Macrocyclic inhibitors of Factor XIa: Discovery of alkyl-substituted macrocyclic amide linkers with improved potency

Article abstract of DOI:10.1016/j.bmcl.2017.06.058

Optimization of macrocyclic inhibitors of FXIa is described which focused on modifications to both the macrocyclic linker and the P1 group. Increases in potency were discovered through interactions with a key hydrophobic region near the S1 prime pocket by

Full text of DOI:10.1016/j.bmcl.2017.06.058

Accepted Manuscript
Macrocyclic Inhibitors of Factor XIa: Discovery of Alkyl-Substituted Macro-
cyclic Amide Linkers with Improved Potency
James R. Corte, Wu Yang, Tianan Fang, Yufeng Wang, Honey Osuna, Amy
Lai, William R. Ewing, Karen A. Rossi, Joseph E. Myers Jr., Steven Sheriff,
Zhen Lou, Joanna J. Zheng, Timothy W. Harper, Jeffrey M. Bozarth, Yiming
Wu, Joseph M. Luettgen, Dietmar A. Seiffert, Mimi L. Quan, Ruth R. Wexler,
Patrick Y.S. Lam
PII:
S0960-894X(17)30660-1
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.06.058
BMCL 25092
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Accepted Date:
2 May 2017
21 June 2017
Please cite this article as: Corte, J.R., Yang, W., Fang, T., Wang, Y., Osuna, H., Lai, A., Ewing, W.R., Rossi, K.A.,
Myers, J.E. Jr., Sheriff, S., Lou, Z., Zheng, J.J., Harper, T.W., Bozarth, J.M., Wu, Y., Luettgen, J.M., Seiffert, D.A.,
Quan, M.L., Wexler, R.R., Lam, P.Y.S., Macrocyclic Inhibitors of Factor XIa: Discovery of Alkyl-Substituted
Macrocyclic Amide Linkers with Improved Potency, Bioorganic & Medicinal Chemistry Letters (2017), doi: http://
dx.doi.org/10.1016/j.bmcl.2017.06.058
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
Macrocyclic Inhibitors of Factor XIa: Discovery of Alkyl-Substituted
Macrocyclic Amide Linkers with Improved Potency.
James R. Cortea,
F
, Wu Yanga,, Tianan Fanga, Yufeng Wanga, Honey Osunaa, Amy Laia, William R.
F
Ewinga, Karen A. Rossia, Joseph E. Myers Jr.b, Steven Sheriff b, Zhen Louc, Joanna J. Zhengc, Timothy W.
Harperc, Jeffrey M. Bozarthc, Yiming Wuc, Joseph M. Luettgenc, Dietmar A. Seiffertc, Mimi L. Quana,
Ruth R. Wexlera and Patrick Y. S. Lama
aBristol-Myers Squibb Company, Research and Development, 350 Carter Road, Hopewell, New Jersey 08540, United States
bBristol-Myers Squibb Company, Research and Development, US Rt. 206 & Province Line Road, Princeton, New Jersey 08540, United States
cBristol-Myers Squibb Company, Research and Development, 311 Pennington-Rocky Hill Road, Pennington, New Jersey 08543-2130, United States
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Optimization of macrocyclic inhibitors of FXIa is described which focused on modifications to
both the macrocyclic linker and the P1 group. Increases in potency were discovered through
interactions with a key hydrophobic region near the S1 prime pocket by substitution of the
macrocyclic linker with small alkyl groups. Both the position of substitution and the absolute
stereochemistry of the alkyl groups on the macrocyclic linker which led to improved potency
varied depending on the ring size of the macrocycle. Replacement of the chlorophenyltetrazole
cinnamide P1 in these optimized macrocycles reduced the polar surface area and improved the
oral bioavailability for the series, albeit at the cost of a decrease in potency. ©2014 Elsevier
Science Ltd. All rights reserved.
Keywords:
Factor XIa inhibitors
FXIa
Activated Partial Thromboplastin Time
aPTT
Thrombosis
———
 Corresponding author. Tel.: +1-609-466-5030; fax: +1-609-818-3331; e-mail: james.corte@bms.com
 Corresponding author. Tel.: +1-609-466-5100; fax: +1-609-818-3331; e-mail: wu.yang@bms.com

Novel oral anticoagulants (NOACs), such as the direct
thrombin inhibitor (dabigatran) and direct Factor Xa
inhibitors (rivaroxaban, apixaban, and edoxaban), are
effective for the prevention and treatment of deep vein
thrombosis (DVT) and pulmonary embolism, and stroke
prevention in non-valvular atrial fibrillation.1 Rivaroxaban
is also effective for the secondary prevention of acute
coronary syndrome (ACS), however an increased risk of
bleeding and intracranial hemorrhage was observed.2,3 Safer
and more efficacious anticoagulants remains a medical need.
compared to either the 12- or 13-membered saturated amide
linkers 1 or 2. Importantly, these amide-linked macrocycles
exhibited excellent in vitro anticoagulant activity as
measured by an activated partial thromboplastin time
(aPTT) clotting assay (EC1.5x = 0.27-1.2 M).13 However,
this macrocyclic series suffered from poor oral
bioavailability as illustrated below for compounds 1-3, most
likely due to low permeability arising from high polar
surface area (PSA = 169 Å2 for macrocycles 1-3).14 Our
optimization strategy for the macrocyclic series was two-
fold. First, we wanted to probe the S1 prime pocket with
modifications to the unexplored alkyl backbone portion of
the macrocyclic linker with the goal of enhancing potency.
Second, we sought to address the poor oral bioavailability
for the series by improving the permeability via reduction of
PSA. In this communication, we describe the discovery of
alkyl-substituted macrocycles with improved FXIa affinity
as well as our initial efforts to improve oral bioavailability
for the series by replacing chlorophenyltetrazole cinnamide
(P1) with a smaller, less polar group.
Based on the lack of an overt bleeding phenotype for
individuals with a genetic FXI deficiency (Hemophilia C),
inhibition of the blood coagulation enzyme Factor XIa
(FXIa), the activated form of the zymogen Factor XI (FXI),
could provide an effective anticoagulant therapy with a safer
bleeding profile.4 Individuals with severe FXI deficiency
appear to be protected from certain cardiovascular events as
they show reduced rates of both ischemic stroke and DVT.5
Importantly, from a safety perspective, the majority of
individuals with Hemophilia C display a minor bleeding
tendency which is in contrast to the bleeding episodes
observed with either a severe Factor VIII deficiency
X-ray crystallography of macrocycles 1 and 2 confirmed a
key hydrogen bond, which had been suggested by molecular
modeling, between the NH of the amide moiety in the
macrocyclic linker and the carbonyl of Leu41 in FXIa
resulting in a significant increase in FXIa affinity.11,15 The
crystal structures also revealed that the alkyl backbone of
the macrocyclic amide linker did not fully access the
hydrophobic region, represented by the yellow surface in the
S1 prime pocket (Figure 1).16 Owing to the fact that
additional potency enhancements might be achievable by
targeting this hydrophobic region, we explored alkyl
substitutions on the linker in both the 12- and 13-membered
macrocycles.
(Hemophilia A) or
a severe Factor IX deficiency
(Hemophilia B).6
Several FXIa inhibitors, such as -ketothiazole
peptidomimetic,7 4-carboxy-2-azetidinone,8 tetrahydro-
quinoline,9 and chloroimidazole derivatives,10 were shown
to be effective in a variety of animal thrombosis models
with minimal effects on bleeding time. Moreover, we have
shown that at equivalent antithrombotic doses,
a
tetrahydroquinoline FXIa inhibitor displayed less bleeding
in a rabbit cuticle bleeding time model when compared to
the direct thrombin inhibitor dabigatran.9b These preclinical
studies support targeting FXIa as a means to achieve greater
protection from thrombosis without increasing the bleeding
risk.
Table 1. 12- and 13-membered macrocyclic FXIa inhibitors (the
macrocyclic amide linker is highlighted in blue).
Figure 1. Overlay of X-ray structures of 1 (yellow) and 2 (cyan) in
FXIa. Yellow surfaces highlight key hydrophobic regions within the
FXIa binding site as determined by SiteMap.16
aPTT
n
FXIa Ki
(nM)12
EC1.5x
(M)13
0.30
Compd #
Alkene
saturated
saturated
E-alkene
(ring size)
We initially explored methyl substitution at the -, -, and
-positions in the 13-membered saturated linker (Table 2).
Compound 2 and its corresponding chloroimidazole analog
4 were included for comparison. From our previously
disclosed SAR, adding a chloro group to the 5 position of
the imidazole generally provides a 5-fold increase in FXIa
affinity and the chloroimidazole scaffold is a common
structural motif in our more potent macrocyclic analogs.11,17
Substitution at the -position relative to the amide resulted
in a loss in FXIa affinity and aPTT activity for both
diastereomers 5 and 6. The stereochemistry of a methyl
group at the -position had a more dramatic effect on FXIa
1
2
3
0 (12)
1 (13)
1 (13)
1.0
3.2
1.2
0.16
0.27
We recently disclosed the discovery of a novel series of 12-
and 13-membered macrocyclic FXIa inhibitors 1-3 (Table
1).11 The macrocyclic amide linker, highlighted in blue, was
critical for FXIa binding affinity. The presence of the E-
alkene in the 13-membered amide-linker 3 reduced the
conformational entropy and showed improved FXIa affinity

affinity. Whereas diastereomer (R)--Me 7 exhibited a 6- to
7-fold loss in FXIa affinity, the corresponding diastereomer
(S)--Me 818 exhibited a 4.5-fold increase in FXIa affinity
compared to the unsubstituted analog 2. In spite of the
improvement in FXIa affinity, the potency in the aPTT
clotting assay was unchanged. The lack of change in the
aPTT clotting assay is likely at least in part due to higher
protein binding with the installation of the methyl group
[human protein binding (%bound) for 2 is 96.2% and 8 is
98.0%]. Incorporation of the (S)--Me group in
chloroimidazole 9 resulted in an 8-fold improvement in
FXIa affinity (FXIa Ki = 0.06 nM), however only a modest
increase in aPTT activity was observed (compare 4 vs 9).
Substitution at the -position led to a loss in FXIa affinity
and aPTT activity for both diastereomers 10 and 11. In
summary, a -methyl group improved FXIa affinity in the
13-membered saturated linker.
Substitution at the -position was explored next. Whereas
the -Me 15 (diastereomer A) showed a modest loss in both
FXIa affinity and aPTT activity compared to 3, the
corresponding -Me 16 (diastereomer B) led to a 16-fold
loss in FXIa affinity and a 4-fold loss in activity in the aPTT
clotting assay. In summary, alkyl substitution on the 13-
membered unsaturated linker did not lead to an
improvement in potency. Since potency enhancements were
not seen, the corresponding chloroimidazoles analogs of 14
and 15 were not prepared.
Table 3. Substitutions on the 13-membered unsaturated macrocyclic
linker
Table 2. Substitutions on the 13-membered saturated macrocyclic
linker
aPTT
EC1.5x
(M)13
0.27
0.28
0.98
FXIa
Ki
Compd #
R
X
(nM)12
0.16
0.03
2.0
3
H
H
H
Cl
H
12
13a
(S)--Meb
(R)--Meb
14a
H
0.13
0.38
15a
H
0.33
0.67
-Me
(diastereomer A)
16a
(diastereomer B)
aPTT
EC1.5x
(M)13
1.2
FXIa
Ki
Compd #
R
X
H
2.7
1.2
-Me
(nM)12
3.2
aAll diastereomers were separated following the ring-closing metathesis
step. The diastereomeric excess of all final compounds was ≥ 95% de
as determined by 1H NMR. bThe absolute stereochemistry was assigned
based on an X-ray co-crystal of 14 with FXIa.
2
H
H
H
Cl
4
0.47
0.92
5a
-Me
H
H
6.5
31
5.3
32
(diastereomer A)
6a
-Me
(diastereomer B)
An X-ray crystal structure of compound 8 bound to human
FXIa active site (2.1 Å resolution, Figure 2A) was obtained
and revealed the absolute stereochemistry of the -methyl to
be in the S-configuration.18 An overlay of compound 8
with the X-ray structure of compound 2 is shown in Figure
2B and highlights the interaction of the methyl group with
the hydrophobic region (yellow surface) in the S1 prime
pocket. Interestingly, an overlay of compound 8 and
compound 14 (Figure 2C) suggests that the -methyl in the
unsaturated amide linker could still access the hydrophobic
pocket. However this was not supported by the SAR
described in Table 3 as 14 did not offer additional
improvement in potency over 3.
7a
H
H
Cl
21
0.69
0.06
27
1.1
0.57
(R)--Meb
(S)--Meb
(S)--Meb
8a
9
10a
H
5.9
2.5
-Me
(diastereomer A)
11a
(diastereomer B)
H
18
12
-Me
aAll diastereomers were separated. The diastereomeric excess for 5 was
91% de, 6 was ≥ 95% de, 10 was 79% de, and 11 was 89% de as
1
determined by H NMR. The diastereomeric excess of 7 and 8 were ≥
99% de as determined by analytical chiral hplc. bThe absolute
stereochemistry was assigned based on an X-ray co-crystal of 8 with
FXIa (Figure 2A).
Since a -methyl group improved FXIa affinity in the 13-
membered saturated linker, it was installed in the 13-
membered unsaturated linker (Table 3). The unsubstituted
amide linkers containing imidazole
3
and the
chloroimidazole 12 were included for comparison. In
contrast to the 13-membered saturated linker, substitution at
the -position on the unsaturated linker did not improve
FXIa affinity. The (S)--Me 13 exhibited a 12-fold loss in
FXIa affinity and the corresponding (R)--Me 1419 was
equipotent to the unsubstituted linker 3 and only slightly
less active than the saturated linker (S)--Me 9. It should be
noted that the configuration of the -stereocenter is the same
for both (R)--Me 14 and (S)--Me 9. However the S
designation changed to R based on the presence of the
alkene in the macrocyclic linker which altered the
prioritization in the Cahn-Ingold-Prelog system.

1
H
H
H
Cl
H
H
H
H
H
1.0
0.39
0.07
1.1
0.18
7.0
0.30
0.39
1.2
0.63
0.52
4.0
17a
(R)--Meb
(R)--Meb
(S)--Meb
(R)--Etb
(S)--Etb
(S)--iPrc
(R)--iPr
18
19
20
21
22
0.14
21
1.1
64
23
24d
H
36
37
-Me
(diastereomer A)
25e
(diastereomer B)
H
35
31
-Me
aAll diastereomers were separated. The diastereomeric excess of all
final compounds was >95% de as determined by 1H NMR. bThe
absolute stereochemistry was assigned based on X-ray co-crystals with
FXIa. cThe configuration of the -stereocenter is the same as for (R)-
-Me 17 and (R)--Et 20, however the R designation changed to S
based on the iPr group which altered the prioritization in the Cahn-
Ingold-Prelog system. dA 4:1 mixture of diastereomers. eA 2.6:1
mixture of diastereomers.
(A)
Since potency enhancements were achieved with 
substitution on the 13-membered saturated linker, we next
explored substitutions on the 12-membered saturated linker
(Table 4). In contrast to the saturated 13-membered
macrocycle, addition of a methyl group at the -position on
the saturated 12-membered macrocycle resulted in over a
30-fold loss in FXIa affinity (24 and 25 vs 1). However,
when the methyl group was introduced at the -position, the
(R)-isomer 17 improved FXIa affinity by 2-fold while the
(S)-isomer 19 did not have a significant impact on FXIa
affinity. Incorporation of the (R)--Me group in
chloroimidazole 18 further improved FXIa Ki by 5-fold to
70 pM, albeit with a 3-fold loss in potency in the aPTT
clotting assay (compare 17 and 18). The 12-membered
macrocycle (R)--Me 18 was equipotent to the 13-
membered macrocycle (S)--Me 9. Furthermore, as the size
of the -alkyl groups increased from methyl (17) to ethyl
(20) to isopropyl (22), the FXIa affinity also improved
stepwise from 0.4 nM to 0.2 nM and 0.1 nM. However,
despite these potency enhancements with the larger alkyl
group, a loss in aPTT clotting activity was observed.
Overall, substitution of small alkyl groups at the -position
of the saturated amide linker improved FXIa affinity in the
12-membered macrocycle.
(B)
(C)
Figure 2. (A) X-ray crystal structure of 13-membered saturated
macrocycle compound 8 bound to Factor XIa with omit electron density
contoured at 3 rmsd (gray mesh). The red spheres depict water
molecules and the dotted lines depict hydrogen bonds. (B) Overlay of
X-ray crystal structures of 8 (orange) and 2 (cyan) in Factor XIa. (C)
Overlay of X-ray crystal structures of 8 (orange) and 14 (wheat) in
Factor XIa. Yellow surfaces highlight key hydrophobic regions within
the FXIa binding site as determined by SiteMap.16
Table 4. Substitutions on the 12-membered saturated macrocyclic
linker
(A)
aPTT
EC1.5x
(M)13
FXIa
Ki
Compd #
R
X
(nM)12

(B)
Figure 3. (A) X-ray crystal structure of 12-membered macrocycle
compound 20 bound to Factor XIa with omit electron density contoured
at 3 rmsd (gray mesh). The red spheres depict water molecules, the
dotted lines depict hydrogen bonds, and ethylene diol is an artifact of
the flash-cooling procedure. (B) Overlay of X-ray crystal structures of
20 (purple) and 8 (orange) in Factor XIa. Yellow surfaces highlight key
hydrophobic regions within the FXIa binding site as determined by
SiteMap.16

Table 5. P1 modifications and rat PK for the 12- and 13-membered macrocycles
n
Caco-2
AB/BA
(nm/s)
<15/58
<15/111
18/251
Rat
PK
F%c
0
aPTT EC1.5x
FXIa Ki
(nM)12
PSAa
(Å2)
HLMb
T1/2 (min)
Compd #
(ring size)
alkene
R
(M)13
12d
26
27
28
29
1 (13)
0 (12)
1 (13)
1 (13)
1 (13)
E-alkene
saturated
saturated
E-alkene
E-alkene
H
0.03
10
44
0.28
2.0
12
2.7
6.6
169
125
125
125
125
58
8
45
35
17
NTe
NTe
16
(R)--Me
(S)--Me
(R)--Me
H
12
<15/189
<15/200
16
41
aPSA = polar surface area. bHLM = human liver microsome stability. Compounds were dosed in rat in an N-in-1 format.23 Vehicle for both iv and po:
c
70% PEG400; 20% water; 10% ethanol. dCompound 12 contains the chlorophenyltetrazole cinnamide P1 and was included for comparison. eNT = not
tested.
An X-ray crystal structure of 20 bound to human FXIa
active site (2.6 Å resolution, Figure 3A) was obtained and
revealed the absolute stereochemistry of the -ethyl in the
12-membered macrocycle to be the R-configuration (the
absolute stereochemistry of the -methyl and -iPr were
also confirmed by x-ray structures).20 Interestingly, the (R)-
stereochemistry is the opposite stereochemistry compared to
the (S)--Me 8 in the saturated 13-membered macrocycle.
An overlay of compound 20 and compound 8 is shown in
Figure 3B and highlights the interaction of the alkyl groups
with the hydrophobic region (yellow surface) in the S1
prime pocket. Despite the differences in both the position of
substitution and the stereochemistry, the (S)--Me 8 on the
13-membered saturated amide linker and the (R)--Et 20 on
the 12-membered saturated amide linker are able to access
the same hydrophobic region near the S1 prime pocket.
smaller benzamide, the 12-membered macrocycle was more
potent (26 vs 27). The 13-membered macrocycles 28 and
29, possessing the E-alkene, were comparable to 26. The
loss in potency with the smaller benzamide P1 can be
explained by the fact that the phenyl moiety does not extend
as deep into the S1 pocket and cannot engage key residues
as was seen with the larger chlorophenyltetrazole P1.22 Even
though replacement of the chlorophenyltetrazole cinnamide
resulted in a significant loss in FXIa affinity, these analogs
(26-29) have a
lower PSA as shown in Table 5.
Permeability for this set of macrocycles was improved
compared to 12 based on the Caco-2 B to A permeability;
however the efflux ratio was very high. Macrocyclic
compounds 28 and 29 were evaluated in
a
rat
pharmacokinetic model, using an N-in-one cassette dosing
protocol23 (Table 5). Replacing the chlorophenyltetrazole
with 2,6-difluoro-4-methyl benzamide resulted in
compounds with low clearance, short half-life, and
improved oral bioavailability (%F = 16 - 41). (Please see
Supplementary Data for detailed pharmacokinetic data)
Multiple alkyl-substituted macrocycles, (S)--Me 9, (R)--
Me 14, and (R)--Me 18, were identified that exhibited
exquisite FXIa affinity and were comparable to the
unsubstituted 13-membered macrocycle 12. With potent
macrocycles in hand, we next sought to address the poor
oral bioavailability for the series which we attributed to low
permeability arising from high polar surface area (PSA).14
For example, unsubstituted 13-membered macrocycle 12
exhibited poor Caco-2 permeability (A to B/B to A < 15/58
nm/sec) and a high PSA (169 Å2), and was found to not
The synthesis of the -methyl substituted 13-membered
macrocycles are described in Scheme 1. For the 13-
membered macrocycles the key macrocyclization step
utilized a ring-closing olefin metathesis (RCM) strategy.24
Aniline 1a11 was coupled with 3-methylpent-4-eonic acid
using T3P which gave amide 1b as a mixture of
diastereomers. Amide 1b was pretreated with p-TsOH to
form the imidazolium ion and then cyclized via RCM using
the 2nd generation Grubbs II catalyst which gave the 13-
membered macrocycle 1c as a mixture of E and Z-alkene
isomers as well as a mixture of (R)- and (S)-stereoisomers.25
Hydrogenation of 1c over palladium on carbon, followed by
deprotection with 4M HCl at elevated temperature afforded
the corresponding amine as a mixture of diastereomers.
Chlorophenyltetrazole cinnamide was installed by coupling
amine with activated carboxylic ester 1i which gave,
following separation of diastereomers by chiral preparatory
have
oral
bioavailability
in
rat.
Since
the
chlorophenyltetrazole cinnamide P1 was a major contributor
to high PSA, new P1 groups with lower PSA values were
desired. Previously we had shown that oral bioavailability
could be achieved in our linear series of FXIa inhibitors
with less polar benzamide P1 groups.21 As a result, the
macrocycles from 18, 9, 14, and 12 were combined with
2,6-difluoro-4-methyl benzamide (Table 5). Replacing
chlorophenyltetrazole cinnamide in the 12-membered
macrocycle 18 with benzamide gave 26, which resulted in a
significant loss (>140-fold) in FXIa affinity but only a
modest loss in potency in the aPTT clotting assay.
Installation of benzamide in the 13-membered saturated
macrocycle led to 27 with an even greater loss (>700-fold)
in FXIa affinity. Interestingly, even though the 13-
membered macrocycle 9 and 12-membered macrocycle 18
both containing the chlorophenyltetrazole cinnamide were
comparable in FXIa affinity, when combined with the
hplc, (R)--Me
7
(not shown) and (S)--Me 8.
Alternatively, the stereoisomers can be separated by reverse
phase preparatory hplc following the RCM step which gave
1d. Hydrogenation of 1d over palladium on carbon,
followed by chlorination of the imidazole with N-
chlorosuccinimide at elevated temperature provided
chloroimidazole 1e. Global deprotection as described above

provided amine 1g, which was then coupled with the
appropriately substituted carboxylic acids to give 9 and 27.
Analogs 14 and 28 were prepared following the sequence
described above. The - and -methyl substituted 13-
membered macrocycle analogs were prepared as described
above by replacing 3-methylpent-4-eonic acid with 2-
methylpent-4-eonic acid and 4-methylpent-4-eonic acid,
respectively.
generation Grubbs II catalyst provided compounds 2c.
Hydrogenation of 2c simultaneously reduced the double
bond and deprotected the benzyl ester, which upon
hydrolysis of the trifluoroacetamide gave acid anilines 2d.
Macrolactamization of 2d under dilute coupling conditions
with BOP and DMAP afforded the desired 12-membered
macrocycles as a mixture of diastereomers which were
separated by reverse phase preparatory hplc. Deprotection
and installation of the P1 groups as described in Scheme 1
provided analogs 17, 20, and 22. Analog 18 was obtained by
chlorination of 17 and intermediate 2e was converted to 26
following a similar sequence. The -methyl substituted 12-
membered amide-linked analogs were prepared as described
above by replacing the benzyl ester of -substituted
butenoic acids 2b with benzyl 3-methylbut-3-enoate.
The synthesis of the -alkyl substituted 12-membered
macrocycles is described in Scheme 2. In contrast to the 13-
membered ring macrocycles described above, the 12-
membered macrocycles were prepared via
a
key
macrolactamization step. Aniline 1a was protected as
trifluoroacetamide 2a. Intermolecular metathesis of 2a with
the benzyl ester of -substituted butenoic acids 2b using 2nd
Scheme 1. Reagents and conditions: (a) 3-methylpent-4-enoic acid, T3P, Hunig’s base, EtOAc, -10 C to rt, 95%; (b) Second generation Grubbs II
catalyst (40 mol%), pTsOH, DCM, reflux, 68%; (c) 10% Pd/C, H2, MeOH; (d) 4M HCl in dioxane, 50 C or 75 C, 55% over two steps; (e) 1i, Hunig’s
base, DMF, 55%, then chiral preparatory hplc to separate diastereomers; (f) reverse phase chromatography to separate diastereomers, 10%; (g) 10% Pd/C,
H2 (55 psi), EtOH, 78%; (h) NCS, ACN, CHCl3, 65 C, 35-55%; (i) 1j or 2,6-difluoro-4-methylbenzoic acid, EDC, HOBt, Hunig’s base or triethylamine,
DMF, 32-44% over two steps.
Scheme 2. Reagents and conditions: (a) 2,2,2-trifluoroacetic anhydride, triethylamine, EtOAc, 0 C 97%; (b) Second generation Grubbs II catalyst (40
mol%), pTsOH, DCM, reflux or microwave 150 C, 18-49%; (c) 10% Pd/C, H2, MeOH, 69-100%; (d) LiOH, MeOH, 60 C, 74-100%; (e) BOP, DMAP,

DIEA, DMF/DCM, then reverse phase HPLC separation of diastereomers, 15-35%; (f) 4M HCl in dioxane, 50 C, 100%; (g) 1i, Hunig’s base, DMF, 48-
73% or 2,6-difluoro-4-methylbenzoic acid, EDC, HOBt, Hunig’s base or triethylamine, DMF; (h) NCS, ACN, CHCl3, 65 C, 35-55%.

Bozarth, J. M.; Crain, E. J.; Harper, T.; Luettgen, J. M.; Myers, J.
E.; Ramamurthy, V.; Rossi, K. A.; Sheriff, S.; Watson, C. A.;
Wei, A.; Zheng, J. J.; Seiffert, D. A.; Wexler, R. R.; Quan, M. L.
ACS Med. Chem. Lett. 2015, 6, 590.
In conclusion, the macrocyclic FXIa series described herein was
optimized by modifying both the macrocyclic linker and the P1
group. SAR exploration of macrocyclic linker identified key
alkyl substitutions which improved FXIa affinity by interacting
with a hydrophobic region near the S1 prime pocket as
demonstrated by ligand bound X-ray crystal structures. Both the
position of substitution and the absolute stereochemistry of the
alkyl moiety on the macrocyclic linker varied with the size of the
macrocyclic ring. The (S)--Me group was more potent in the 13-
membered saturated linker while the (R)--Me moiety was more
potent in the 12-membered saturated linker. Replacement of the
P1 group, chlorophenyltetrazole cinnamide, was a successful
approach for reducing the PSA and improving the oral
bioavailability for the macrocyclic series. However, this P1
modification came with a loss in potency. Further optimization of
the P1 to improve FXIa potency and maintain the oral
bioavailability will be reported in due course.
11. Corte, J. R.; Fang, T.; Osuna, H.; Pinto, D. J. P.; Rossi, K. A.;
Myers, J. E.; Sheriff, S.; Lou, Z.; Zheng, J. J.; Harper, T. W.;
Bozarth, J. M.; Wu, Y.; Luettgen, J. M.; Seiffert, D. A.; Decicco,
C. P.; Wexler, R. R.; Quan, M. L. J. Med. Chem. 2017, 60, 1060.
12. Ki values were obtained from purified human enzyme at 37 °C
and were averaged from multiple determinations (n=2), as
described in Refs. 9b and 11.
13. aPTT (activated partial thromboplastin time) in vitro clotting
assay was performed in human plasma as described in Refs. 9b
and 11. The reported EC1.5x values are the FXIa inhibitor plasma
concentrations which produce a 50% increase in the clotting time
relative to the clotting time in the absence of the inhibitor.
14. Ertl, P. Polar Surface Area. In Molecular Drug Properties.
Measurement and Prediction; Mannhold, R., Ed.; Wiley-VCH
Verlag GmbH & Co. KGaA: Weinheim, 2008; pp 111-126.
15. PDB deposition number for 1 is 5Q0D.
16. The yellow surfaces highlight key hydrophobic regions within the
FXIa binding site as determined by SiteMap, please see Halgreen,
T. Chem. Biol. Drug Des. 2007, 69, 146.
Acknowledgements: The authors thank the Department of
Discovery Synthesis for the synthesis of intermediates at BMS
Biocon Research Center; the Department of Lead Discovery &
Optimization; and Mojgan Abousleiman for in vitro work.
17. Hangeland, J. J.; Friends, T. J.; Rossi, K. A.; Smallheer, J. M.;
Wang, C.; Sun, Z.; Corte, J. R.; Fang, T.; Wong, P. C.; Rendina,
A. R.; Barbera, F. A.; Bozarth, J. M.; Luettgen, J. M.; Watson, C.
A.; Zhang, G.; Wei, A.; Ramamurthy, V.; Morin, P. E.; Bisacchi,
G. S.; Subramaniam, S.; Arunachalam, P.; Mathur, A.; Seiffert,
D. A.; Wexler, R. R.; Quan, M. L. J. Med. Chem. 2014, 57, 9915.
18. The absolute stereochemistry was assigned based on an X-ray co-
crystal of (S)--Me 8 with FXIa (Figure 2A). The PDB
deposition number for (S)--Me 8 is 5Q0E.
19. The absolute stereochemistry was assigned based on an X-ray co-
crystal of (R)--Me 14 with FXIa. The PDB deposition number
for (R)--Me 14 is 5Q0F.
20. The absolute stereochemistry was assigned based on an X-ray co-
crystal of (R)--Et 20 with FXIa. The PDB deposition number
for (R)--Et 20 is 5Q0G.
21. Pinto, D. J. P; Smallheer, J. M.; Corte, J. R.; Austin, E. J. D.;
Wang, C.; Fang, T., Smith, L. M.; Rossi, K. A.; Rendina, A. R.;
Bozarth, J. M.; Zhang, G.; Wei, A.; Ramamurthy, V.; Sheriff, S.;
Myers, J. E.; Morin, P. E.; Luettgen, J. M.; Seiffert, D. A.; Quan,
M. L.; Wexler, R. R. Bioorg. Med. Chem. Lett. 2015, 25, 1635.
22. For the chlorophenyltetrazole cinnamide P1: The chlorophenyl
portion fills the S1 pocket with the chlorine atom forming a -Cl
interaction with Tyr228; the tetrazole portion extends out of the
S1 pocket and interacts with the Cys191-Cys219 disulfide bridge;
and the tetrazole N4 nitrogen forms an H-bond, via a series of
water molecules, to Gly218. The smaller benzamide P1 does not
interact with these key residues. Please see Supplementary data
for an overlay of compounds 28 and 14.
23. He, K.; Qian, M.; Wong, H.; Bai, S. A.; He, B.; Brogdon, B.;
Grace, J. E.; Xin, B.; Wu, J.; Ren, S. X.; Zeng, H.; Deng, Y.;
Graden, D. M.; Olah, T. V.; Unger, S. E.; Luettgen, J. M.; Knabb,
R. M.; Pinto, D. J.; Lam, P. Y. S.; Duan, J.; Wexler, R. R.;
Decicco, C. P.; Christ, D. D.; Grossman, S. J. J. Pharm. Sci. 2008,
97 (7), 2568.
A. Supplementary Data.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/
References and Notes
1.
(a) Morais, J.; De Caterina, R. Cardiovasc. Drugs Ther. 2016, 30,
201-214. (b) Schaefer, J. K.; McBane, R. D.; Wysokinski, W. E.
Ann. Hematol. 2016, 95, 437. (c) Levy, J. H.; Spyropoulos, A. C.;
Samama, C. M.; Douketis, J. J. Am. Coll. Cardiol. Intv. 2014, 7,
1333.
2.
3.
Gulpen, A. J. W.; Cate-Hoek, A. J. T.; Cate, H. T. Expert Review
of Cardiovascular Therapy, 2016.
In 2013 the European Union approved rivaroxaban for secondary
prevention in ACS despite an increased risk of bleeding and
intracranial hemorrhage. Please see (a) Behnes, M.; Fastner, C.;
Ansari, U.; Akin, I. Cardiovasc. Hematol. Disord.: Drug Targets
2015, 15, 101. (b) Oldgren, J.; Wallentin, L.; Alexander, J. H.;
James, S.; Jonelid, B.; Steg, G.; Sundstrom, J. Eur. Heart J. 2013,
34, 1670.
4.
5.
(a) Bane, C. E.; Gailani, D. Drug Discovery Today 2014, 19 (9),
1454. (b) Muller, F.; Gailani, D.; Renne, T. Curr. Opin. Hematol.
2011, 18, 349.
(a) Salomon, O.; Steinberg, D. M.; Koren-Morag, N.; Tanne, D.;
Seligsohn, U. Blood 2008, 111 (8), 4113. (b) Salomon, O.;
Steinberg, D. M.; Zucker, M.; Varon, D.; Zivelin, A.; Seligsohn,
U. Thromb. Haemost. 2011, 105, 269.
6.
7.
Bane, C. E.; Neff, A. T.; Gailani, D. Factor XI deficiency or
hemophilia C.
In Hemostasis and Thrombosis: Practical
24. Van Lierop, B. J.; Lummiss, J. A. M.; Fogg, D. E. Ring-closing
metathesis. In Olefin Metathesis: Theory and Practice; Grela,
K., Ed.; John Wiley & Sons Inc.: Hoboken, New Jersey, 2014; pp
85-152.
Guidelines in Clinical Management; Saba, H.I., Roberts, H. R.,
Eds.; John Wiley & Sons, Ltd.: Oxford, 2014; pp 71-81.
Lin, J.; Deng, H.; Jin, L.; Pandey, P.; Quinn, J.; Cantin, S.;
Rynkiewicz, M. J.; Gorga, J. C.; Bibbins, F.; Celatka, C. A.;
Nagafuji, P.; Bannister, T. D.; Meyers, H. V.; Babine, R. E.;
Hayward, N. J.; Weaver, D.; Benjamin, H.; Stassen, F.; Abdel-
Meguid, S. S.; Strickler, J. E. J. Med. Chem. 2006, 49, 7781.
(a) Schumacher, W. A.; Seiler, S. E.; Steinbacher, T. E.; Stewart,
A. B.; Bostwick, J. S.; Hartl, K. S.; Liu, E. C.; Ogletree, M.L.
Eur. J. Pharmacol. 2007, 570, 167. (b) Wong, P. C.; Crain, E. J.;
Watson, C. A.; Schumacher, W. A. J. Thromb. Thrombolysis
2011, 32, 129.
(a) Quan, M .L.; Wong, P. C.; Wang, C.; Woerner, F.; Smallheer,
J. M.; Barbera, F. A.; Bozarth, J. M.; Brown, R. L.; Harpel, M. R.;
Luettgen, J. M.; Morin, P. E.; Peterson, T.; Ramamurthy, V.;
Rendina, A. R.; Rossi, K. A.; Watson, C. A.; Wei, A.; Zhang, G.;
Seiffert, D.; Wexler, R. R. J. Med. Chem. 2014, 57(3), 955. (b)
Wong, P. C.; Quan, M. L.; Watson, C. A.; Crain, E. J.; Harpel, M.
R.; Rendina, A. R.; Luettgen, J. M.; Wexler, R. R.; Schumacher,
W. A.; Seiffert, D. A. J. Thromb. Thrombolysis 2015, 40(4), 416.
25. Chen, Y.; Dias, H. V. R.; Lovely, C. J. Tetrahedron Lett. 2003,
44, 1379.
8.
9.
10. Hu, Z.; Wong, P. C.; Gilligan, P. J.; Han, W.; Pabbisetty, K. B.;

Graphical Abstract
Macrocyclic Inhibitors of Factor XIa:
Discovery of Alkyl-Substituted Macrocyclic
Amide Linkers with Improved Potency.
James R. Corte, Wu Yang, Tianan Fang, Yufeng Wang, Honey Osuna, Amy Lai, William R. Ewing, Karen A.
Rossi, Joseph E. Myers Jr., Steven Sheriff, Zhen Lou, Joanna J. Zheng, Timothy W. Harper, Jeffrey M. Bozarth,
Yiming Wu, Joseph M. Luettgen, Dietmar A. Seiffert, Mimi L. Quan, Patrick Y. S. Lam, and Ruth R. Wexler