Olefin ring-closing metathesis as a powerful tool in drug discovery and development - potent macrocyclic inhibitors of the hepatitis C virus NS3 protease

Article abstract of DOI:10.1016/j.jorganchem.2006.09.027

Peptidomimetic inhibitors of the hepatitis C NS3 protease often exhibit poor biopharmaceutical properties. Structure modification of a substrate-based tripeptide into a β-strand 15-membered ring scaffold provided a new class of peptidomimetics that are si

Full text of DOI:10.1016/j.jorganchem.2006.09.027

Journal of Organometallic Chemistry 691 (2006) 5163–5171
www.elsevier.com/locate/jorganchem
Olefin ring-closing metathesis as a powerful tool in drug discovery
and development – potent macrocyclic inhibitors of the hepatitis C
virus NS3 protease
a,
a
a
Youla S. Tsantrizos *, Jean-Marie Ferland , Andrew McClory a,1, Martin Poirier ,
b,
b
b
b
b
Vittorio Farina *, Nathan K. Yee , Xiao-jun Wang , Nizar Haddad , Xudong Wei ,
b
b
Jinghua Xu , Li Zhang
a
Boehringer Ingelheim (Canada) Ltd., Research and Development, 2100 Cunard Street, Laval, Que., Canada H7S 2G5
Boehringer Ingelheim Pharmaceuticals, Department of Chemical Development, 900 Ridgebury Road, Ridgefield, CT 06877, USA
b
Received 31 August 2006; accepted 12 September 2006
Available online 19 September 2006
Abstract
Peptidomimetic inhibitors of the hepatitis C NS3 protease often exhibit poor biopharmaceutical properties. Structure modification of
a substrate-based tripeptide into a b-strand 15-membered ring scaffold provided a new class of peptidomimetics that are significantly
superior as drug candidates to their acyclic precursors. Tripeptide dienes composed of three unnatural amino acid residues with numer-
ous chiral centers were efficiently converted to macrocyclic peptides, in high diastereomeric purity, using ring-closing metathesis (RCM).
The conformation of the acyclic diene and the protocol for the RCM reaction were investigated and optimized extensively in order to
achieve an efficient synthesis of potential therapeutic agents for the treatment of hepatitis C infections. These studies provided the fist
small molecule (BILN 2061) that was clinically validated for the treatment of hepatitis C infection in man and opened the door to a
plethora of new pre-clinical pharmaceutical agents that can be made in multi kilogram quantities using RCM chemistry.
Ó 2006 Elsevier B.V. All rights reserved.
1. Introduction
cules, short peptides are conformationally heterogeneous
and exhibit low affinity for their intended binding sites.
In addition, these peptides often suffer from poor meta-
bolic stability and low bioavailability, further magnifying
the challenge of designing a small peptidomimetic com-
pound as a therapeutic drug. Nonetheless, during the last
decade, considerable effort has been devoted to the design
and synthesis of substrate-based peptidomimetic inhibitors
that target specifically the NS3 serine protease enzyme of
the hepatitis C virus (HCV). HCV is a severe medical prob-
lem with a global infection afflicting more than 170 million
people that can cause cirrhosis of the liver and liver cancer.
Therefore, disrupting the catalytic function of its essential
enzyme NS3 protease can provide a treatment option of
significant socioeconomic benefit.
Small molecules which can attenuate the function of a
biological process by mimicking the secondary structure
of proteins, or the critical molecular recognition element
(structural hot spots) between polypeptides and proteins,
are of enormous scientific interest [1]. Such interactions
play a key role in numerous cellular functions (e.g. protein
glycosylation [2], intracellular recognition [3], proteolysis
and others) and can be of particular relevance in drug dis-
covery. However, unlike the structurally unique peptides
found at the contact points between two large biomole-
*
Corresponding authors.
E-mail addresses: ytsantrizos@lav.boehringer-ingelheim.com (Y.S.
Initially, the quest for the discovery of a substrate-
based, small-molecule inhibitor targeting the HCV NS3
Tsantrizos), farinav@rdg.boehringer-ingelheim.com (Vittorio Farina).
1
Current address: Chemistry Department, Stanford University.
0022-328X/$ - see front matter Ó 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.09.027

5164
Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
serine protease was taken on by a number of pharmaceuti-
cal companies. However, to the dismay of medicinal chem-
ists, the shallow and solvent-exposed active site of this
enzyme proved to be an exceedingly difficult target for drug
design and many companies abandoned their efforts. At
Boehringer Ingelheim, medicinal chemistry activities were
initiated by the discovery that N-terminal peptide prod-
ucts, derived from sequences of the NS3 protease polypep-
tide substrate, were competitive inhibitors of the HCV NS3
protease [4]. Hexapeptides, such as model compound 1
(Fig. 1), were synthesized and used as NMR probes to
investigate the critical interactions between the active site
of the enzyme and its polypeptide substrate [5]. Armed with
the insight provided by these studies, we designed macrocy-
clic b-strand tripeptides (e.g. compounds 3b–d) that adopt
in the free state the enzyme-bound secondary structure of
the hexapeptides and bind tightly to the active site of the
enzyme [6]. The key ring-closing step was performed on
the acyclic tripeptide dienes 2 using RCM chemistry. While
this work was in progress, other investigators also reported
similar efforts towards the design of macrocyclic peptides
as potential inhibitors of metallo, aspartic, cysteine, and
serine proteases [7].
of structurally complex macrocyclic peptides is often very
challenging and the synthetic methodologies employed usu-
ally hinder the development of such compounds into com-
mercial products. Therefore, following our initial discovery
of macrocyclic compounds inhibiting the HCV NS3 prote-
ase [6], two of our key objectives were (a) the optimization
of the biopharmaceutical properties and (b) the develop-
ment of a synthetic process that was economically feasible
and could support clinical trials of a therapeutic agent. In
this report, we examine the effects of substitution along
the hydrocarbon linker moiety of the prototype compound
3d, in light of the structural data previously published for
this compound [6]. In addition, we provide an account of
the synthetic developments that led to the first pharmaceu-
tical application of ring-closing metathesis in multi-kilo-
gram scale.
2. Results and discussion
2.1. Medicinal chemistry
Over the last 10 years, the developments of stable orga-
nometallic reagents that catalyze ring-closing metathesis
(RCM) reactions have revolutionized the field of organic
synthesis [8]. These reagents provide new means of synthe-
sizing polyfunctional, large heterocyclic rings that are
exceptionally challenging to prepare by any other synthetic
methods. Peptide scaffolds, such as those typified by the
HCV inhibitor 3d, were originally synthesized using the
first-generation Grubbs catalyst, bis-(tricyclohexyl-phos-
phine)benzylidene-ruthenium dichloride (4) [9]. This
The design of macrocyclic drugs which are peptidic in
nature is undoubtedly inspired by nature’s non-ribosomally
formed secondary metabolites, which are often composed
of unusual amino acid fragments and are constrained into
cyclic peptides with unique bioactive conformations; well-
known pharmaceuticals include cyclosporin, vancomycin
and chloropeptin. However, unlike the production of natu-
ral products via fermentation, the chemical manufacturing
O
CO2H
O
O
O
H
H
N
N
N
N
N
H
H
NH
O
O
O
CO2H
CO2H
Substrate-based hexapeptide inhibitor 1
H3CO
N
H3CO
N
O
R2
O
H
H
CO2R1
N
CO2H
N
N
N
H
O
O
O
N
R3
O
O
O
Tripeptide diene 2
Macrocyclic prototype 3d (Boehringer Ingelheim)
Fig. 1. Substrate-based peptidomimetic inhibitors of HCV NS3 protease.

Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
5165
15-membered macrocyclic scaffold is characterized by the
(1R,2S)-vinyl aminocyclopropyl-carboxylic acid moiety
(vinyl ACCA moiety; P1) [10], a proline unit substituted
at Cc with a 4-hydroxy-7-methoxyquinoline (P2) [11] and
a linker moiety composed of (2S)-N-Boc-amino-8-none-
noic acid (P3) [9,12,13]. The backbone conformation of
the acyclic diene precursor 2, and in particular the ratio
of the proline cis:trans rotamers, was found to be crucial
for the yield of the RCM reaction and diastereomeric pur-
ity of the macrocyclic product(s) formed [14]. Recently, we
reported that when the Ca of the olefinic ‘‘linker’’ moiety
(P3) was unsubstituted (Scheme 1: diene 2a, R3 = H), the
cis:trans rotamer ratio of the P2–P3 amide bond was
approximately 1:1 and the outcome of the RCM reaction
was very poor (mixtures of product were obtained with
<30% overall RCM conversion) [14]. In contrast, when
the bulky –NHBoc group was attached to the Ca of the
olefinic linker (Scheme 1: diene 2b, R3 = NHBoc), the cis:-
trans ratio was approximately 1:9 and the overall conver-
sion of 2b to macrocyclic ester 3b was achieved in 40%
yield (ꢀ60% of total RCM conversion) [14]. Variable tem-
in 80% yield to the desired diastereomer 3c (Scheme 1).
These observations were analogous to those initially
reported by Grubbs and coworkers on template-directed
RCM reactions [15] and more recently by Lubell’s group
for the RCM macrocyclization of dipeptides having sec-
ondary or tertiary amide bonds [16]. Remote control effects
on the outcome of the RCM reaction were also previously
reported in the synthesis of macrocyclic natural products,
such as salicylihalamide [17] and epothilones [18].
Our initial success with RCM-mediated macrocycliza-
tion of tripeptide dienes [9] encouraged medicinal chemis-
try efforts to focus on exploring the structure–activity
relationship (SAR) of these analogs. Our SAR studies
included evaluation of modifications along the hydrocar-
bon linker connecting the P1 to the P3 moiety. Previously,
the position of the double bond was found to play a signif-
icant role in the potency of the compounds [9]. For exam-
ple, inhibitor 3d (IC50 = 10 nM, EC50 = 77 nM) was found
to be 5–7-fold more potent than 6a (Scheme 2;
IC50 = 50 nM, EC50 = 410 nM) in our in vitro and cell-
based assays [19]. Therefore, the presence of a trans double
bond connecting the Ce0 of (2S)-N-Boc-amino-6-heptenoic
acid (P3) to the Ce of a homoallyl ACCA unit at P1 was
decreasing the binding affinity of 6a for its target (Scheme
2) [9]. However, saturation of the double bond to give ana-
log 6b (IC50 = 16 nM, EC50 = 120 nM) resulted in the re-
gain of potency in both the enzymatic and the cell-based
assays [9].
1
perature H NMR data of the acyclic tripeptide 2a sug-
gested a tertiary amide rotational barrier of >13 kcal/
mol. Finally, the overall conformational preference of
diene 2c clearly pre-organized the peptidic backbone in
the same b-strand conformation as that of the desired mac-
rocyclic product (3c), thus facilitating the RCM reaction;
RCM-induced macrocyclization of diene 2c was achieved
H3CO
N
O
R2
H3CO
N
O
R2
H
N
H
N
CO2R1
α
α
CO2R1
N
N
RCM
β"
β"
α'
O
R3
α'
O
β
β
R3
γ
δ
γ
O
O
β'
β'
η'
ε'
θ'
X
X
η'
R5
θ'
R5
γ'
γ'
ι'
R4
R4
ε'
δ'
δ'
2a R1 = CH3, R2 = R3 = R4 = R5 = H, X = CH2
3a R1 = CH3, R2 = R3 = R4 = R5 = H, X = CH2 (<15% yield)
2b R1 = CH3, R2 = R4 = R5 = H, R3 = NHBoc, X = CH2
2c R1 = CH3, R2 = Ph, R3 = NHBoc, R4 = R5 = H, X = CH2
2d R1 = R4 = R5 = CH3, R2 = Ph, R3 = NHBoc, X = S
3b R1 = CH3, R2 = R4 = R5 = H, R3 = NHBoc, X = CH2 (40% yield)
3c R1 = CH3, R2 = Ph, R3 = NHBoc, R4 = R5 = H, X = CH2 (~80% yield)
LiOH
THF/CH3OH/H2O
3d R1 = H, R2 = Ph, R3 = NHBoc, R4 = R5 = H, X = CH2 (90% yield)
3e R1 = H, R2 = Ph, R3 = NHBoc, R4 = R5 = CH3, X = S
HN
O
N
BILN 2061
3f R1 = H, R2
=
R3
=
R4 = R5 = H, X = CH2
S
N
H
O
Scheme 1. RCM-induced macrocyclization of dienes 2.

5166
Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
H3CO
N
O
H3CO
N
O
R2
H
N
H
N
CO2H
α
α
CO2R1
N
N
β"
β"
1) RCM
α'
O
BocHN
R5
α'
O
β
β
BocHN
η
γ
O
O
δ
ε
γ
2) LiOH
THF/CH3OH/H2O
β'
R4
β'
δ
X
X
η'
ε
R5
γ'
γ'
R4
IC50 (nM)
ε'
δ'
ε'
δ'
50
16
74
12
17
4
5a X = CH2, R1 = R2 = H
5b X = O, R1 = R2 =H
6a X = CH2, R1 = R2 = H,
E
-double bond
iii
iii
6b X = CH2, R1 = R2 = H, saturated linker
6c X = O, R1 = R2 = H, E- double bond
6d X = O, R1 = R2 = H, saturated linker
6e X = O, R1 = CH3, R2 = H, E- double bond
6f X = O, R1 = CH3, R2 = H, saturated linker
6g X = O, R1 = H, R2 = CH3, E- double bond
6h X = O, R1 = H, R2 = CH3, saturated linker
6i X = S, R1 = CH3, R2 = H, E- double bond
5c X = O, R1 = CH3, R2 = H
5d X = O, R1 = H, R2 = CH3
5e X = S, R1 = CH3, R2 = H
iii
iii
600
340
27
Scheme 2. SAR of covalent linker connecting P1–P3.
A co-crystal structure of inhibitor 3d with the NS3 pro-
tease was also obtained, revealing the binding interactions
between the inhibitor and the active site of the enzyme [6].
This structure indicated that the Cc0 of P3 was located near
P3 was substituted with a methyl group at the R5 position
(Scheme 1, compound 3e and Scheme 2, compounds 6g and
6h), a dramatic loss in potency was observed (e.g. >20-fold
loss in potency was observed between compound 6b and
6h). This observation is consistent with our X-ray data,
suggesting that placing a methyl group (i.e. R5 = CH3)
directly into the binding pocket of the enzyme would dis-
rupt many of the favorable interactions observed in the
co-crystal of the NS3 protease-3d complex [6]. Finally,
inhibitor 6f was found to be only slightly more potent than
3d in both the enzymatic and cell-based assay, most likely
due to further rigidification of the linker moiety by the
introduction of the Cb-methyl at P3. However, based on
the overall biopharmaceutical properties of these inhibi-
tors, optimization of the critical macrocyclization step via
RCM chemistry was focused on analogs of 3d. Further-
more, SAR optimization of the P2 quinoline and the
N-Boc capping group of the P3 moiety [24] led to the dis-
covery of the first small molecule inhibitor of HCV NS3
protease, BILN 2061 (Scheme 1, 3f), that exhibited cellular
potency of ꢀ1 nM, good metabolic stability and oral bio-
availability suitable for a drug substance. BILN 2061
(3f), was the first clinically relevant NS3 protease inhibitor
that achieved dramatic viral load reduction in HCV-
infected humans after treatment of only 48 h [25].
˚
the sulfur atom of residue cysteine 159 (ꢀ3.9 A from P3-Cc
to the S of Cys159) and suggested that binding affinity
could plausibly be increased through a H-bond, or a
dipole–dipole interaction with the Cys159 thiol. Hence, a
variety of new tripeptide dienes were synthesized, replacing
the Cc of P3 with oxygen or sulfur (compounds 2d and 5b–
e). In addition, in order to compare the SAR effects with
those previously reported for acyclic tetrapeptides
[5a,10b,11] and tripeptides [20], as well the activated car-
bonyl inhibitors VX-950 (Vertex Pharmaceuticals) [21]
and SCH 503034 (Schering-Plough) [22], the effects of
branching at the Cb of the P3 moiety were also evaluated.
The building blocks required for the synthesis of the new
dienes were prepared from commercially available N-Boc
protected L-serine, L-threonine, (2S,3S)-allo threonine and
L-penicillamine [23]. RCM-induced macrocyclization of
dienes 2d and 5b–e was catalyzed by the first generation
Ru-based Grubbs catalyst (4) using the same protocol as
previously reported [9]. Interestingly, neither substitution
of R4 by a methyl (Scheme 2), nor replacement of the Cc
by oxygen or sulfur, had any dramatic effects on the
in vitro potency (IC50) of the corresponding macrocyclic
compounds (Scheme 1 and 2). However, when the Cb of
Since the disclosure of the first series of macrocyclic inhib-
itors by Boehringer Ingelheim[26] and particularly the

Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
5167
Phase I clinical data of BILN 2061 (3f) [25], many com-
pounds characterized by the same 15-member ring pepti-
domimetic scaffold have appeared in the drug-discovery
pipelines of the pharmaceutical industry [27]; these include
(Fig. 2) derivatives from Bristol-Myers Squibb (e.g. com-
pound 7), Enanta-Chiron (e.g. compound 8), Intermune-
Array Biopharma (e.g. compound 9) and Pfizer (e.g. com-
pound 10).
the vinyl ACCA moiety [14,28]. To the best of our knowl-
edge, the reaction pathway leading to this epimerizion was
unprecedented in the literature. In fact, RCM reactions
involving vinylcyclopropanes had been previously reported
in the synthesis of radicicol-type macroclides [29], coron-
anes [30], and oligo-gem-difluorocyclopropanes [31], with-
out any evidence of stereomutation.
This led us to the examination of other Ru-based cata-
lysts (Fig. 3), which included the Hoveyda catalyst 11
[32], as well as the second-generation, imidazolium-based
catalysts 12 and 13, which normally provide better RCM
reactivity and higher turnover number (TON) [33]. With
none of these catalysts was the epimerization reaction
observed, and this led us to abandon catalyst 4 in our opti-
mization/scale-up studies, focusing instead on 11, 12, and
13. Faced with developing a route to BILN 2061 (3f) and
related analogs that could be implemented at least up to
the 100 kg scale, we had to address a number of complex
challenges:
2.2. RCM optimization chemistry
Faced with the problem of producing large amounts of
BILN 2061 (3f) and related analogs, and considering the
synthetic difficulties associated with a total re-design of
the synthesis, we considered defining a scalable process
around the RCM reaction. The (1R,2S)-vinyl ACCA moi-
ety (P1) of 3, as well as compounds 7–10, plays a critical
role in the conformation of the macrocyclic scaffold and,
consequently, the affinity of these compounds for their
intended biological target [6,9,10b]. Therefore, maintaining
absolute stereochemical fidelity during the RCM conver-
sion of an acyclic diene to the desired macrocyclic product
was critical not only to our drug discovery program, but
also and especially to the process development of these
compounds. Our initial evaluation of the factors that mod-
ulate the outcome of the RCM reaction, using the unique
tripeptide precursors from this class of compounds, evi-
denced a Ru-mediated epimerization reaction involving
1. Throughput: this is the most important factor in large-
scale execution of a reaction that requires high dilu-
tion. Our ability to prepare multikilo amounts of
BILN 2061 is tied to the effective molarity[34] of the
particular ring-closure to the 15-membered ring.
Because the kinetic and thermodynamic effective
molarities of these cyclization may differ, both had
to be elucidated.
H3CO
N
N
N
S
O
S
O
O
R
O
O
O
H
N
H
NH
OH
N
N
N
H
H
N
O
O
O
O
N
O
O
O
O
8 (Enanta-Chiron)
7 (Bristol-Myers Squibb)
N
N
N
N
F
N
O
S
O
S
O
O
O
O
O
H
N
H
NH
OH
N
N
N
H
N
H
N
O
O
O
O
O
O
O
9 (Intermune-Array Biopharma)
10 (Pfizer)
Fig. 2. Examples of macrocyclic peptides currently under investigation as potential therapeutic agents for the treatment of hepatitis C infections.

5168
Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
PCy3
PCy3
Cl
Cl
Cl
Cl
Ru
O
Ru
N
N
Ph
N
N
PCy3
Cl
Cl
Ru
O
Ru
Cl
13 (2H)
12 (2G)
Cl
4 (1G)
Ph
11 (1H)
PCy3
Fig. 3. Ruthenium catalysts described in this study.
2. Catalyst load: in initial experiments, a load of 5–
10 mol% catalyst 11 had been necessary. This may prove
prohibitive from a cost standpoint. In addition, removal
of such large levels of Ru (to <10 ppm levels) from the
drug substance is quite challenging.
3. Reaction time: initial runs required 20–24 h reaction
times. This adds to the cost of the process in terms of
plant utilization. A shorter reaction time is needed.
4. Reproducibility: this turned out to be a major issue in
many runs, especially as the TON improved from 5–
10% to <1%. The epimerization reaction described
above re-surfaced even with catalysts such as 11, and
this required tight quality control of the starting materi-
als, solvent and all reagents. The need to perform acid
washes on the solvent used, to eliminate traces of amines
and/or phosphines, has been described already [35].
Table 1. It must be noted that, for fair comparison, all
these substrates were used in the crude state, i.e. as emerg-
ing from the assembly sequence of the tripeptide diene
without further purification. The dienes are typically rather
pure (>95% w/w), and further purification, in our experi-
ence, only affected the TON obtained and not the course
of the reaction, with one major exception: the presence of
even small amounts (1–2% mol) of phosphines or amines,
as mentioned above, can cause appreciable amounts of epi-
merization even with catalyst 11.
None of the reactions shown in Table 1 suffered from this
problem. In all experiments the catalyst load represents the
amount of catalyst (added in portions) needed to bring the
particular reaction to completion (<1–2% starting material).
As shown in the table, RCM of diene 14a proceeds
uneventfully in excellent yield with Hoveyda catalyst 11
(entries 1–2). As long as the concentration does not exceed
10 mM, no cyclic dimers are detected by LC-MS. However,
the reaction requires as much as 3.5–5 mol% catalyst as well
as prolonged reaction times (20–24 h) to go to completion.
When more active catalysts 12 and 13 are used, loads of
0.5–1 mol% suffice for completion in only 1–4 h, but the
product is accompanied by a number of cyclic dimers total-
ing 8–10% in yield; these are problematic to remove and
require extensive recrystallization steps (entries 3–4). A con-
trol experiment (entry 5) indicates that the reaction reaches
an apparent equilibrium in about 5 min, reaching 95% con-
version only as long as ethylene is not allowed to evaporate
and shift the equilibrium to the right. About 7% dimer con-
tent seems to be the thermodynamic value at this concentra-
tion with this substrate. It appears therefore that, under the
conditions of entries 1 and 2, the reaction is under kinetic
Several complementary strategies for the assembly of
dienes as substrates to the RCM reaction, as well as the
installation of the quinoline moiety at P2, have been care-
fully examined, and the results are discussed elsewhere [36].
For the purpose of the RCM studies, it suffices to note that
a number of related diene precursors were available for
development studies (Scheme 3).
Substrates 14a–d represent potential intermediates
where the hydroxyl group on the P2 moiety is protected
(14a), unprotected (14b), activated for later introduction
of the quinoline moiety by SN2 reaction (14c), and already
bearing the complete P2 construct (14d). Clearly, the rela-
tive efficiency of each substrate assembly must be weighed
against their performance in the RCM reaction under a
variety of catalytic conditions. These are collected in
RO
RO
H
H
CO2Me
N
CO2Me
N
N
H
N
H
Ru catalyst
solvent
O
N
O
O
N
O
O
O
O
O
15a-d
14
NHi-Pr
S
14d R =
14a R = PNB
N
MeO
N
14b R = OH
14c R = OSO2-C6H4-p-Br
Scheme 3. RCM reaction development using substrates 14a–d.

Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
5169
Table 1
Development of RCM reaction for the pilot plant (Scheme 3)
Entry
Substrate
(10 mM)
Catalyst (equiv.) + additives
Solvent,
Temperature (°C)
Time (h)
Yield of 15a–d
(HPLC assay)
% Cyclic dimers
(HPLC area %)
Unreacted
substrate (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
a
14a
14a
14a
14a
14aa
14b
14b
14b
14b
14b
14b
14b
14b
14c
14c
14c
14c
14c
14d
14d
14d
11 (5%)
CH2Cl2, 40
PhMe, 60
PhMe, 60
24
20
4
90
90
87
85
88
80
78
72
65
77
80
81
88
83
87
85
72
69
50
93
86
<1
<1
8
10
7
13
13
18
10
9
12
12
6
5
5
5
15
17
<2
<2
5
<2
<2
<1
<1
5
<2
<1
<1
<1
<2
<1
<1
<1
2
11 (3.5%)
12 (0.5%)
13 (1%)
12 (2.5%)
11 (2.5%)
12 (2%)
12 (2%)
13 (1%)
13 (2%)
12 (0.7%)
12 (0.5%)
13 (1.2%)
11 (4%)
11 (4%)
11 (2%)
12 (2%)
13 (2%)
11 (7%) + 1 equiv. CuI
12 (10%) + 1 equiv. CuI
12 (5%)
PhMe, 55
1
CH2Cl2, 40
CH2Cl2, 40
CH2Cl2, 40
PhMe, 60
PhMe, 55
THF, 60
EtOAc, 60
EtOAc/PhMe 3:2, 60
THF/PhMe 4:1, 60
CH2Cl2, 40
PhMe, 80
PhMe, 60
PhMe, 60
PhMe, 60
CH2Cl2, 40
CH2Cl2, 40
PhMe, 60
0.5
20
20
1
1
22
6.5
4
8
20
20
20
2
<1
<1
<1
<1
50
2
40
24
1
n.d.
<2
Experiment carried out in sealed tube.
control, whereas underthe conditions in entry 5 the control is
thermodynamic; evidently, the kinetic effective molarity
of this reaction under catalysis by 11 is slightly more favor-
able than the thermodynamic one.
19). The addition of one full equivalent of CuI (presumably
binding the quinoline nitrogen and freeing more catalyst)
was necessary in order to reach good conversion (entry
20). Cuprous iodide was the most active co-catalyst of
the many Cu-based species screened [33b]. However, CuI
is destructive to the catalyst, and even with highly active
12, as much as 5–10% mol catalyst is needed, as long as
the solvent is dichloromethane. Surprisingly, we found that
nonpolar solvents like PhMe and higher temperatures
obviate the need for CuI, and high yields can be obtained
without it. In this case also, however, rather large catalyst
loads are needed (entry 21). Dimer levels can be limited to
5%, and this protocol becomes therefore also relatively
attractive. We do not fully understand the origin of the
dichotomy between experiments in entries 20 and 21: the
effect appears to be both solvent- and temperature-related.
Perhaps both factors affect binding of the catalyst to the P2
quinoline moiety, which may be the culprit for the slower
reaction rate and the higher catalyst load needed.
Clearly, a variety of subtle factors are at play here, and the
most surprising finding in this study was the difference in
effective molarity between closely related substrates such as
14a–d, for which we cannot advance a reasonable explana-
tion. In general, catalysts 12 and 13 can be used in much
smaller proportions (TONs up to 200 for 12 versus 20–30
for 11) and yield faster rates. Unfortunately, they always lead
to sizable amounts of dimeric product even at 10 mM sub-
strate concentration, probably a result of partial or total
thermodynamic control. The observation that minor struc-
tural differences can have a major influence on the effective
molarity of the reaction suggests a clear strategy for future
optimization of this reaction.
Other substrates perform less effectively than 14a. Thus
14b, a very desirable starting material because it eliminates
the C-4 protection/deprotection of the hydroxyproline sub-
unit, apparently has a less favorable thermodynamic effec-
tive molarity and it is kinetically easier to equilibrate than
14a. The conformational/energetic difference versus 14a
must be slight, but is enough to cause more dimer forma-
tion under all conditions tested: thus, 13% dimers are
observed even with the milder catalyst 11, whereas higher
proportions of dimers are obtained with second-generation
catalysts (entries 6–9). This led to an extensive solvent
screen aimed at utilizing this reaction for large-scale pro-
duction: of all the solvent tested, THF and EtOAc gave
the best yields (77–80%, entries 10–11), but led to very slow
RCM rates. Better results were obtained with solvent mix-
tures: use of THF/PhMe gave the best results, achieving a
yield of 88%, with only 6% dimers (entry 13). The difference
in effective molarity (both kinetic and thermodynamic)
between 14a and 14b is very hard to explain, given the
small structural difference and the remote nature of the
substituent at P2.
Substrate 14c behaved slightly more favorably than 14b,
and here dimer levels could be kept at 5% (entries 14–18),
as long as the first-generation Hoveyda catalyst was used
(entries 14–16), whereas dimer levels climbed up to 17%
otherwise (entry 18).
Finally, use of substrate 14d, containing the quinoline
moiety at P2, gave markedly reduced RCM rates (entry

5170
Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
ult, D. Wernic, D. Lamarre, Bioorg. Med. Chem. Lett. 8 (1998)
1713;
3. Conclusion
`
´
(b) M. Llinas-Brunet, M. Bailey, R. Deziel, G. Fazal, V. Gorys,
S. Goulet, T. Halmos, R. Maurice, M. Poirier, M.-A. Poupart,
J. Rancourt, D. Thibeault, D. Wernic, D. Lamarre, Bioorg.
Med. Chem. Lett. 8 (1998) 2719;
In summary, the b-strand macrocyclic core of the antivi-
ral agent BILN 2061 (3f) was synthesized from a tripeptide
diene in excellent yield using ring-closing metathesis. A side
reaction that led to epimerized by-products, and was
shown to be catalyst dependent, was eliminated by careful
selection of the ligands attached to the Ru center. To the
best of our knowledge, this is the first application of a
Ru-induced RCM reaction to the large-scale synthesis of
a class of molecules that have significant potential value
as therapeutic agents [27].
It is well known that the frequency at which novel syn-
thetic methodologies make it out of the academic laborato-
ries and into an industrial manufacturing environment is
relatively low, particularly when the chemistry involves
potentially hazardous metals, complex intellectual property
scenarios, costly reagents/catalysts and reactions requiring
very high dilutions. In this report, we have provided a brief
account of the studies that led to the first practical applica-
tion of ring-closing metathesis in the pharmaceutical indus-
try and that allowed the production of BILN 2061 (3f) in
>100 kg quantities.
(c) C. Steinkuhler, G. Biasiol, M. Brunetti, A. Urbani, U. Koch,
¨
R. Cortese, A. Pessi, R. De Francesco, Biochemistry 37 (1998)
8899;
(d) P. Ingallinella, S. Altamura, E. Bianchi, M. Taliani, R.
Ingenito, R. Cortese, R. De Francesco, C. Steinkuhler, A. Pessi,
¨
Biochemistry 37 (1998) 8906.
`
[5] (a) M. Llinas-Brunet, M. Bailey, G. Fazal, E. Ghiro, V. Gorys, S.
Goulet, T. Halmos, R. Maurice, M. Poirier, M.-A. Poupart, J.
Rancourt, D. Thibeault, D. Wernic, D. Lamarre, Bioorg. Med.
Chem. Lett. 10 (2000) 2267;
(b) S.R. LaPlante, D.R. Cameron, N. Aubry, S. Lefebvre, G. Kukolj,
R. Maurice, D. Thibeault, D. Lamarre, M. Llinas-Brunet, J. Biol.
`
Chem. 274 (1999) 18618.
[6] Y.S. Tsantrizos, G. Bolger, P. Bonneau, D.R. Cameron, N. Goud-
`
reau, G. Kukolj, S.R. LaPlante, M. Llinas-Brunet, H. Nar, D.
Lamarre, Angew. Chem., Int. Ed. Engl. 42 (2003) 1356.
[7] (a) R.C. Reid, L.K. Pattenden, J.D.A. Tyndall, J.L. Martin, T.
Walsh, D.P.J. Fairlie, Med. Chem. 47 (2004) 1641;
(b) P.A. Bartlett, M. Rezac, S. Olson, S. Phillips, PTC International
Application [The Regents of the University of California, USA] WO
02/099045, December 2002;
(c) E. Dumez, J.S. Snaith, R.F.W. Jackson, A.B. McElroy, J.
Overington, M.J. Wythes, J.M. Withka, T.J. McLellan, J. Org.
Chem. 67 (2002) 4882;
(d) R.C. Reid, M.J. Kelso, M.J. Scanion, D.P. Fairlie, J. Am. Chem.
Soc. 124 (2002) 5673;
(e) M.P. Glenn, L.K. Pattenden, R.C. Reid, D.P. Tyssen, J.D.A.
Tyndall, C.J. Birch, D.P. Fairlie, J. Med. Chem. 45 (2002) 371;
(f) J.S. Nowick, E.M. Smith, J.W. Ziller, A.J. Shaka, Tetrahedron 58
(2002) 727;
With further improvements in catalyst performance and
reaction throughput, it is our firm belief that the olefin
RCM reaction can be used for the manufacturing of BILN
2061, or closely-related analogs, on full production scale
i.e. multiton quantities. Such developments will be
described in due course.
(g) A.S. Ripka, K.A. Satyshur, R.S. Bohacek, D.H. Rich, Org. Lett.
3 (2001) 2309;
Acknowledgments
(h) C.C. Mak, V.-D. Le, Y.-C. Lin, J.H. Elder, C.-H. Wong, Bioorg.
Med. Chem. Lett. 11 (2001) 219;
(i) J.D.A. Tyndall, D.P. Fairlie, Curr. Med. Chem. 8 (2001) 893;
(j) D.P. Fairlie, M.L. West, A.K. Wong, Curr. Med. Chem. 5 (1998)
29;
(k) Z.-Q. Tian, B.B. Brown, D.P. Mack, C.A. Hutton, P.A. Bartlett,
J. Org. Chem. 62 (1997) 514;
(l) J.S. Nowick, M. Pairish, I.Q. Lee, D.L. Holmes, J.W. Ziller, J.
Am. Chem. Soc. 119 (1997) 5413;
(m) A. Boumendjel, J.C. Roberts, E. Hu, P.V. Pallai, J. Org. Chem.
61 (1996) 4434;
(n) A.B. Smith, R. Hirschmann, A. Pasternak, M.C. Guzman, A.
Yokoyama, P.A. Sprengeler, P.L. Darke, E.A. Emini, W.A. Schleif,
J. Am. Chem. Soc. 117 (1995) 11113;
We are grateful to Professors Amir H. Hoveyda
(Department of Chemistry, Boston College) and Steven
P. Nolan (Department of Chemistry, University of New
Orleans) for providing us with samples of their catalysts.
We thank Drs. Paul Anderson, Michael Bo¨s, Pierre Beau-
lieu, Michael Cordingley, Anne-Marie Faucher, Daniel
`
´
Lamarre, Montse Llinas-Brunet, Marc-Andre Poupart,
Bruno Simoneau, Ioannis Houpis, Rogelio Frutos, Fabrice
Gallou, Robert D. Simpson and Xuwu Feng for their sup-
port with this project.
Appendix A. Supplementary data
(o) D.P. Fairlie, G. Abbenante, D.R. March, Curr. Med. Chem. 2
(1995) 654;
(p) S.F. Martin, C.J. Oalmann, S. Linas, Tetrahedron 49 (1993) 3532.
[8] (a) Recent reviews on ROM and RCM: R.H. Grubbs, Tetrahedron
60 (2004) 7117;
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.jorganchem.2006.09.027.
(b) R.H. Grubbs (Ed.), Handbook of Metathesis, vols. 1–3, Wiley-
VCH, 2003;
References
(c) R.R. Schrock, A.H. Hoveyda, Angew. Chem. Int. Ed. 42 (2003)
4592;
(d) T.M. Trnka, R.H. Grubbs, Acc. Chem. Res. 34 (2001) 18;
(e) A. Furstner, Angew. Chem., Int. Ed. Engl. 39 (2000) 3012.
¨
[1] (a) A.G. Cochran, Curr. Opin. Chem. Biol. 5 (2001) 654;
(b) M.W. Peczuh, A.D. Hamilton, Chem. Rev. 100 (2000) 2479;
(c) A.G. Cochran, Chem. Biol. 7 (2000) R85.
[9] N. Goudreau, C. Brochu, D.R. Cameron, J.-S. Duceppe, A.-M.
Faucher, J.-M. Ferland, C. Grand-Ma ıtre, M. Poirier, B. Simoneau,
Y.S. Tsantrizos, J. Org. Chem. 69 (2004) 6185.
[2] E. Weerapana, B. Imperiali, Org. Biomol. Chem. 1 (2003) 93.
`
[3] A. Loregian, H.S. Marsden, G. Palu, Rev. Med. Virol. 12 (2002) 239.
[10] (a) 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.
`
[4] (a) M. Llinas-Brunet, M. Bailey, G. Fazal, S. Goulet, T. Halmos,
S. LaPlante, R. Maurice, M. Poirier, M.-A. Poupart, D. Thibea-

Y.S. Tsantrizos et al. / Journal of Organometallic Chemistry 691 (2006) 5163–5171
5171
`
Farina, H. Heimroth, T. Krueger, J. Schnaubelt, J. Org. Chem. 70
(2005) 5869;
Tsantrizos, S.M. Weldon, C.-L. Yong, M. Llinas-Brunet, Nature 426
(2003) 186;
(b) J. Rancourt, D.R. Cameron, V. Gorys, D. Lamarre, M. Poirier,
(c) C.M. Rice, Nature 426 (2003) 129.
`
D. Thibeault, M. Llinas-Brunet, J. Med. Chem. 47 (2004) 2511.
[26] Y.S. Tsantrizos, D.R. Cameron, A.-M. Faucher, E. Ghiro, N.
`
[11] (a) N. Goudreau, D.R. Cameron, P. Bonneau, V. Gorys, C. Plouffe,
Goudreau, T. Halmos, M. Llinas-Brunet, WO 2000/59929.
`
M. Poirier, D. Lamarre, M. Llinas-Brunet, J. Med. Chem. 47 (2004)
123;
[27] (a) S. Venkatraman, V.M. Girijavallabhan, Schering Corporation
WO 2005/051980 A1;
(b) M.-A. Poupart, D.R. Cameron, C. Chabot, E. Ghiro, N.
Goudreau, S. Goulet, M. Poirier, Y.S. Tsantrizos, J. Org. Chem. 66
(2001) 4743.
(b) L.M. Blatt, S.M. Wenglowsky, S.W. Andrews, Y. Jiang, A.L.
Kennedy, K.R. Condroski, J.A. Josey, P.J. Stengel, M.R. Madduru,
G.A. Doherty, B.T. Woodard, Intermune, Inc./ Array Biopharma
Inc. WO 2005/037214 A2;
[12] Labeling of the unnatural amino acid residues within the peptidic
scaffold is based on the protease subsite nonmenclature: I. Schechter,
A. Berger, Biochem. Biophys. Res. Commun. 27 (1967) 157.
[13] A.-M. Faucher, M.D. Bailey, P.L. Beaulieu, C. Brochu, J.-S.
Duceppe, J.-M. Ferland, E. Ghiro, V. Gorys, T. Halmos, S.H.
`
(c) M. Llinas-Brunet, M. Bailey, P. Bhardwaj, P. Forgione, E. Ghiro,
N. Goudreau, T. Halmos, J. Rancourt, Boehringer Ingelheim
(Canada) Ltd. WO 2005/028501 A1;
(d) S. Nakajima, Y. Sun, D. Tang, G. Xu, B. Porter, Y.S. Or, Z.
Wang, Z. Miao, Enanta Pharmaceuticals Inc. WO 2004/093798 A2;
(e) F. McPhee, J.A. Campbell, W. Li, S. D’Andrea, Z.B. Zheng, A.C.
Good, D.J. Carini, B.L. Johnson, P.M. Scola, Bristol-Myers Squibb
Company WO 2004/094452 A2;
`
Kawai, M. Poirier, B. Simoneau, Y.S. Tsantrizos, M. Llinas-Brunet,
Org. Lett. 6 (2004) 2901.
[14] M. Poirier, N. Aubry, C. Boucher, J.-M. Ferland, S. LaPlante,
Y.S.J. Tsantrizos, J. Org. Chem. 70 (2005) 10765.
[15] (a) M. Weck, B. Mohr, J.-P. Sauvage, R.H. Grubbs, J. Org. Chem.
64 (1999) 5463;
(f) S. Chen, X. Mei, Z. Wang, Boehringer Ingelheim International
GMBH. WO 2004/093915 A1;
(b) B. Mohr, M. Weck, J.-P. Sauvage, Grubbs, Angew. Chem., Int.
Ed. Engl. 36 (1997) 1308;
(c) M.J. Marsella, H.D. Maynard, R.H. Grubbs, Angew. Chem., Int.
Ed. Engl. 36 (1997) 1101.
(g) K. Donsbach, D. Ecker, R.P. Frutos, F. Gallou, D. Gutheil, N.
Haddad, R. Hagenkoetter, D. Kemmer, J. Kroeber, T. Nicola, J.
Schnaubelt, M. Schul, R.D. Simpson, X. Wei, E. Winter, Y. Xu, N.K.
Yee, J. Brandenburg, Boehringer Ingelheim International GMBH.
WO 2004/092203 A2;
[16] R. Kaul, S. Surprenant, W.D. Lubell, J. Org. Chem. 70 (2005) 3838.
[17] (a) A. Furstner, T. Dierkes, O.R. Thiel, G. Blanda, Chem. Eur. J. 7
¨
(h) W. Dersch, W. Samstag, Boehringer Ingelheim International
GMBH. WO 2004/089974 A1;
(2001) 5286;
(b) A. Furstner, O.R. Thiel, G. Blanda, Org. Lett. 2 (2000) 3731.
¨
(i) Z. Miao, Y. Sun, F. Wu, S. Nakajima, G. Xu, Z. Wang, Enanta
Pharmaceuticals Inc. WO 2004/072243 A2;
[18] D. Meng, D.-S. Su, A. Balog, P. Bertinato, E.J. Sorensen, S.J.
Danishefsky, Y.-H. Zheng, T.-C. Chou, L. He, S.B. Horwitz, J. Am.
Chem. Soc. 119 (1997) 2733.
`
(j) M. Llinas-Brunet, M.D. Bailey, Boehringer Ingelheim (Canada)
Ltd. WO 2004/037855 A1;
`
[19] A. Pause, G. Kukolj, M. Bailey, M. Brault, F. Doˆ, T. Halmos, L.
Lagace´, R. Maurice, M. Marquis, G. McKercher, C. Pellerin, L.
Pilote, D. Thibeault, D.D. Lamarre, J. Biol. Chem. 278 (2003) 20374.
(k) M. Llinas-Brunet, V.J. Gorys, Boehringer Ingelheim (Canada)
Ltd. WO 2003/064455 A2
[28] 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., in press.
[29] (a) Z.-Q. Yang, X. Geng, D. Solit, C.A. Pratilas, N. Rosen, S.J.
Danishefsky, J. Am. Chem. Soc. 126 (2004) 7881;
(b) K. Yamamoto, K. Biswas, C. Gaul, S. Danishefsky, J. Tetrahe-
dron Lett. 44 (2003) 3297.
[30] A.G.M. Barrett, D. Hamprecht, R.A. James, M. Ohkubo, P.A.
Procopiou, M.A. Toledo, A.J.P. White, D.J. Williams, J. Org. Chem.
66 (2001) 2187.
[31] T. Itoh, K. Mitsukura, N. Ishida, K. Uneyama, Org. Lett. 2 (2000)
1431.
`
[20] M. Llinas-Brunet, M.D. Bailey, E. Ghiro, V. Gorys, T. Halmos, M.
Poirier, J. Rancourt, N. Goudeau, J. Med. Chem. 47 (2004) 6585.
[21] (a) R.B. Perni, S.J. Almquist, R.A. Byrn, G. Chandorkar, P.R.
Chaturvedi, L.F. Courtneay, C.J. Decker, K. Dinehart, C.A. Gates,
S.L. Harbeson, A. Heiser, G. Kalkeri, E. Kolaczkowski, K. Lin,
Y.-P. Luong, B.G. Rao, W.P. Taylor, J.A. Thomson, R.D. Tung,
Y. Wei, A.D. Kwong, C. Lin, Antimicrob. Agents Chemother. 50
(2006) 899;
(b) K. Lin, R.B. Perni, A.D. Kwong, C. Lin, Antimicrob. Agents
Chemother. 50 (2006) 1813.
[22] B.A. Malcolm, R. Liu, F. Lahser, S. Agrawal, B. Belanger, N.
Butkiewicz, R. Chase, F. Gheyas, A. Hart, D. Hesk, P. Ingravallo,
C. Jiang, R. Kong, J. Lu, J. Pichardo, A. Prongay, A. Skelton, X.
Tong, S. Venkatraman, E. Xia, V. Girijavallabhan, F.G. Njoroge,
Antimicrob. Agents Chemother. 50 (2006) 1013.
[32] J.S. Kingsbury, J.P.A. Harrity, P.J. Bonitatebus, A.H. Hoveyda, J.
Am. Chem. Soc. 121 (1999) 791.
[33] (a) M. Scholl, T.M. Trnka, J.P. Morgan, R.H. Grubbs, Tetrahedron
Lett. 40 (1999) 2247;
[23] Details on the synthesis of dienes 5b–e and a typical procedure for
kilo-scale RCM reaction can be found in the Supplementary
Materials.
(b) T. Weskamp, F.J. Kohl, W. Hieringer, D. Gleich, W.A.
Herrmann, Angew. Chem., Int. Ed. Engl. 38 (1999) 2416;
(c) J. Huang, E.D. Stevens, S.P. Nolan, J.L. Petersen, J. Am. Chem.
Soc. 121 (1999) 2674;
`
[24] M. Llinas-Brunet, M.D. Bailey, G. Bolger, C. Brochu, A.-M.
Faucher, J.-M. Ferland, M. Garneau, E. Ghiro, V. Gorys, C.
Grand-Ma ıtre, T. Halmos, N. Lapeyre-Paquette, F. Liard, M.
(d) S.B. Garber, J.S. Kingsbury, B.L. Gray, A.H. Hoveyda, J. Am.
Chem. Soc. 122 (2000) 8168.
´
Poirier, M. Rheaume, Y.S. Tsantrizos, D. Lamarre, J. Med. Chem.
[34] G. Illuminati, L. Mandolini, Acc. Chem. Res. 14 (1981) 95.
[35] T. Nicola, M. Brenner, K. Donsbach, P. Kreye, Org. Process Res.
Dev. 9 (2005) 513.
[36] (a) N.K. Yee, V. Farina, I. Houpis, N. Haddad, R.P. Frutos, F.
Gallou, X.-j. Wang, X. Wei, R.D. Simpson, X. Feng, V. Fuchs, Y.
Xu, J. Tan, L. Zhang, J. Xu, L.S. Smith-Keenan, J. Vitous, M.D.
Ridges, E.M. Spinelli, M. Johnson, K. Donsbach, T. Nicola, M.
Brenner, E. Winter, P. Kreye, W.J. Samstag, J. Org. Chem. 71 (2006)
7133;
47 (2004) 1605.
[25] (a) H. Hinrichsen, Y. Benhamou, H. Wedemeyer, M. Reiser, R.E.
Sentjens, J.L. Calleja, X. Forns, A. Erhardt, J. Cro¨nlein, R.L.
Chaves, C.-L. Yong, G. Nehmiz, G.G. Steinmann, Gastroenterology
127 (2004) 1347;
(b) D. Lamarre, P.C. Anderson, M. Bailey, P. Beaulieu, G. Bolger,
P. Bonneau, M. Bo¨s, D. Cameron, M. Cartier, M.G. Cordingley,
A.-M. Faucher, N. Goudreau, S.H. Kawai, G. Kukolj, L. Lagace,
´
S.R. LaPlante, H. Narjes, M.-A. Poupart, J. Rancourt, R.E. Sentjens,
R. St George, B. Simoneau, G. Steinmann, D. Thibeault, Y.S.
(b) N. Haddad, X.-j. Wang, X. Wei, L. Zhang, J. Xu, V. Farina (in
preparation).