General method for the synthesis of cyclic peptidomimetic compounds

Article abstract of DOI:10.1016/S0040-4039(01)01134-0

A number of cyclic peptides, wherein the i and i+1 residue side chains are joined by an alkyl linker and a Weinreb amide is present at the C-terminus, were synthesized. Using LiCl to solubilize these peptides in THF the C-terminus can be readily converted to an activated carbonyl such as an aldehyde or ketone.

Full text of DOI:10.1016/S0040-4039(01)01134-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 5801–5804
General method for the synthesis of cyclic peptidomimetic
compounds
Michael A. Walker* and Timothy Johnson
Bristol-Myers Squibb Pharmaceutical Research Inst., 5 Research Parkway, Wallingford, CT 06492, USA
Received 21 May 2001; accepted 6 June 2001
Abstract—A number of cyclic peptides, wherein the i and i+1 residue side chains are joined by an alkyl linker and a Weinreb
amide is present at the C-terminus, were synthesized. Using LiCl to solubilize these peptides in THF the C-terminus can be readily
converted to an activated carbonyl such as an aldehyde or ketone. © 2001 Elsevier Science Ltd. All rights reserved.
Cyclic peptides have been exploited as exploratory tools
and potential drug candidates in medicinal chemistry
because they provide rigid scaffolds for locking short
peptides into specific configurations.1,2 Therefore, it is
not surprising to find that this approach has been used
in the area of protease inhibitor design since it is
generally accepted that substrate binding to the enzyme
takes the form of a b-sheet. However, not only does
cyclization provide conformational rigidity, in certain
cases cyclization can have an effect on metabolic stabil-
ity, bioavailability and pharmacokinetics.3
convenient but also flexible enough to allow the rapid
assembly of a diverse set of structures. Of the available
methods for synthesizing cyclic peptides we focused on
the method of Mosberg, wherein the cyclic peptide is
formed by linking two cysteines via a linker (cf. Fig.
1).5 This methodology has been examined by others6 in
the area of aspartyl protease inhibitors; however, to
apply this to the synthesis of a serine protease inhibitor
P
X
P
X
O
O
O
H
N
H
N
O
N
H
N
H
N
We became interested in cyclic peptides as potential
trypsin/chymotrypsin based serine protease inhibitors.
Published X-ray structures of peptides (or peptido-
mimetics) bound to the substrate binding domain of
serine proteases clearly show that a tether between the
P1 and P3 side chains can easily be accommodated. For
example, the X-ray structure of turkey ovomucoid
inhibitor complexed to elastase clearly shows that the
P1 and P3 side chains (Schecter and Berger4 nomencla-
O
O
X
P
1
X
P
X
X
O
O
X
O
H
N
H
N
O
N
H
N
H
N
O
O
X
,
ture) are separated by no more than 7.5 A and that the
2
approach is unimpeded by enzyme residues. In addi-
tion, P2 and P4 are oriented on the same face of the
‘substrate’ b-strand, facing away from the binding
groove of the enzyme.
X
X
O
O
X
O
H
N
H
N
To be useful, and to further extend the current state-of-
the-art we needed a synthesis, which was not only
N
H
N
H
Z
O
O
X
3
Keywords: cyclic peptide; Weinreb amide; peptidomimetic, thiazole,
cysteine and aldehyde.
* Corresponding author. Fax: 203-677-7702; e-mail: michael.a.walker
@bms.com
Figure 1. General plan for peptide and cyclic peptide serine
protease inhibitors (X=S, P=protecting group and Z=H or
heterocycle).
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01134-0

5802
M. A. Walker, T. Johnson / Tetrahedron Letters 42 (2001) 5801–5804
we needed to modify the method to allow the introduc-
tion of a ‘serine-trap’ at the C-terminus of the peptide.
To complicate matters further we desired a synthesis
which would save the C-terminal modification until the
end of the sequence in order to allow us the conve-
nience of examining SAR at this site without having to
repeat the whole route. To the best of our knowledge
this sequence has never been disclosed before pre-
sumably, as we were to learn, due to the poor solubility
of peptides in THF. Therefore, this paper not only
describes the synthesis of a number of cyclic peptides
but includes new methodology for their subsequent
transformation into potential serine protease inhibitors.
The synthesis of the macrocycle is demonstrated for
compound 10 (Scheme 2). In our first attempt at form-
ing this ring we intended to take advantage of the
orthogonal cysteine protecting groups by carrying out
the cyclization in a two-step sequence. Thus, the t-butyl
dissulfide group was removed with Bu3P and the result-
ing thiol reacted with 1,4-dibromobutane to yield 9.
The yield of this alkylation reaction was much lower
than expected due to the formation of the bis-alkylated
byproduct. Nonetheless this intermediate was carried
on to cyclic peptide 10 by first treating with TFA/
Et3SiH to remove the trityl protecting group followed
by DBU to effect the cyclization. Although unopti-
mized, the yield of this step was also disappointingly
low. In the end we found that we could achieve a
slightly better yield by removing both protecting groups
prior to forming the macrocycle. This latter method
was used to introduce a wide variety of linkers as
shown in Fig. 2.
A representative synthesis of the peptide cyclization
precursor is shown in Scheme 1. The synthesis com-
menced with the introduction of a Weinreb amine in
order to set up the C-terminal modification envisioned
later in the sequence. Standard Fmoc peptide synthesis
procedures were then used to introduce the remaining
amino acids in the sequence. The less than optimal yield
encountered in the transformation of 6 to 7 is due
principally to the formation of the piperizane
byproduct during the deprotection reaction. This side
reaction was eliminated when piperidine/DMA was
used to remove the Fmoc group during the synthesis of
subsequent analogues. The N-terminus was capped as
the corresponding benzoyl amide.
Modification of the C-terminus is shown is Scheme 3
for the synthesis of 12a and 12b.7 As mentioned previ-
ously, we found that the cyclic tripeptide was insoluble
in THF, the solvent of choice for air and moisture
sensitive organometallic reagents. This turned out to be
a problem with all of the peptides longer than three
residues, requiring a general solution. Fortunately, we
found that the addition of LiCl, according to the
STr
OH
STr
STr
O
a
b,c,
N
FmocHN
N
FmocHN
FmocHN
OMe
N
H
OMe
(100%)
(76%)
O
O
O
O
4
5
6
StBu
StBu
S
S
STr
STr
O
O
H
N
H
N
b,d
e,f
(97%)
N
N
N
H
N
H
OMe
FmocHN
N
H
OMe
(45%)
O
O
O
O
8
7
Scheme 1. (a) 1.25 equiv. (MeO)MeNH2Cl, 3.0 equiv. iPr2NEt, 1.5 equiv. HBTU, 1.0 equiv. BtOH·H2O DMA; (b) 1:1
Et2NH/CH2Cl2; (c) 1.5 equiv. Fmoc-Ala-OH·H2O, 3.0 equiv. iPr2NEt, 1.5 equiv. HBTU, 1.0 equiv. BtOH·H2O, DMA; (d) 1.25
equiv. Fmoc-Cys(StBu)-OH, 3.0 equiv. iPr2Net, 1.5 equiv. HBTU, 1.0 equiv. BtOH·H2O, DMA; (e) 50 equiv. piperidine, DMA;
(f) 1.2 equiv. BnCOCl, CH2Cl2, satd NaHCO3 (aq.).
StBu
S
STr
SH
STr
O
O
O
O
H
N
H
N
N
N
a
N
H
N
H
OMe
N
H
N
H
OMe
O
O
O
11
O
(87%)
8
a,b (X = Br)
(32%)
c,b (X = I)
(25%)
Br
S
STr
S
S
O
O
O
O
H
H
N
c,d
N
N
N
N
H
N
OMe
N
H
N
H
OMe
H
(17%)
O
O
O
O
9
10
Scheme 2. (a) 4.0 equiv. Bu3P, THF/H2O; (b) X-(CH2)4-X, DBU, PhH; (c) 2.0 equiv. Et3SiH, 3:1 CH2Cl2/TFA; (d) DBU, PhH.

M. A. Walker, T. Johnson / Tetrahedron Letters 42 (2001) 5801–5804
5803
S
S
S
O
O
O
O
S
S
H
N
O
O
O
O
S
S
H
N
N
N
H
N
H
N
N
N
H
N
H
O
O
O
O
O
O
13
14
S
S
S
S
H
N
H
N
N
H
N
H
R
N
H
N
H
O
O
15
16a, R = H
16b, R = thiazoly
N
N
O
N
S
S
S
S
O
S
O
O
H
H
N
N
N
R
N
N
H
N
H
N
H
H
O
O
O
O
18a, R = H
18b, R = thiazolyl
17a, R = H
17b, R = thiazolyl
S
S
S
S
O
O
H
S
O
O
H
S
N
N
H
N
N
N
H
N
N
H
N
H
O
O
O
O
19 isomer B
19 isomer A
S
S
S
S
SH
O
SH
O
O
O
O
N
O
O
O
N
H
N
H
N
H
H
N
N
N
S
S
N
H
N
N
H
N
N
H
H
H
O
H
O
O
O
20
21
Figure 2. Cyclic peptiditomimetics.
S
S
S
S
O
O
O
O
H
H
N
a or b
N
N
R
N
H
N
H
OMe
N
H
N
H
O
O
O
O
10
12a, R = H (96% yield)
S
12b, R =
(65% yield)
N
Scheme 3. (a) 5.0 equiv. LiAlH4, 5.0 equiv. LiCl; (b) 8.0 equiv. thiazole, 8.0 equiv. n-BuLi, 5.0 equiv. LiCl.
procedure first described by Seebach,8 resulted in the
dissolution of the tripeptide Weinreb intermediates.
This appears to be one of the few instances we are
aware of where formation of an activated carbonyl has
been accomplished as the last step in the synthesis of a
peptide-based serine protease inhibitor.9 Normally, the
activated carbonyl functionality is introduced prior to
the synthesis of the peptidic portion of the molecule.
Formation of the activated carbonyl at such an early
stage can potentially be problematic due to epimeriza-
tion at the corresponding a-position. In our case, only
one diastereomer was observed in the synthesis of 12a
and 12b. Moreover, the current methodology is more
versatile since a variety of C-terminal modifications can
be rapidly explored.
The same chemistry was applied to the synthesis of the
related tripeptide analogues shown in Fig. 2. In our
view the linker not only functions as a device for
conformational control but also serves as a platform
for probing regions adjacent to the substrate-binding
pocket. Thus, in the current study we have employed
linkers which contain aryl functionality such as phenyl
(14), pyridyl (17) and quinazolinyl (18). The cyclic

5804
M. A. Walker, T. Johnson / Tetrahedron Letters 42 (2001) 5801–5804
compounds 19 (isomers A and B) were formed as
separable diasteromers derived from a single reaction.
We believe these compounds to be atropisomers due to
rotational hindrance about the bond connecting the
two phenyl groups. Lastly we have been able to apply
the same chemistry to the synthesis of pentapeptide
analogues 20 and 21.
6. Szewczuk, Z.; Rebholz, K. L.; Rich, D. H. Int. J. Peptide
Protein Res. 1992, 40, 233.
7. (a) Procedure for the synthesis of 12a: Thiazole (23 mL,
0.32 mmol) was dissolved in 1.0 of THF under N2. The
resulting solution was cooled to −78°C and 130 mL of 2.5
M nBuLi (THF) added dropwise to form a slurry. In a
separate flask lithium chloride (8.0 mg, 0.19 mmol), dried
overnight at 150°C under vacuum, was combined with
peptide 10 (20.0 mg, 40.3 mmol) and dissolved in 0.5 mL
of THF. The peptide solution was added to the thiazole-
anion mixture by cannula and the resulting mixture
stirred for 45 min. Saturated NH4Cl was added to
quench the reaction. EtOAc was added and the mixture
transferred to a separatory funnel. The organic layer was
separated, dried over Na2SO4, filtered and solvent
removed under vacuum. The crude product was purified
by flash column chromatography (SiO2, 96:4 CH2Cl2/
EtOH) to yield 13 mg (65% yield) of 12a. (b) Procedure
for the synthesis of 12b: Peptide 10 (30 mg, 60 mmol)
was combined with lithium chloride (13 mg, 310 mmol)
which had been dried overnight at 150°C under vacuum,
and the mixture dissolved in 1.0 mL of THF. The solu-
tion was cooled to −78°C and 300 mL of 1 M LiAlH
(THF) added dropwise. The resulting mixture was stirred
for 0.5 h, then quenched with 1 M KHSO4. The reaction
was worked up as above to yield 25 mg (96% yield) of
12b.
In conclusion we have presented a method for the rapid
synthesis of a diverse set of cyclipeptides as potential
inhibitors of serine protease. Moreover, we have suc-
cessfully applied the peptide solubilizing method of
Seebach in the introduction of the C-terminal activated
carbonyl late in the synthetic sequence. It is interesting
to note that shortly after we completed our investiga-
tion a polymer bound version of the Weinreb amine
became commercially available (Novabiochem). This
could be of use in the synthesis of libraries based on
our methodology.
References
1. Lambert, J. N.; Mitchell, J. P.; Roberts, K. D. J. Chem.
Soc., Perkin Trans. 1 2001, 471.
2. Fairlie, D. P.; Abbenante, G.; March, D. P. Curr. Med.
Chem. 1995, 2, 654.
3. Tyndall, J. D. A.; Reid, R. C.; Tyssen, D. P.; Jardine, D.
K.; Todd, B.; Passmore, M.; March, D. R.; Pattenden, L.
K.; Bergman, D. A.; Alewood, D.; Hu, S.-H.; Alewood,
P. F.; Birch, C. J.; Martin, J. L.; Fairlie, D. P. J. Med.
Chem. 2000, 43, 3495.
8. Seebach, D.; Thaler, A.; Beck, A. K. Helv. Chim. Acta
1989, 72, 857.
9. Dragovich, P. S.; Zhou, R.; Webber, S. E.; Prins, T. J.;
Kwok, A. K.; Okano, K.; Fuhrman, S. A.; Zalman, L.
S.; Maldonado, F. C.; Brown, E. D.; Meador, III, J. W.;
Patick, A. K.; Ford, C. E.; Brothers, M. A.; Binford, S.
L.; Matthews, D. A.; Ferre, R. A.; Worland, S. T.
Bioorg. Med. Chem. Lett. 2000, 10, 45.
4. Schecter, I.; Berger, A. Biochem. Biophys. Res. Commun.
1967, 27, 157.
5. Mosberg, H. I.; Omnaas, J. R. J. Am. Chem. Soc. 1985,
107, 2986.
.