Application of ring-closing metathesis for the synthesis of macrocyclic peptidomimetics as inhibitors of HCV NS3 protease

Article abstract of DOI:10.1021/ol0711265

An efficient synthetic approach for the preparation of macrocyclic peptidomimetics for inhibition of HCV NS3 is presented. The macrocyclic core is built using ring-closing metathesis (RCM) of a tripeptidic diene. The presented approach allows the introduction of heteroatoms in strategic places along the macrocyclic ring. The methyl ester moiety in the RCM products was synthetically manipulated to install a ketoamide moiety via a Passerini reaction.

Full text of DOI:10.1021/ol0711265

ORGANIC
LETTERS
2007
Vol. 9, No. 16
3061-3064
Application of Ring-Closing Metathesis
for the Synthesis of Macrocyclic
Peptidomimetics as Inhibitors of HCV
NS3 Protease
Francisco Vela´zquez,* Srikanth Venkatraman,* Wanli Wu, Melissa Blackman,
Andrew Prongay, Viyyoor Girijavallabhan, Neng-Yang Shih, and
F. George Njoroge
Schering-Plough Research Institute, 2015 Galloping Hill Road, K-15-3 3545,
Kenilworth, New Jersey 07033-1300
francisco.Velazquez@spcorp.com; srikanth.Venkatraman@spcorp.com
Received May 15, 2007
ABSTRACT
An efficient synthetic approach for the preparation of macrocyclic peptidomimetics for inhibition of HCV NS3 is presented. The macrocyclic
core is built using ring-closing metathesis (RCM) of a tripeptidic diene. The presented approach allows the introduction of heteroatoms in
strategic places along the macrocyclic ring. The methyl ester moiety in the RCM products was synthetically manipulated to install a keto-
amide moiety via a Passerini reaction.
Hepatitis C (HCV) is a viral infection afflicting more than
3% of the world population. It is the leading cause of liver
transplants in the United States and if left untreated could
result in liver failure and hepatocellular carcinoma.1 HCV
NS3 protease is an enzyme which plays a pivotal role in
replication of the HCV virus.2 Inhibition of this enzyme has
proven effective in reducing viral loads in humans, and
considerable efforts by different research groups have been
directed toward development of HCV NS3 protease inhibi-
tors.3 We recently reported the development of SCH 503034,
a novel, orally bioavailable HCV NS3 protease inhibitor that
is currently undergoing clinical trials.4 In an effort to modify
the highly peptidic nature of SCH 503034, we designed a
new class of macrocyclic peptidomimetics as HCV NS3
protease inhibitors. These compounds were expected to be
more potent and possess improved pharmacokinetic proper-
ties.
In this paper, we describe a synthetic methodology for
preparation of these macrocyclic inhibitors of the HCV NS3
protease.5 This approach is efficient in the construction of
(3) See the following reviews and references therein: (a) Chen, S. H.;
Tan, S. L. Curr. Med. Chem 2005, 12, 2317. (b) Goudreau, N.; Montse, L.
B. Expert Opin. InVest Drugs 2005, 14, 1129.
(4) Venkatraman, S. et al. J. Med. Chem. 2006, 49, 6074.
(5) BILN 2061 was the first compound to undergo clinical trials. It is
also a macrocyclic inhibitor of HCV NS3 protease. See: (a) Faucher, A.
M. et al. Org. Lett. 2004, 6, 2901. (b) Llinas-Brunet, M. et al. J. Med.
Chem. 2004, 47, 1605.
(1) (a) Cohen, J. Science 1999, 285, 26. (b) Brown, R. S., Jr.; Gaglio, P.
J. LiVer Transplant. 2003, 9, S10.
(2) (a) Lindenbach, B. D.; Rice, C. M. Nature 2005, 436, 933. (b)
Bartenschlager, R.; Ahlborn-Laake, L.; Mous, J.; Jacobsen, H. J. Virol. 1993,
67, 3835.
10.1021/ol0711265 CCC: $37.00
© 2007 American Chemical Society
Published on Web 07/04/2007

macrocylic compounds with different ring sizes. Moreover,
the versatility of this methodology allows thorough explora-
tion of structure-activity relationships in different regions
of the molecules. Our goal was to investigate different classes
of compounds including macrocycles containing all-carbon
ring systems and to introduce heteroatoms in strategic places
along the macrocylic core as well. Our approach was based
on the application of ring-closing metathesis (RCM) for
construction of the key macrocyclic cores.6 This approach
has been employed for the preparation of other bioactive
macrocyclic peptidomimetics such as apicidin A and related
analogues for treatment of parasite-mediated infections7 and
inhibitors of BACE-1 for treatment of Alzheimer’s disease.8
Burk’s catalyst for asymmetric hydrogenation was employed
for the synthesis of the required amino acids, and the
Passerini reaction was used for the construction of a keto-
amide moiety required in HCV NS3 inhibitors.9
could be obtained from peptide couplings using ω-unsatur-
ated amino acids such as compound D, which in turn can be
obtained through asymmetric hydrogenation using Burk’s
catalyst.11
A. Synthesis of 15-, 16-, and 17-Membered Macrocycles
Containing an All-Carbon Aliphatic Chain.12 For the
preparation of macrocycles containing an all-carbon aliphatic
chain, we first synthesized ω-unsaturated N-Boc-protected
amino acid 5a and amine hydrochloride salt 6b (Scheme 1).13
Scheme 1. Synthesis of ω-Unsaturated Amino Acids
Our retrosynthetic analysis is shown in Figure 1. We
As mentioned above, synthesis of these intermediates
involved a highly efficient asymmetric hydrogenation of R,â-
unsaturated esters using Burk’s catalyst.14 The presence of
the ω-unsaturation in 2a and 2b presented an additional
challenge to the synthetic transformation since hydrogenation
had to proceed chemoselectively to avoid overreduction of
the substrates. Thus, Knoevenagel condensation of malonate
derived monoester 1 with pent-4-enal gave R,â-unsaturated
ester 2a in moderate yield (43%). Likewise, the reaction of
ester 1 with hex-5-enal gave R,â-unsaturated ester 2b (26%).
The chemo- and stereoselective asymmetric hydrogenation
of the conjugated olefinic bond in the presence of the
terminal olefin for compounds 2a and 2b proceeded with
excellent efficiency. Thus, hydrogenation using rhodium (Et-
DuPhos)OTf delivered the desired amino acids 3a and 3b
in nearly quantitative yields and high enantiomeric excess
(>98% ee). The amino acids 3a and 3b had the required
S-configuration at the newly created stereogenic center, and
their terminal olefin remained intact in the process. It is
important to mention that chemoselectivity in the hydrogena-
tion step was directed by the N-acetyl group in compounds
2a and 2b. The N-Boc-protected amino acids 4a and 4b were
obtained from 3a and 3b in 98 and 73% yields, respectively.
Finally, hydrolysis of the ethyl ester and N-acetyl func-
tionalities in 4a using lithium hydroxide gave the desired
Figure 1. Retrosynthetic analysis for HCV NS3 macrocyclic
inhibitors.
envisioned making inhibitors such as compound A, which
are 16-membered macrocycles (15- and 17-membered mac-
rocycles were also investigated), that contained an all-carbon
aliphatic chain or had an oxygen atom incorporated into the
macocyclic core. The keto-amide moiety, which acts as an
electrophilic serine trap, was installed via Passerini reaction
of the corresponding aldehyde derived from compound B.10
The macrocyclic core, which is the main feature of these
compounds, was obtained by hydrogenation of the ring-
closing metathesis product of diene C. The RCM precursor
(6) (a) Deiters, A.; Martin, S. F. Chem. ReV. 2004, 104, 2199. (b)
Tsantrizos, Y. S. J. Organomet. Chem. 2006, 691, 5163.
(7) Deshmukh, P. H.; Schulz-Fademrecht, C.; Procopiou, P. A.; Vigushin,
D. A.; Coombes, R. C.; Barrett, A. G. M. AdV. Synth. Catal. 2007, 349,
175.
(8) Ghosh, A. K. et al. Bioorg. Med. Chem. Lett. 2005, 15, 15.
(9) See the following examples and references therein for SAR of
electrophilic serine traps in HCV NS3 inhibitors: (a) Bennett, J. M., et al.
Bioorg. Med. Chem. Lett. 2001, 11, 355. (b) Han, W.; Hu, Z.; Jiang, X.;
Decicco, C. P. Bioorg. Med. Chem. Lett. 2000, 10, 711.
(10) (a) Passerini, M. Gazz. Chim. Ital. 1921, 51, 126. (b) Ugi, I.; Meyr,
R. Chem. Ber. 1961, 94, 2229.
(11) (a) Burk, M. J. J. Am. Chem. Soc. 1991, 113, 8518. (b) Burk, M. J.
Acc. Chem. Res. 2000, 33, 363.
(12) The synthesis of 16-membered macrocycles is shown in the
manuscript. See the Supporting Information for synthesis and characteriza-
tion of 15- and 17-membered macrocycles.
(13) For preparation of 5a using an alternative approach, see: (a) Ripka,
A. S.; Bohacek, R. S.; Rich, D. H. Bioorg. Med. Chem. Lett. 1998, 8, 357.
(b) Goudreau, N. et al. J. Org. Chem. 2004, 69, 6185.
(14) For examples of rhodium-catalyzed asymmetric hydrogenations,
see: Chi, Y.; Tang, W.; Zhang, X. Modern Rhodium-Catalyzed Organic
Reactions; Evans, P. A., Ed.; Wiley-VCH: Weinheim, 2005; pp 1-31.
3062
Org. Lett., Vol. 9, No. 16, 2007

Scheme 2. Synthesis of 16-Membered Macrocylic Compounds Containing an All-Carbon Aliphatic Chain
N-Boc-protected amino acid 5a in nearly quantitative yield.
On the other hand, N-acetyl cleavage followed by N-Boc
deprotection in compound 4b gave the corresponding amine
hydrochloride salt 6b in 68% yield.
B. Synthesis of 16-Membered Oxygen-Containing Mac-
rocycles. As mentioned above, it was in our interest to
investigate the introduction of heteroatoms in specific places
along the macrocyclic system. Introduction of heteroatoms
was aimed at changing properties such as solubility and
pharmacokinetics in our inhibitors. We envisioned applying
the RCM approach described above for preparation of these
oxygen-containing macrocyclic systems. Thus, Boc-L-serine
(16) was used as starting material for the synthesis of
macrocycle 22. Palladium-catalyzed O-allylation of 16 was
required to avoid erosion of enantiomeric purity. Thus,
treatment of 16 with methyl allyl carbonate and a catalytic
amount of tetrakis(triphenylphosphine) palladium afforded
17 in 64% yield. Saponification of the ester functionality
followed by HATU coupling with proline derivative 7 gave
dipeptide 19. The required diene 20 for ring-closing
metathesis was obtained in a two-step process similar to that
described above for the all-carbon macrocyclic analogues.19
Thus, the crucial macrocyclization of diene 20 was carried
out via RCM using Grubbs’ first-generation catalyst to give
the metathesis product as an inconsecuential mixture of cis/
trans isomers. This reaction represents a new example of
olefin metathesis for the construction of large heterocyclic
systems.
Hydrogenation of the metathesis product gave macrocycle
21 in 63% yield along with a small amount of a side product
arising from O-deallylation (ca. 15%). The methyl ester
functionality in 21 was transformed into the required keto-
amide moiety using the conditions previously described in
the all-carbon analogues. Thus, the oxygen-containing mac-
rocycle 22 was obtained in five steps from ester 21.
The inhibition activity against HCV NS3 of these mac-
rocycles was measured and found to be in the nanomolar
range.20 The all-carbon macrocycle 15 (Scheme 2) had a Ki*
) 36 nM, whereas the oxygen-containing analogue 22
Having on hand the required ω-unsaturated amino acids,
we began construction of the macrocyclic core of our
inhibitors. Synthesis of the macrocyclic core started with the
HATU coupling of ω-unsaturated acid 5a with dimethyl-
cyclopropyl proline 74,15 to give dipeptide 8 in 75% yield.
Hydrolysis of the methyl ester 8 gave acid 9 in nearly
quantitave yield. At this point, the ω-unsaturated amino acid
6b was introduced to install the required terminal double
bond for RCM giving tripeptide 10 in 88% yield. The crucial
RCM reaction of diene 10 was carried out using Grubbs’
first-generation catalyst16 to deliver the corresponding alkene
in 93% yield as a mixture of cis/trans isomers which were
hydrogenated to afford the desired 16-membered macrocycle
11 (95%). This approach proved to be remarkably efficient
in the preparation of other macrocycles, including 15- and
17-membered macrocycles, giving similar results.
After the successful synthesis of the macrocyclic core, we
proceeded to the construction of the keto-amide moiety which
serves as a serine trap in the class of inhibitors that were of
interest. Ethyl ester 11 was converted to aldehyde 12 in a
two-step sequence. Then, a Passerini reaction of aldehyde
12 and commercially available allyl isocyanide17 gave the
corresponding R-acetoxy amide 13 as an inconsequential
mixture of diastereomers. Hydrolysis of the acetate in 13
proceeded in almost quantitative yield to give allyl hy-
droxyamide 14. Finally, Dess-Martin18 oxidation of 14
delivered the desired 16-membered macrocyle 15 which
contains the required keto-amide serine trap.
(15) Zhang, R.; Mamai, A.; Madalengoitia, J. S. J. Org. Chem. 1999,
64, 547.
(16) (a) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H. Angew.
Chem. Int. Ed. 1995, 34, 2039. (b) Trnka, T. M.; Grubbs, R. H. Acc. Chem.
Res. 2001, 34, 18.
(17) Allyl isocyanide can be prepared by dehydration of N-allylformamide
following the procedure described for the preparation of cyclopropyl
isocyanide in: Scho¨llkopf, U. et al. Liebigs Ann. Chem. 1976, 1, 183.
(18) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(19) The methyl ester analogue of 6b was used in the synthesis of oxygen-
containing macrocycles.
(20) A manuscript describing the synthesis and structure-activity
relationship (potency and pharmacokinetic profiles) of macrocyclic inhibitors
using this approach is in preparation.
Org. Lett., Vol. 9, No. 16, 2007
3063

Scheme 3. Synthesis of 16-Membered Oxygen-Containing
Macrocycles
(Scheme 3) had a Ki* ) 150 nM.21 Moreover, experimental
data suggests that a 16-membered macrocycle might be the
optimum ring size for activity although 15-membered
compounds were also well accommodated. A thorough study
of the potency and pharmacokinetic profiles for these
macrocylic inhibitors was carried out and will be disclosed
in a separate publication. The synthetic strategy described
herein proved to be efficient and flexible for SAR investiga-
tions at both ends of the macrocyclic core.
Macrocyclic keto-amide 23 was synthesized using the
RMC approach described herein.20 This compound is a 17-
membered macrocycle with Ki* ) 6 nM. Figure 2 shows
compound 23 bound to HCV NS3 protease, and several key
interactions can be observed. From the X-ray, it was clear
that the aliphatic chain of the macrocylic core made excellent
contact in the lipophilic region of the active site. Moreover,
the keto-amide moiety adopts a conformation which facili-
tates nuclephilic attack by Ser-139 leading to enzyme
inhibition.
Figure 2. Macrocyclic keto-amide 23. A 17-membered all-carbon
macrocycle bound to the active site of HCV NS3 protease.
NS3 was developed using RCM. The flexibility of this
approach allowed the synthesis of 15-, 16-, and 17-membered
macrocycles from easily accessible intermediates. A simple
modification of this methodology allowed introduction of
heteroatoms in strategic places of the macrocylic system.
Also, required ω-unsaturated amino acid starting materials
were efficiently synthesized using a chemoselective asym-
metric hydrogenation. The Passerini reaction allowed instal-
lation of the crucial keto-amide moiety present in this class
of inhibitors.
Acknowledgment. We thank John Pichardo, Sony Agraw-
al, Andrea Hart, and Xiao Tong, Department of Virology,
Schering-Plough Research Institute, for their support in this
project.
In conclusion, a highly efficient approach for the synthesis
of macrocyclic peptidomimetics for the inhibition of HCV
Supporting Information Available: Experimental pro-
cedures, spectroscopical data, and copies of spectra for
selected compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
(21) The HCV NS3 serine protease inhibitory value is provided as Ki*.
For a definition of Ki* and discussion, see: Morrison, J. F.; Walsh, C. T.
AdVances in Enzymology and Related Areas of Molecular Biology; Meister,
A., Ed.; John Wiley & Sons: Hoboken, NJ, 1988; Vol. 61, pp 201.
OL0711265
3064
Org. Lett., Vol. 9, No. 16, 2007