Bismacrocyclic inhibitors of hepatitis C NS3/4a protease

Article abstract of DOI:10.1002/anie.200803298

(Chemical Equation Presented) Double time: The bismacrocycle 2 was prepared from 1 by a selective double ring-closing metathesis (RCM) reaction to form the 18- and 15-membered rings simultaneously. Derivative 3 shows excellent potency against NS3/4a protease.

Full text of DOI:10.1002/anie.200803298

Communications
DOI: 10.1002/anie.200803298
Ring-Closing Metathesis
Bismacrocyclic Inhibitors of Hepatitis C NS3/4a Protease**
John A. McCauley,* Michael T. Rudd, Kevin T. Nguyen, Charles J. McIntyre, Joseph J. Romano,
Kimberly J. Bush, Sandor L. Varga, Charles W. Ross, III, Steven S. Carroll, Jillian DiMuzio,
Mark W. Stahlhut, David B. Olsen, Terry A. Lyle, Joseph P. Vacca, and Nigel J. Liverton
Hepatitis C virus (HCV) is a chronic infection that affects an
estimated 170–200 million people worldwide.[1] The positive-
RNA-strand virus of the Flaviviridae family replicates
primarily in the liver, and although disease progression is
typically a slow process, a significant fraction of those infected
develop serious liver disease, including cirrhosis and hepato-
cellular carcinoma.[2] HCV is currently a leading cause of
death in HIV-coinfected patients[3] and is the most common
indication for liver transplantation.[4] Of several promising
antiviral targets for HCV that have emerged in recent years,[5]
NS3/4a protease inhibitors show perhaps the most dramatic
antiviral effects.[6] Clinical proof of concept for this mecha-
nism was first established with BILN-2061 (1, Scheme 1).[7]
We recently disclosed our molecular-modeling-derived
strategy that led us to design HCV NS3/4a protease inhibitors,
such as 2, which contain a P2–P4 macrocyclic constraint
(Scheme 1).[8] These compounds were prepared by using a
ring-closing metathesis (RCM) reaction as the key step, and
further refinement of this series resulted in the synthesis of
compound 3. With the success of this alternative macrocycle
design strategy, we were intrigued by the possibility of
accessing a compound that contained both the P2–P4 macro-
cycle found in compounds 2 and 3 and the P1–P3 macrocycle
seen in compounds such as 1.
Scheme 1. Macrocyclic NS3/4a protease inhibitors.
The bismacrocycle 4 (Scheme 1) is an interesting target
from a biological perspective and, we quickly realized, a
formidable synthetic challenge. Compound 4 contains both
18- and 15-membered-ring macrocycles, as well as a trans
alkene and a cis alkene. Our synthetic planning began with
the retrosynthetic analysis of ester 5 (Scheme 2), which could
be transformed in a straightforward manner into the ultimate
target 4. Initial disconnection of the P1–P3 macrocycle gives
6, which could theoretically be transformed into 5 in three
steps. The synthesis of compound 6, however, would be
[*] Dr. J. A. McCauley, Dr. M. T. Rudd, K. T. Nguyen, C. J. McIntyre,
J. J. Romano, K. J. Bush, S. L. Varga, Dr. C. W. Ross, III, Dr. T. A. Lyle,
Dr. J. P. Vacca, Dr. N. J. Liverton
Department ofMedicinal Chemistry, Merck Research Laboratories
West Point, PA 19486 (USA)
Fax: (+1)215-652-3971
E-mail: john_mccauley@merck.com
Dr. S. S. Carroll, J. DiMuzio, M. W. Stahlhut, Dr. D. B. Olsen
Department ofAntiviral Research, Merck Research Laboratories
(USA)
Scheme 2. Retrosynthetic analysis ofthe bismacrocyclic core. Boc =
tert-butoxycarbonyl.
[**] We thank Professor Samuel J. Danishefsky and Professor Barry M.
Trost for helpful discussions and their valuable insight. We also
thank Joan S. Murphy for high-resolution mass spectra.
nontrivial, as it would require selective RCM of a triene,
protection of the terminal alkene contained in the P3 side
chain, or exploration of non-RCM synthetic routes. Alter-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200803298.
9104
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9104 –9107

Angewandte
Chemie
natively, initial disconnection of the P2–P4 macrocycle gives
the more readily accessible target 7. We felt that this route
would most likely deliver the target 4, but only after a mostly
linear sequence of more than 10 steps. A third possibility
would be to form both rings simultaneously through a double
ring-closing metathesis reaction of the tetraene 8. This route
was attractive to us, because 8 could be synthesized in a
convergent manner by formation of the central amide bond
from readily available fragments. Furthermore, each macro-
cycle had been formed independently through RCM reactions
previously, and the alkene geometry had been controlled in
both cases.[7a,8]
Although this planned synthesis would be a very con-
vergent and rapid approach to a difficult target, the double
RCM reaction could have a number of selectivity issues that
would make the isolation of the desired product technically
difficult, if indeed it was formed. For example, an alternative
cyclization of 8 may lead to compound 9 (Scheme 3), which
Scheme 4. Synthesis of 4 through a selective bis-RCM reaction:
a) HATU, DIPEA, DMF, 79%; b) Zhan 1b catalyst,[11] DCE, 1 mm, 708C,
1 h, 66%; c) LiOH, H2O, THF, EtOH, 86%; d) CDI, THF; cyclopropyl-
sulfonamide, DBU, 408C, 68%. HATU=O-(7-azabenzotriazol-1-yl)-
N,N,N’,N’-tetramethyluronium hexafluorophosphate, DIPEA=N,N-dii-
sopropylethylamine, DMF=N,N-dimethylformamide, DCE=1,2-
dichloroethane, CDI=1,1’-carbonyldiimidazole, DBU=1,8-
diazabicyclo[5.4.0]undec-7-ene.
Scheme 3. Potential alternative cyclization modes for the bismacro-
cyclization reaction.
were formed preferentially over at least nine other possible
products with macrocyclic rings of similar sizes. These
potential products include the undesired alkene isomers of 5
and the six compounds represented by structures 9 and 10
(Scheme 3).
We then undertook optimization studies of the bis-RCM
step. When the dilution of the reaction mixture was increased
to 1 mm, compound 5 was obtained in a remarkable yield of
66%, and the approximately 10:1 selectivity was maintained.
Presumably, this high dilution leads to a decrease in the
quantity of oligomeric products that may have been formed
under our standard reaction conditions at a concentration of
5 mm.[13] Compound 5 was then hydrolyzed to the carboxylic
acid 13, which was coupled to cyclopropylsulfonamide[14] to
complete the target 4.
contains 16- and 17-membered rings; in this case, the alkene
geometry in both rings would be in question. Another
possibility we considered was the initial reaction of the two
alkene functionalities at the termini of the more flexible
carbon chains to give the 14-membered ring present in
compound 10. A recent review of metathesis reactions of tri-
and tetraene substrates has shown the utility of these
reactions in the synthesis of certain natural products, although
many of the examples involve the formation of smaller rings,
and the potential selectivity issues that may occur with 8 do
not apply.[9]
In practice, compounds 11[10] and 12[10] could be coupled
under standard conditions in high yield to give the bis-RCM
precursor 8 (Scheme 4). When 8 was exposed to the Zhan 1b
catalyst[11] and p-benzoquinone[12] in 1,2-dichloroethane
(5 mm) at 708C for 1 h, a single major peak was observed
When compared with the P2–P4 monomacrocycles 2 and 3
and the P1–P3 monomacrocycle 1, bismacrocycle 4 maintains
excellent enzyme potency against the genotype 1b NS3/4a
1
[15]
by HPLC, and one compound was isolated in 45% yield. H
protease (Table 1).
Compound 4, however, showed
and 13C NMR spectroscopic analysis, including 2D experi-
ments, confirmed that the major reaction product was the
desired bismacrocycle 5.[10] HPLC–MS analysis of the reac-
tion mixture showed the presence of three other products
with masses consistent with bismacrocyclic structures; how-
ever, 5 was formed with approximately 10:1 selectivity over
each of these compounds. Thus, the 18- and 15-membered
rings of 5, which contain a trans and a cis alkene, respectively,
improved potency in the cellular replicon assay under both
high- and low-serum conditions.[16] Notably, in comparison to
compound 3, the addition of the P1–P3 macrocycle in
compound 4 leads to a fivefold increase in potency against
the genotype 3a NS3/4a protease enzyme. Compound 13,
which lacks the potency-enhancing acyl sulfonamide, also
maintains good activity, comparable to that of carboxylic acid
1, against the genotype 1b enzyme.
Angew. Chem. Int. Ed. 2008, 47, 9104 –9107ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9105

Communications
Table 1: Inhibitory activity against HCV NS3/4a protease.[a]
Furthermore, the conversion of 8 into 5 seems to generate
fewer RCM side products, which indicates perhaps that
selectivities are reinforced through a form of double chemo-
selection.
Another possibility for the highly selective formation of 5
from 8 is that 5 could be the thermodynamic product, and
intermediate compounds equilibrate to product 5 under the
reaction conditions. We initially thought that compound 10
would be a likely product of the RCM reaction of 8 because of
the flexibility of the chains containing the relevant alkenes,
although, in practice, we did not detect intermediate products
in the cyclization of 8.[17] To probe this possibility, we prepared
10 as shown in Scheme 6. Compound 18 was synthesized in a
Compound Ki(genotype 1b) Ki(genotype 3a) EC50(1b replicon) [nm]
[nm]
0.30
<0.016
0.034
0.029
0.13
[nm]
10% FBS/50% NHS
1
2
3
4
13
–
35
6.8
1.2
140
3/19
13/25
3/14
2/10
10/340
[a] Data are geometric averages ofthe values ofr at least three
measurements. FBS=fetal bovine serum, NHS=normal human serum.
In addition to the excellent biological profile of 4, we were
fascinated by the remarkable selectivity of the bis-RCM
reaction. We designed two competition experiments to
explore the selectivity of the individual ring-forming reactions
(Scheme 5). We first studied the reactivity of triene 14, which
Scheme 6. Equilibration experiment: a) Zhan 1b catalyst,[11] DCE,
=
5 mm, 708C, 1 h, 95%; b) (CH2 CH)SnBu3, [Pd(PPh3)4], toluene,
1008C, 74%; c) LiOH, H2O, THF, EtOH; d) HATU, DIPEA, 20, DMF,
60% (2 steps); e) Zhan 1b catalyst, DCE, 5 mm, 708C, 92 h, 26%.
standard fashion, and the 14-membered ring was formed by
RCM to give 19 as a mixture of alkene isomers in good yield.
A three-step sequence of vinylation, hydrolysis, and amide
coupling with 20[18] then yielded compound 10. The extended
exposure of 10 to RCM conditions, including the addition of
up to 20 mol% of the Zhan 1b catalyst, led to bismacrocycle 5
as the major product. This result demonstrates that 5 is likely
to be the thermodynamic product.[19] Since the equilibration
of 10 to 5 requires extended heating, as opposed to the direct
conversion of tetraene 8 into 5 (92 h versus 1 h), we do not
believe that compound 10 is an intermediate in the conversion
of 8 into 5.[10] This result is particularly noteworthy, as the 14-
membered ring in 19 is formed quickly and efficiently (1 h,
95% yield) from a substrate, 18, with only two alkene
functionalities.
Scheme 5. Competition experiments: a) Zhan 1b catalyst,[11] p-benzo-
quinone, DCE, 5 mm, 708C, 1 h; 15: 38%; 17: 42%.
lacks the styryl alkene to eliminate the possibility of forming
the P2–P4 macrocycle. Under standard RCM conditions,
compound 15 was formed as the major product with greater
than 10:1 selectivity over four other RCM products, as
determined by HPLC–MS analysis. Thus, the desired P1–P3
macrocyclization was favored.[10] A complementary experi-
ment with 16, which lacks the alkene in the P1 side chain, also
showed selectivity for the desired macrocycle, 17, under the
RCM conditions; however, this reaction is slightly less
selective (> 6:1 selectivity over five other products).[10]
These experiments demonstrate that the remarkable selec-
tivity observed in the formation of 5 is not a result of a driving
of overall efficiency by a single chemoselective reaction, but is
instead dictated by the high degree of chemoselectivity in the
formation of both the P1–P3 and P2–P4 macrocycles.
In summary, we have synthesized the bismacrocyclic HCV
NS3/4a protease inhibitor 4 through a highly convergent
synthetic sequence of amide-bond formation followed by a
chemo- and stereoselective double RCM reaction to form the
18- and 15-membered rings simultaneously. Compound 4
9106 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9104 –9107

Angewandte
Chemie
Zeng, M.-H. Hao, X. Wei, N. K. Yee, C. A. Busacca, Z. Han, V.
shows excellent activity against both the genotype 1b and the
genotype 3a NS3/4a enzymes as well as very good cellular
activity. Further studies on this unique class of NS3/4a
protease inhibitors and into the origins of the bismacrocyc-
lization selectivity are ongoing.
Farina, C. H. Senanayake, Org. Lett. 2008, 10, 1303 – 1306.
[8] N. J. Liverton, M. K. Holloway, J. A. McCauley, M. T. Rudd,
J. W. Butcher, S. S. Carroll, J. DiMuzio, C. Fandozzi, K. F.
Gilbert, S.-S. Mao, C. J. McIntyre, K. T. Nguyen, J. J. Romano,
M. Stahlhut, B.-L. Wan, D. B. Olsen, J. P. Vacca, J. Am. Chem.
Soc. 2008, 130, 4607 – 4609.
[9] a) B. Schmidt, J. Hermanns, Curr. Org. Chem. 2006, 10, 1363 –
1396; for examples of diastereoselective RCM reactions with tri-
and tetraene substrates, see: b) C. M. Huwe, J. Velder, S.
Blechert, Angew. Chem. 1996, 108, 2542 – 2544; Angew. Chem.
Int. Ed. Engl. 1996, 35, 2376 – 2378; c) M. Lautens, G. Hughes,
Angew. Chem. 1999, 111, 160 – 162; Angew. Chem. Int. Ed. 1999,
38, 129 – 131; d) D. J. Wallace, P. G. Bulger, D. J. Kennedy, M. S.
Ashwood, I. F. Cottrell, U.-H. Dolling, Synlett 2001, 357 – 360;
e) P. A. Evans, J. Cui, G. P. Buffone, Angew. Chem. 2003, 115,
1776 – 1779; Angew. Chem. Int. Ed. 2003, 42, 1734 – 1737;
f) A. M. Hooper, S. Dufour, S. Willaert, S. Pouvreau, J. A.
Pickett, Tetrahedron Lett. 2007, 48, 5991 – 5994.
[10] See the Supporting Information for synthetic procedures and
spectral data for 4, 5, 8, 11, 12, 13, 15, and 17, proton assignments
for 5, and HPLC traces of the bismacrocyclization reaction
mixture and for the competition and equilibration experiments.
[11] The Zhan 1b catalyst is available from Strem, catalogue number
44-0082.
Experimental Section
5: A solution of 8 (50 mg, 0.06 mmol) in 1,2-dichloroethane (61 mL;
1 mm) was purged with nitrogen gas for 10 min. The Zhan 1b catalyst
(4.5 mg, 0.006 mmol) was then added, and the reaction mixture was
heated to 708C and stirred for 1 h. The reaction mixture was then
cooled, concentrated, and purified by silica-gel chromatography
(gradient elution, 10–100% EtOAc in hexane) to provide 5 (31 mg,
1
66%). H NMR (600 MHz, CD3CN): d = 8.22 (s, 1H), 8.20–8.18 (m,
2H), 7.55–7.52 (m, 2H), 7.50–7.46 (m, 1H), 7.33 (s, 1H), 7.26 (br s,
1H), 7.24 (s, 1H), 6.76 (d, J = 16.1 Hz, 1H), 6.36 (ddd, J = 16.1, 10.5,
5.6 Hz, 1H), 6.26 (d, J = 8.3 Hz, 1H), 5.49 (apparent t, J = 3.0 Hz,
1H), 5.44 (apparent q, J = 8.4 Hz, 1H), 5.18 (apparent t, J = 9.8 Hz,
1H), 4.75 (dd, J = 11.5, 1.6 Hz, 1H), 4.53 (ddd, J = 11.1, 8.5, 3.2 Hz,
1H), 4.45 (d, J = 11.1 Hz, 1H), 4.37 (dd, J = 9.3, 7.7 Hz, 1H), 4.02–
3.96 (m, 2H), 3.99 (s, 3H), 3.89 (dd, J = 11.5, 2.8 Hz, 1H), 3.31 (d, J =
11.1 Hz, 1H), 2.63 (ddd, J = 14.3, 7.5, 1.6 Hz, 1H), 2.47–2.39 (m, 2H),
2.33 (dd, J = 13.5, 5.0 Hz, 1H), 2.19–2.14 (obscured m, 1H), 2.03 (dd,
J = 13.5, 10.5 Hz, 1H), 1.84–1.79 (m, 2H), 1.61–1.54 (m, 1H), 1.51 (dd,
J = 9.5, 5.0 Hz, 1H), 1.47 (dd, J = 8.5, 5.0 Hz, 1H), 1.45–1.36 (m, 3H),
1.28–1.14 (m, 3H), 1.11 (s, 3H), 1.10 (s, 3H), 0.86 (s, 3H) ppm; HRMS
(ESI): m/z calcd for C44H53N4O8: 765.3858 [M+H]+; found: 765.3880.
Received: July 7, 2008
Revised: August 28, 2008
Published online: October 16, 2008
Keywords: hepatitis C · inhibitors · macrocycles ·
.
ring-closing metathesis · ruthenium
[12] S. H. Hong, D. P. Sanders, C. W. Lee, R. H. Grubbs, J. Am.
Chem. Soc. 2005, 127, 17160 – 17161.
[1] WHO, Weekly Epidemiology Record 1999, 74, 425 – 427.
[2] T. J. Liang, T. Heller,Gastroenterology 2004, 127, S62 – S71.
[3] D. Salmon-Ceron, C. Lewden, P. Morlat, S. BØvilacqua, E.
Jougla, F. Bonnet, L. HØripret, D. Costagliola, T. May, G. ChÞne,
J. Hepatol. 2005, 42, 799 – 805.
[13] High-resolution LC–MS/MS analysis (Bruker apex-Qe FTICR
MS) of the crude 5 mm reaction mixture confirmed the presence
of at least one dimeric product.
[14] J. A. Campbell, A. Good, WO2002/060926, 2002.
[15] S.-S. Mao, J. DiMuzio, C. McHale, C. Burlein, D. B. Olsen, S. S.
Carroll, Anal. Biochem. 2008, 373, 1 – 8.
[4] R. S. Brown, Nature 2005, 436, 973 – 978.
[5] C. P. Gordon, P. A. Keller, J. Med. Chem. 2005, 48, 1 – 20.
[6] For recent reviews, see: a) J. A. Thomson, R. B. Perni, Curr.
Opin. Drug Discovery Dev. 2006, 9, 606 – 617; b) S.-H. Chen, S.-
L. Tan, Curr. Med. Chem. 2005, 12, 2317 – 2342.
[16] G. Migliaccio, J. E. Tomassini, S. S. Carroll, L. Tomei, S.
Altamura, B. Bhat, L. Bartholomew, M. R. Bosserman, A.
Ceccacci, L. F. Colwell, R. Cortese, R. De Francesco, A. B.
Eldrup, K. L. Getty, X. S. Hou, R. L. LaFemina, S. W. Ludmerer,
M. MacCoss, D. R. McMasters, M. W. Stahlhut, D. B. Olsen, D. J.
Hazuda, O. A. Flores, J. Biol. Chem. 2003, 278, 49164 – 49170.
[17] In certain cases related to the synthesis of BILN-2061 (1), it has
been shown that the RCM process is initiated on the vinyl-
cyclopropane double bond; see reference [7d] and X. Zeng, X.
Wei, V. Farina, E. Napolitano, Y. Xu, L. Zhang, N. Haddad, N. K.
Yee, N. Grinberg, S. Shen, C. H. Senanayake, J. Org. Chem. 2006,
71, 8864 – 8875.
[18] P. L. Beaulieu, J. Gillard, M. D. Bailey, C. Boucher, J.-S.
Duceppe, B. Simoneau, X.-J. Wang, L. Zhang, K. Grozinger, I.
Houpis, V. Farina, H. Heimroth, T. Krueger, J. Schnaubelt, J.
Org. Chem. 2005, 70, 5869 – 5879.
[19] For the use of a thermodynamically controlled RCM strategy in
natural product synthesis, see: A. B. Smith, III, C. M. Adams, S.
A. Kozmin, D. V. Paone, J. Am. Chem. Soc. 2001, 123, 5925-5937.
[7] a) M. Llinàs-Brunet, M. D. Bailey, G. Bolger, C. Brochu, A.-M.
Faucher, J. M. Ferland, M. Garneau, E. Ghiro, V. Gorys, C.
Grand-Maître, T. Halmos, N. Lapeyre-Paquette, F. Liard, M.
Poirier, M. RhØaume, Y. S. Tsantrizos, D. Lamarre, J. Med.
Chem. 2004, 47, 1605 – 1608; b) D. Lamarre, P. C. Anderson, M.
Bailey, P. Beaulieu, G. Bolger, P. Bonneau, M. Bös, D. R.
Cameron, M. Cartier, M. G. Cordingley, A.-M. Faucher, N.
Goudreau, S. H. Kawai, G. Kukolj, L. LagacØ, S. R. LaPlante, H.
Narjes, M.-A. Poupart, J. Rancourt, R. E. Sentjens, R. St.
George, B. Simoneau, G. Steinmann, D. Thibeault, Y. S.
Tsantrizos, S. M. Weldon, C.-L. Yong, M. Llinàs-Brunet,
Nature 2003, 426, 186 – 189; for a large-scale synthesis of
BILN-2061, see: c) T. Nicola, M. Brenner, K. Donsbach, P.
Kreye, Org. Process Res. Dev. 2005, 9, 513 – 515; for recent
optimization of the synthesis of BILN-2061, see: d) C. Shu, X.
Angew. Chem. Int. Ed. 2008, 47, 9104 –9107ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9107