Discovery of novel cobactin-t based matrix metalloproteinase inhibitors via a ring closing metathesis strategy

Article abstract of DOI:10.1016/j.bmcl.2011.08.068

The discovery of potent N-hydroxyl caprolactam matrix metalloproteinase (MMP) inhibitors (6) based on the natural product Cobactin-T (2) is described. The synthetic method, which utilizes the ring closing metathesis reaction, is compatible to provide comp

Full text of DOI:10.1016/j.bmcl.2011.08.068

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6485–6490
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel Cobactin-T based matrix metalloproteinase inhibitors via
a ring closing metathesis strategy
⇑
Lawrence J. Wilson , Bingbing Wang, Shyh-Ming Yang, Robert H. Scannevin, Sharon L. Burke,
Prabha Karnachi, Kenneth J. Rhodes, William V. Murray
High-Throughput Synthesis and Neurological Disorders Teams, Johnson & Johnson Pharmaceutical Research & Development, 920 Route 202, Raritan, NJ 08869, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery of potent N-hydroxyl caprolactam matrix metalloproteinase (MMP) inhibitors (6) based on
the natural product Cobactin-T (2) is described. The synthetic method, which utilizes the ring closing
metathesis reaction, is compatible to provide complementary (R) and (S) enantiomers. These compounds
tested against MMP-2 and 9, show that the R stereochemistry (i.e., 16), which is opposite for that found in
the natural product Cobactin-T is >1000-fold more active with IC50 values of 0.2–0.6 nM against both
enzymes. The variation in the incorporation of the sulfonamide enzyme recognition element
(Ar2XAr1SO2N(R1), 6), along with alterations in the RCM/double bond chemistry (R2) provided a series
of sub nanomolar MMP inhibitors. For example, compounds 16 and 34 were found to be the most potent
with IC50 values against MMP-2 and MMP-9 found to be between 0.2 and 0.6 nM with 34 being the most
potent compound discovered (MMP-2 IC50 = 0.39 nM and MMP-9 IC50 = 0.22 nM). Compounds 16 and 34
showed acceptable drug-like properties in vivo with compound 34 showing oral bioavailability.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 19 July 2011
Revised 12 August 2011
Accepted 15 August 2011
Available online 28 August 2011
Keywords:
Matrix metalloproteinases
Ring closing metathesis
MMP-2
MMP-9
Cobactin-T
Matrix metalloproteinases are a family of zinc containing metal
enzymes responsible for degrading all types of extracellular matrix
proteins, as well as processing bioactive molecules. These include
the digestion of a large range of connective and soft tissues such as
collagen and gelatin, the cleavage of cell surface receptors, and the
release of apoptotic ligands. There are currently about 27 MMPs,
which are classified by function and substrate specificity. Specific
examples are: collagenases (MMP-1, 8, and 13); stromelysins
(MMP-3, 10, and 11); and gelatinases (MMP-2 and 9). These
enzymes have become the focus of therapeutic intervention in many
diseases including osteoarthritis, cancer, macular degeneration, car-
diovascular disease, and stroke.1,2 The gelatinases play a role in reg-
ulating angiogenesis and arterial remodeling in vascular pathologies
such as heart disease, tumor metathesis, and ischemic stroke. During
ischemic stroke, both gelatinases are up regulated and involved in
degradation of the extra-cellular matrix and blood brain barrier.
Much of the brain damage is linked to this cerebral hemorrhaging
as inflammatory cells are allowed to penetrate the brain tissue and
cause further neurological injury, as well as edema. Initial results
have shown that treatment with hydroxamic acid based broad spec-
trum MMP inhibitors such as BB-94 (1, Fig. 1) in animal models of
ischemic stroke lead to diminished hemorrhaging and infarct.2,3
As part of a drug discovery program, we became interested in
the role of the gelatin degrading matrix metalloproteinases 2
O
O
S
XH
O
H
N
HO
N
NHMe
N
H
R
N
H
OH
O
O
2 (Cobactin T - X=O, R=Me)
S
3
(X=S, R=CH2Ph)
1 (BB-94)
O
O
HO
N
H
O
4
O
R2
O
S
N
O
O
OH
N
H
O
S
N
Ar2-X-Ar1
O
N
R1
OH
*
O
5
6
⇑
Corresponding author at present address: Emory Institute for Drug Discovery,
Emory University, Atlanta, GA 30322, USA.
Cl
E-mail address: wilsonlarr@gmail.com (L.J. Wilson).
Figure 1. Structures of matrix metalloproteinase inhibitors.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.08.068

6486
L. J. Wilson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6485–6490
(MMP-2) and 9 (MMP-9) in ischemic stroke to block soft tissue
degradation. Historically, there are many types of inhibitors that
have been discovered and most are proposed to bind to the zinc
metal in the active site of the enzyme, while both non-zinc binding
and zinc mechanism based inhibitors have shown a more limited
interest.4,5 Recently, there have been significant developments
involving development of heterocyclic MMP inhibitors where the
zinc binding group (ZBG) is buried in the heterocycle ring system.5
These include ring systems such as hydroxy-pyrones (4) and pyrid-
inones (5) termed HOPO’s (Fig. 1).6,7 These ring systems have been
studied in MMP inhibitor development with success due to their
known ability to chelate a range of metal ions including zinc. We
envisioned utilizing the lysine derived N-hydroxyl caprolactam
portion of Cobactin-T (2) as a zinc binding element for matrix
metalloproteinases (6, Fig. 1). Synthetic studies on Cobactin-T have
resulted in its total synthesis, as well as analogs of the mycobac-
tins.8,9 The amino-N-hydroxy caprolactam has also been exploited
in the design of inhibitors of other zinc metal chelating metallo
enzymes (3).10 The naturally occurring S-isomer of the saturated
lactam was reported previously by Miller, and in our enantiomeric
synthesis of Cobactin-T utilizing the ring closing metathesis (RCM)
reaction.11 Placing an aryl sulfonamide on the nitrogen is a well
known approach to getting proper enzyme recognition in all types
of MMP inhibitors including carboxylic acid, hydroxamic acid, and
heterocyclic compounds.4,5 Occupancy of the S10 pocket with a
hydrophobic ligand should determine selectivity preferring deep
S10 pocket (MMP-2, 9, 13) over shallow S10 pocket (MMP-1, 7) en-
zymes.12 With this design criteria in mind, our approach was to
construct the ring in an expedient manner using the ring closing
metathesis (RCM) reaction from readily available allyl glycine
based amino acid building blocks. This effort focuses on access to
an enantiomeric route to explore the influence of stereochemistry
on the biological properties of the molecules. Furthermore, to date
none of the studies have provided complementary enantiomeric
isomers, which could prove extremely useful in understanding bio-
logical activity and structure activity relationships.
The medicinal chemistry approach starts by assembling the cap-
rolactam ring system using the ring closing metathesis reaction. The
synthetic route begins from chiral Boc protected allyl glycines (7a/
7b, Scheme 1). First, Boc-glycine coupling with O-benzyl-hydroxyl
amine was accomplished to give 8a/8b. Use of normal coupling con-
ditions (i.e., DCC, EDC) resulted in racemization of the alpha carbon
and loss of chirality. This was avoided by employing the more reac-
tive coupling reagents HATU and HOAt giving greater than 96%
enantioneric purity determined by chiral HPLC analysis. Next, alkyl-
ation with allyl bromide proceeded smoothly and in high yield to
give diallyl compounds 9a/9b. Reaction of these species with
Grubb’s second generation ruthenium catalyst II gave the
O
O
O
H
N
H
N
b-d
Boc
N
Bn
R
Boc
XH
(R)-7a (X=O)
(S)-7b (X=O)
(R)-9a (R=H)
(S)-9b (R=H)
a
(R)-8a (X=NOBn)
(S)-8b (X=NOBn)
10
(R)- (R=CH3)
(R)-11 (R=CH2Cl)
N
N
Mes
Mes
Ph
H
O
O
Cl
Cl
S
Ru
e, f
S
N
N
OH
PCy3
II
H
O
O
15
R
N
g, h, k
g, h, i
O
O
Cl
Bn
S
Boc
N
O
N
N
H
H
H
O
O
O
12a
(R)-
(R=H)
18
O
O
R
(S)-12b (R=H)
S
N
13
(R)- (R=CH3)
N
H
OH
14
(R)- (R=CH2Cl)
O
O
Cl
16
g, h, i
j, g, h, i
l, g, h, i
Cl
N
O
O
O
O
O
O
O
S
N
S
S
N
N
N
OH
N
OH
N
OH
H
H
H
O
O
O
O
O
O
17
20
19
Cl
Cl
Cl
Scheme 1. Reagents and conditions: (a) BnONH2, HATU, HOAt, N-methylmorphine, DCM; (b) for 9: Cs2CO3, DMF, allyl bromide, rt; (c) for 10: 2-methyl-allyl chloride,
Cs2CO3, DMF, rt; (d) for 11: 2-chloromethyl-allyl chloride, Cs2CO3, DMF, 60 °C; (e) for 12: II, DCM, rt; (f) for 13 and 14: II, 1,2-DCE, microwave, 150 °C; (g) TFA, DCM, rt;
(h) 4-(4-CIPh)OPhSO2CI, pyridine, DCM; (i) MeSO3H, rt; (j) [CODIrPCy3Pyr]PF6, THF, MeOH, H2; (k) Sml2, THF, rt; (l) morpholine, CH3CN, 60 °C.

L. J. Wilson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6485–6490
6487
corresponding lactams 12a/b.13 At this point, the 4-(4-chlorophen-
oxy) phenyl sulfonamide group could be attached by first deprotect-
ing the Boc group and then reacting the resulting primary amine
with 4-(4-chloro-phenoxy)-phenyl sulfonyl chloride. Removal of
the oxygen benzyl group using methane sulfonic acid gave the
resulting (S) and (R) sulfonamido N-hydroxy caprolactams 15 and
16. Next, we synthesized the saturated derivative. First, we reduced
the double bond of lactam 12a by using a heterogeneous iridium cat-
alyst. This was followed by the Boc group deprotection, sulfonamide
formation and benzyl ether deprotection sequence to provide the
saturated compound 17. Variation of the substituents on the double
bond of the ring could be provided through alteration of the RCM
reaction sequence, and two examples were explored. Thus, both a
methyl and a morpholino-methyl substituent were installed by
the following reactions: alkylation with either 2-methyl-allyl chlo-
ride (8a–10) or 2-chloromethyl-allyl chloride (8a–11); microwave
promoted RCM (10/11–13/14),14 morpholine addition (for 14);
and sulfonamide attachment followed by O-benzyl deprotection
(19 and 20). Finally, to round out the heterocycle core modifications,
reduction of the nitrogen oxygen bond was performed to provide
lactam 18 to test the influence of the system without the zinc bind-
ing group.
reported FMOC-Trityl protected lactam (21) by FMOC group
removal, followed by amide formation, and trityl group
deprotection.11 All others consisted of sulfonamide modifactions
made from the (R)-Boc-benzyl lactam 12a. The first of these con-
sisted of removing the biphenyl ether linkage (23), followed by biar-
yl ether expansion (24–28). These compounds were synthesized by
the sequence of Boc group deprotection, sulfonamide formation
with the corresponding commercially available sulfonyl chloride,
and O-benzyl deprotection. Alternatively, 5-arylthiophene contain-
ing P10 groups were installed by sufonylation with 5-bro-
mothiophenesulfonyl chloride, followed by Suzuki coupling with 4
substituted aryl boronic acids, and benzyl group removal (29–31).
Further extension
included the basic amine linkers replacing the aryl oxygen atom (32,
33) synthesized by a 4-fluoro-phenyl sulfonamide and amine dis-
placement sequence. Finally, sulfonamide N alkylation was accom-
plished by alkylation of the corresponding N-sulfonamido-O-
benzyl lactam of 16 with N-chloroethyl heterocycles, followed by
O-benzyl group deprotection (34–36) and also on the 5-arylthioph-
ene series through the previously described chemistries
(37–39) for the synthesis of 29–31 and 34–36.
The testing of these compounds against the enzymes of interest
(MMP-2 and 9) was concurrent with the synthesis efforts. We
found some very encouraging and exciting results (Table 1). First,
the (S)-analog 15, which contains the stereochemistry found in
the Cobactin-T (2) containing natural products, showed modest
Modifications in the appending nitrogen correspond to changes
in the P10 binding element. As such we focused on changes on the
nitrogen atom and aromatic portions (Scheme 2). The amide
version of 16 (compound 22) was synthesized from the previously
O
O
O
O
O
S
S
N
N
N
Y
N
OH
N
OH
N
H
OH
H
H
O
O
O
X
O
O
24 (X=CH,Y=CH)
23
22
Cl
25
(X=CH,Y=OMe)
26 (X=CH,Y=CF3)
27 (X=N,Y=CH)
Cl
28
(X=CH,Y=N)
d,g,f
d,e,f
a,b,c
d,h,i,f
O
O
R
S
S
N
R1
N
R2
N
OH
N
H
O
H
O
O
29 (R=Me)
30 (R=Cl)
12a (R1=Boc, R2=Bn)
21
(R1=Fmoc, R2=Trityl)
d,e,l,f
31 (R=OMe)
d,h,l,i/m,f
N
d,j,k,f
O
O
O
O
S
S
S
N
R
N
OH
N
OH
H
O
O
O
O
N
S
N
X
N
OH
N
O
32 (X=CH)
33 (X=N)
O
O
N
X
37
38
(R=p-Cl-Ph)
(R=p-Me-Ph)
39 (R=p-Me-Ph
)
Cl
34 (X=O)
35
(X=CH2)
36 (X=NMe)
Scheme 2. Reagents and conditions: (a) Piperidine, DMF; (b) 4-(4-CIPhO)PhCOCI, iPr2NEt, DCM; (c) Amberlyst 15 ion exchange resin, DCM, MeOH; (d) TFA, DCM; (e) 4-(4-
CIPh)PhSO2CI, Pyr, DCM; (f) MeSO3H; (g) 4-(Ar)OPhSO2CI, Pyr, DCM; (h) 5-Br-thiophene-SO2CI, Pyr, DCM; (i) 4-ArB(OH)2, Pd(OAc)2, DMF, microwave, 150 °C; (j) 4-F-PhSO2CI,
Pyr, DCM; (k) 4-Ph-piperdine or 4-Ph-piperazine, DMSO; (l) Na2CO3, DMF, N-(CICH2CH2)-morpholine, N-(CICH2CH2)-piperidine, or N-(CICH2CH2)-N0-Me-piperazine;
(m) 4-Me-PhCCH, Pd(PPh3)4, CuCI2, NEt3, DMF.

6488
L. J. Wilson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6485–6490
Table 1
IC50 Values for cobactin based mmp inhibitors (15–39)
a
Compound
MMP
Enzyme
MMP-1
IC50 (nM)
MMP-2
1599 36
0.57 0.04
3.7
>1000
2
0.53 0.01
2319
85
1
2
2
15
3
4
11
MMP-9
1117 18
0.58 0.03
5.9
>1000
1.3
0.46 0.02
5091
1506
2
1
2
23
12
9
16
MMP-3
86
MMP-13
0.4
15
16
17
18
19
20
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
>1000
>1000
39
0.2
3.9
11
>1000
>1000
493
90
3
>1000
>1000
0.39 0.01
0.55
0.49
16
>1000
>1000
0.22 0.03
0.21
0.2
17
11
0.8
0.7
7
2.4
a
Values are means of two to three separate experiments. Each dose response was ran in duplicate.
activity against both enzymes in the single digit micromolar range.
However, when testing the (R)-enantiomer 16 potencies were
found to be below 1 nM in both MMP-2 and 9 assays. The large dif-
ference in activity regarding stereochemistry has been observed in
sulfonamide phosphate MMP inhibitors but not carboxylic acid
sulfonamides.5,15 However stereo chemical preference is an
unknown in the case of Cobactin-T containing compounds, and
represents a novel development. Noteworthy is the compound
without the double bond (17) has about 10-fold less activity than
16. Furthermore, compound 16 is greater than 100-fold more
potent than activity reported for compound 4 suggesting the
azepenone scaffold (6) from the ring closing metathesis is the most
potent heterocycle system reported to date.7 The importance of the
zinc binding properties of this system were confirmed since the
normal lactam 18 which lacks the hydroxyl amine functionality
had no activity in our assays.
Further testing in MMP-2 and 9 provided interesting SAR trends
(Table 1). Replacing the sulfonamide with a normal amide (22)
resulted in greater than 1000-fold drop off in activity. The activity
observed here is similar to other heterocycle amides such as 3, and
suggests these other scaffolds may be optimized by replacing the
amide to a sulfonamide.5,6 The biaryl ether oxygen was needed
for the potent activity with the biphenyl series and a large drop
off in activity (P100-fold) was seen between 16 and the biaryl
compound 23, although this compound had 10–20-fold selectivity
towards MMP-2. The inclusion of other substituent patterns than
the chloride at the para position (24–26) and a pyridine ring
replacing the phenyl ring (27, 28) in the biaryl ether assembly
resulted in slightly less active (10–50-fold) compounds. The
5-aryl-thiophene compounds (29–31) were almost as potent and
a modest drop off in activity was observed (about 10-fold) with
similar activity to the aryl-pyridyl ether compounds (27, 28). The
replacement of the aryl oxygen with piperidine and piperazine
rings (32 and 33) resulted in complete loss of activity. Alkylations
on the sulfonamide nitrogen (34–36) and substitutions on the
azepinone ring double bond (19, 20) slightly increased (20,
34–36) or decreased (19) the potency compared to 16. Although
not as potent as the biaryl ether series, it is worth noting the best
5-aryl-thiophene compound discovered was the one with an
alkyne linker and morpholine ethyl (39) which had sub nanmolar
potency against MMP-9. Another interesting trend is that the het-
eroalkyl sulfonamides (34–36 and 39) seemed to have a 2–3-fold
selectivity towards MMP-9 versus MMP-2. Overall, the p-chloro-
biarylether and heterocycle (e.g., morpholine) substituted com-
pounds 16, 20 and 34–36 were the most potent with sub nanomo-
lar IC50 values.
To determine interfamily MMP selectivity, we tested select
compounds against three other MMP’s (1, 3, and 13; Table 1). All
four compounds (16, 20, 31, and 34) showed the following trends.
As it has been hypothesized that MMP-1 activity is associated with
the side effect of musculo-skeletal syndrome (MSS) that is
observed in the clinic, having MMP-1 inhibitory activity would
be undesirable.16 Therefore, it was significant that all were selec-
tive against MMP-1, showing no activity up to 1000-fold higher
than their efficacy against MMP’s 2 and 9. All had approximately
100-fold selectivity against MMP-3. None had selectivity against
MMP-13, showing comparable values for all three enzymes
(MMP 2, 9, and 13). This selectivity pattern is not completely
surprising since MMP-2, 9, and 13 are all deep S10 pocket enzymes
while both MMP-1 and 3 have shallow S10 pockets. As such, the
selectivity is primarily due to the choice of the biaryl ether group
on the sulfonamide.12
To get a better understanding of how these compounds might
be fitting into the enzyme active site, a homology model of
MMP-9 was built. Next, compounds 20 and 34 were docked into
the active site containing the zinc metal (Fig. 2). We found that
the best docking mode for these compounds was a pose in which
the biaryl ether sulfonamide group fits into the deep S10 pocket
on the enzyme. The bent shape of this ligand is supported by the
loss in activity observed upon removal of the biaryl oxygen (23)
as well as replacement by thiophenes (29–31). In addition to the
N-hydroxyl-lactam portion binding to the zinc atom, the model
also picks up hydrogen bonding between the sulfonamide N–H
and the amide carbonyl C–O of Pro-421. The other interaction

L. J. Wilson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6485–6490
6489
Figure 2. Pictures of compounds 20 and 34 docked in a homology model of MMP-9.
Table 2
Pharmacokinetic parameters for compounds 16 and 34
Compounda (dose-mg/kg)
Cmax
(
lM)
T1/2 (h)
AUC (ng-h/mL)
Vss (L/kg)
CL (mL/ min/kg)
16 (iv-0.5)
34 (iv-0.5)
34 (po-10)
4.6
4.4
38.3
14
2
1
1226
1831
—
3.3
1.4
—
7
19
—
a
Iv experiments performed in rats, po experiment performed in mice.
identified is between the sulfonamide S–O group and the amide N–
H of Leu-188. It is likely that this later interaction is more signifi-
cant since it would prevail for all the compounds, especially in
the case of N-alkylated derivatives 32 thru 34 as no sulfonamide
N–H bond donor exists for these compounds. Furthermore, alter-
ation of the sulfonamide to an amide (22) results in significant loss
of activity. Lastly, compound 20 is an interesting example in that
the model predicts the morpholine ring to occupy the S1 pocket
of the enzyme. Although this might be a significant finding, there
is no increase in potency observed (16 vs 20).
The highest activity resulted from the biaryl ether series. The other
modifications provided potent inhibitors but with about equal
activity to the initial discovery. Testing against other MMP’s
showed high to low selectivity and was dependent upon the clas-
sification (deep vs shallow S10 pocket enzymes). Docking studies
provided some rationale to understanding the SAR and were con-
sistent with the perceived modes of binding. Finally, pharmacoki-
netic experiments showed these compounds to be highly bio
available via several routes of administration.
Next, we became interested in exploring the drug-like features
of these compounds, specifically the pharmacokinetic and other
important ADME parameters. Compounds 16 and 34 were evalu-
ated in pharmacokinetic experiments in rats and mice (Table 2).
Intravenous administration provided exposure data, with com-
pound 16 showing a half-life of 14 h, while compound 34 had a
shorter half-life of 2 h. Compound 34 was evaluated in an oral
pharmacokinetic experiment in mice with strong evidence of good
Supplementary data
Supplementary data (experimental procedures and details are
provided for the MMP assays used and also for the synthesis of
compounds 8–20, 23–29 and 34) associated with this article
can be found, in the online version, at doi:10.1016/j.bmcl.2011.
08.068.
exposure (38 lM for a 10 mg/kg dose, Table 2). Some of the earlier
References and notes
developed MMP inhibitors, especially hydroxamic acids, suffer
from lower exposure due to metabolism related to N–O bond
reduction and glucaronidation. These results show that these com-
pounds may have better metabolic stability imparted by introduc-
tion of the caprolactam ring system onto the hydroxamic acid
functionality. Other experiments provided further evidence of
favorable drug-like features.17
In summary, we have provided a series of novel MMP inhibitors
based on the natural product series Cobactin-T (2). The incorpora-
tion of the double bond in the ring adds a series of benefits span-
ning from ease of synthesis via the RCM reaction to increased
potency against the biologic target. The access to both optical
isomers in this case elucidated a high degree of stereo recognition
with the MMP enzymes, and also showed that the more active con-
figuration was that opposite to the natural product itself. The ensu-
ing SAR study performed showed an activity pattern resulting in a
series of extremely potent enzyme inhibitors dependent mainly on
the S10 pocket recognition element contained on the sulfonamide.
1. Fingleton, B. Curr. Pharm. Des. 2007, 13, 333.
2. Lapchak, P. A.; Chapman, D. F.; Zivin, J. A. Stroke 2000, 3034.
3. Jiang, X.-F.; Namura, S.; Nagata, I. Neurosci. Lett. 2001, 305, 41.
4. Matter, H.; Schudok, M. Curr. Opin. Drug Des. Devel. 2004, 7, 513.
5. Jacobsen, J. A.; Major Jourden, J. A.; Miller, M. T.; Cohen, S. M. Biochim. Biophys.
Acta 2010, 1803, 72–94.
6. Yan, Y.-L.; Cohen, S. M. Org. Lett. 2007, 9, 2517.
7. Zhang, Y.-M.; Fan, X.; Chakaravarty, D.; Xiang, B.; Scannevin, R. H.; Huang, Z.;
Ma, J.; Burke, S. L.; Karnachi, P.; Rhodes, K. J.; Jackson, P. F. Bioorg. Med. Chem.
Lett. 2008, 18, 405.
8. Hu, J.; Miller, M. J. Tetrahedron Lett. 1995, 36, 6379.
9. Hu, J.; Miller, M. J. J. Am. Chem. Soc. 1997, 119, 3462.
10. Walz, A. J.; Miller, M. Org. Lett. 2002, 4, 2047.
11. Yang, S.-Y.; Lagu, B.; Wilson, L. J. J. Org. Chem. 2007, 72, 8123.
12. Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. H. Chem. Rev. 1999, 99,
2735.
13. Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18.
14. Yang, C.; Murray, W. M.; Wilson, L. J. Tetrahedron Lett. 2003, 44, 1783.
15. See compound 13d (R/S) in: Pikul, S.; Ohler, N. E.; Ciszewski, G.; Laufersweiler,
M. C.; Almstead, N. G.; De, B.; Natchus, M. G.; Hsieh, L. C.; Janusz, M. J.; Peng, S.
X.; Branch, T. M.; King, S. L.; Taiwo, Y. O.; Mieling, G. E. J. Med. Chem. 2001, 44,
2499.

6490
L. J. Wilson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6485–6490
16. Sparano, J. A. J. Clin. Oncol. 2004, 22, 4683.
II = non-mutageneic; CEREP 50 targets: <50% @ 10
discontinued due to business reasons.
lM. This program was
17. Experiments with in vitro ADME-Tox assays for 16 and 34 are as follows: hERG
binding: <50% @ 10 M; CYP450 inhibition (3A4, 2D6) <50% @ 1 M; AMES
l
l