitu-
tion to the aromatic ring of D-Phe side chain (3 vs 4).
Development of nonpeptidic MCR agonists by coupling
8 with other dipeptide mimics has also been briefly ex-
plored. In addition to the approach of replacing amide
bond with a CH2–CH2 unit, another strategy we have
investigated for developing peptidomimetic melanocor-
tin agonists involves the design of conformationally con-
strained dipeptide mimetics to substitute two adjacent
amino acids of tetrapeptide leads. Our efforts along this
line have led to the discovery of pyrrolidine-based Arg-
2-Nal, two amino acids at the right side of numerous tet-
rapeptide leads, dipeptide mimetics as represented by 21.
Appending Xaa-D-Phe dipeptides, which correlate with
two amino acids at the left side of tetrapeptide leads,
to the Arg-2-Nal dipeptide mimetics resulted in the syn-
thesis of a number of potent MCR agonists, as exempli-
fied by 22 and 23 (Fig. 4 and Table 2, EC50 values at
MC1R for these two compounds are 8 and 30 nM,
respectively).11 A variety of amino acids are tolerated
as the capping groups of the D-Phe reside at the top side
chain and the resulting analogs exhibited excellent affin-
ity and agonist potency at MC4R and, in many cases, at
MC1R and MC3R as well. The synthetic availability of
Tic-D-Phe Ôcarba’ analog 8 using the approach described
above has provided an opportunity to expand the utility
of these new pyrrolidine dipeptide mimetics to generate
non-peptidic melanocortin agonists. Thus, coupling of 8
In summary, we have described a synthetic approach to
the Tic-D-Phe W[CH2–CH2] isostere 8 that features
alkylation of a fused tricyclic lactam 14. Unlike the case
of monocyclic 5-substituted d-lactam, the alkylation of
conformationally constrained tricyclic lactam system
led to a mixture of two isomers (15 and 16) under a vari-
ety of conditions, and the major isomer 15 possessed the
stereochemistry corresponding to D-Phe configuration.
Additionally, the reported approach also allows access
to Tic-Phe W[CH2–CH2] isostere, although it was
formed as the minor product of the alkylation step.
The use of protected dipeptide isostere 8 in the develop-
ment of MCR agonists with reduced peptidic characters
has been demonstrated by the synthesis of peptidomi-
metic 7 and non-peptidic analog 27.
References and notes
1. Eberle, A. N. In The Melanocortin Receptors; Cone, R. D.,
Ed.; Humana Press: Totwa, NJ, 2000, P3-68.
2. (a) Hruby, V. J.; Wilkes, B. C.; Hadley, M. E.; Al-Obeidi,
F.; Sawyer, T. K.; Staples, D. J.; deVaux, A. E.; Dym, O.;
Castrucci, A. M.; Hintz, M. F.; Riehm, J. P.; Rao, K. R. J.
Med. Chem. 1987, 30, 2126; (b) Castrucci, A. M. L.;
Hadley, M. E.; Sawyer, T. K.; Wilkes, B. C.; Al-Obeidi,
F.; Staples, D. J.; DeVaux, A. E.; Dym, O.; Hintz, M. F.;
Scheme 3. Synthesis of non-peptidic analog 27. Reagents and condi-
tions: (a) 8, EDCI, HOBt, NMM, DMF, 18 h; (b) H2, Pd/BaSO4,
acetic acid (10%), CH3OH, 12 h, 34% yield over two steps; (c)TFA,
CH2Cl2, 3 h, ꢁ100%.

X. Tian et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1721–1725
1725
Riehm, J.; Rao, K. R.; Hruby, V. J. Gen. Comp.
Endocrinol. 1989, 73, 157.
3. Sawyer, T. K.; Sanfillippo, P. J.; Hruby, V. J.; Engel, M.
H.; Heward, C. B.; Burnett, J. B.; Hadley, M. Proc. Natl.
Acad. Sci. U.S.A. 1980, 73, 157.
7. Tian, X.; Field, T.; Mazur, A. W.; Ebetino, F. H.; Wos, J.
A.; Crossdoersen, D.; Pinney, B. B.; Sheldon, R. J. Bioorg.
Med. Chem. Lett. 2005, 15, 2819.
8. Casimir, J. R.; Didierjean, C.; Aubry, A.; Rodriguez, M.;
Briand, J.; Gilles, G. Org. Lett. 2000, 2, 895.
4. (a) Koikov, L. N.; Ebetino, F. H.; Hayes, J. C.; Cross-
Doersen, D.; Knittel, J. J. Bioorg. Med. Chem. Lett.
2004, 14, 4839; (b) Koikov, L. N.; Ebetino, F. H.;
Solinsky, M. G.; Cross-Doersen, D.; Knittel, J. J. Bioorg.
Med. Chem. Lett. 2004, 14, 3997; (c) Todorovic, A.;
Holder, J. R.; Scott, J. W.; Haskell-Luevano, C. J.
Peptide Res. 2004, 63, 270; (d) Koikov, L. N.; Ebetino,
F. H.; Solinsky, M. G.; Crossdoersen, D.; Knittel, J. J.
Bioorg. Med. Chem. Lett. 2003, 13, 2647; (e) Holder, J.
R.; Xiang, Z.; Bauzo, R. M.; Haskell-Luevano, C.
Peptides 2003, 24, 73; (f) Holder, J. R.; Xiang, Z.;
Bauzo, R. M.; Haskell-Luevano, C. J. Med. Chem. 2002,
45, 5736; (g) Holder, J. R.; Bauzo, R. M.; Xiang, Z.;
Haskell-Luevano, C. J. Med. Chem. 2002, 45, 3073; (h)
Holder, J. R.; Bauzo, R. M.; Xiang, Z.; Haskell-
Luevano, C. J. Med. Chem. 2002, 45, 2801.
9. (a) Attwood, M. R.; Conway, E. A.; Dunsdon, R. M.;
Greening, J. R.; Handa, B. K.; Jones, P. S.; Jordan, S. C.;
Keech, E.; Wilson, F. X. Bioorg. Med. Chem. Lett. 1997, 7,
429; (b) Rodriguez, M.; Heitz, A.; Martinez, J. Tetrahe-
dron Lett. 1990, 31, 7319.
10. Spectral and analytic data for 15 and 16.
1
Compound 15: H NMR (300 MHz, CDCl3) d 7.10–7.40
(m, 9H), 5.41 (d, J = 17.2 Hz, 1H), 4.31 (d, J = 17.2 Hz,
1H), 3.72 (m, 1H), 3.51 (m, 1H), 2.95 (dd, J = 15.4, 11.6 Hz,
1H), 2.60–2.80 (m, 3H), 2.04 (m, 1H), 1.60–1.90 (m, 3H); 13
C
NMR (75 MHz, CDCl3) d 172.17 (C@O),140.54, 135.90,
135.50, 130.03, 128.76, 128.18, 127.05, 126.99, 126.85,
126.52, 54.56, 45.79, 44.13, 38.21, 37.95, 29.13, 24.37; MS
(CI) m/z 292 (M+1); HRMS calcd for (C13H15NO+H):
292.1701. Found: 292.1692.
1
Compound 16: H NMR (300 MHz, CDCl3) d 7.10–7.40
5. Fotsch, C.; Smith, D. M.; Adams, J. A.; Cheetham, J.;
Croghan, M.; Doherty, E. M.; Hale, C.; Jarosinski, M. A.;
Kelly, M. G.; Norman, M. H.; Tamayo, N. A.; Xi, N.;
Baumgartner, J. W. Bioorg. Med. Chem. Lett. 2003, 13,
2337.
6. Spatola, A. F. In Chemistry and Biochemistry of Amino
Acids, Peptides and Proteins; Weinstein, B., Ed.; Marcel
Dekker: New York, 1983; Vol. 7, pp 267–357.
(m, 9H), 5.20 (d, J = 17.7 Hz, 1H), 4.43 (d, J = 17.7 Hz,
1H), 3.55 (m, 2H), 2.60–2.90 (m, 2H), 2.17 (m, 1H), 1.84 (m,
1H), 1.40–1.70 (m, 2H); 13C NMR (75 MHz, CDCl3) d
172.02 (C@O), 140.48, 134.31, 133.53, 129.59, 128.77,
126.92, 126.82, 126.55, 53.61, 45.63, 43.92, 38.19, 36.64,
26.91, 22.56; MS (CI) m/z 292 (M+1); HRMS calcd for
(C13H15NO+H): 292.1701. Found: 292.1701.
11. Details of this research will be reported elsewhere.