Structure and property based design of factor Xa inhibitors: Pyrrolidin-2-ones with aminoindane and phenylpyrrolidine P4 motifs

Article abstract of DOI:10.1016/j.bmcl.2011.01.131

The rational design, syntheses and evaluation of potent sulfonamidopyrrolidin-2-one-based factor Xa inhibitors incorporating aminoindane and phenylpyrrolidine P4 motifs are described. These series delivered highly potent anticoagulant compounds with excellent oral pharmacokinetic profiles; however, significant time dependant P450 inhibition was an issue for the aminoindane series, but this was not observed with the phenylpyrrolidine motif, which produced candidate quality molecules with potential for once-daily oral dosing in humans.

Full text of DOI:10.1016/j.bmcl.2011.01.131

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1582–1587
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure and property based design of factor Xa inhibitors: Pyrrolidin-2-ones
with aminoindane and phenylpyrrolidine P4 motifs
Robert J. Young a, , Carl Adams a, Mike Blows a, David Brown a, Cynthia L. Burns-Kurtis b, Chuen Chan a,
⇑
Laiq Chaudry a, Máire A. Convery a, David E. Davies a, Anne M. Exall a, Graham Foster a, John D. Harling a,
Eric Hortense a, Stephanie Irvine a, Wendy R. Irving a, Steve Jackson a, Savvas Kleanthous a,
Anthony J. Pateman a, Angela N. Patikis a, Theresa J. Roethka b, Stefan Senger a, Gary J. Stelman b,
John R. Toomey b, Robert I. West a, Caroline Whittaker a, Ping Zhou a, Nigel S. Watson a
a GlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, United Kingdom
b GlaxoSmithKline, 709 Swedeland Road, King of Prussia, PA 19406, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The rational design, syntheses and evaluation of potent sulfonamidopyrrolidin-2-one-based factor Xa
inhibitors incorporating aminoindane and phenylpyrrolidine P4 motifs are described. These series deliv-
ered highly potent anticoagulant compounds with excellent oral pharmacokinetic profiles; however, sig-
nificant time dependant P450 inhibition was an issue for the aminoindane series, but this was not
observed with the phenylpyrrolidine motif, which produced candidate quality molecules with potential
for once-daily oral dosing in humans.
Received 17 December 2010
Revised 26 January 2011
Accepted 27 January 2011
Available online 1 February 2011
Keywords:
Factor Xa
Ó 2011 Elsevier Ltd. All rights reserved.
Oral inhibitors
Property based design
Structure based design
P450 inhibition
The approvals for the selective thrombin inhibitor dabigatran1
and factor Xa (fXa) inhibitor rivaroxaban,2 were significant mile-
stones in the search for new, small molecule, treatments for throm-
bosis, through targeting single protease enzymes in the blood
coagulation cascade. These selective medicines do not require
extensive monitoring due to their reduced bleeding liability com-
pared with established agents such as the coumarins.3 This highly
competitive field has seen many selective inhibitors of the
O
H
O
S
O
S
N
S
R
H
N
H
N
Cl
O
O
O
O
N
O
S
O
N
X = H or F
N
Cl
S1 pocket
X
O
F
1
2
3
S1 pocket
N
A
C
N
B
O
NHMe
S4 pocket
S4 pocket
Figure 1. Clinical candidate 1, aMBA 2 and generic constrained structures 3.
⇑
Corresponding author. Tel.: +44 1438 768372.
E-mail address: Rob.J.Young@gsk.com (R.J. Young).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.131

R. J. Young et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1582–1587
1583
NHBoc
NHBoc
NH2
e (for R2 =H)
c, d
a,b
O
O
N
N
4
O
6
5
R1
N
R2
O
[f], g, h, [j]
(for R1 = R2 = Me)
R3
g, h, [i, j]
O
S
O
O
HN
7
N
R2
R1
N
N
8
R1
CF3
O
Scheme 1. Reagents and conditions: (a) Boc-Met-OH, HATU, DIPEA, DCM, rt 95%; (b) MeI, MeCN, rt, then Cs2CO3, MeCN, 60 °C, 67%; (c) NaBH4, MeOH, rt, 95%; (d) MeSO2Cl,
Et3N, DCM, ꢀ76 °C; then R1R2NH, THF, rt, 40%; (e) TFAA, DCM, rt, 62%; [(f) chiral hplc]; (g) HCl, MeOH, rt, 95%; (h) R3SO2Cl, pyridine, MeCN, rt, 60–90%, (i) Na2CO3, aq MeOH, rt,
45%; (j) for N-TIPS-protected indole-P1; Et4NF, THF, rt, 90%.
trypsin-like serine proteases thrombin and fXa progress to the
clinic and further trials continue.2c,3f
sign.5 Through strict attention to molecular properties, acceptable
levels of oral pharmacokinetic exposure were achieved with these
combinations, whereby moderately basic P4 groups were exploited
to modulate hydrophobicity. However, time-dependent inhibition
(TDI) of Cyp3A4 curtailed the progress of the recently disclosed
Our programme of work on fXa inhibitors has focussed on the
sulfonamidopyrrolidin-2-one template, exemplified by our first
clinical candidate 1,4 featuring the morpholine alanyl amide P4 li-
gand (Fig. 1). The structurally similar S1 binding pockets of throm-
bin and fXa normally bind a basic amino acid, forming a salt bridge
with asparate189; however, chloroaromatics have been used to ex-
ploit binding to tyrosine228 in our fXa series4,5 and, for example, in
rivaroxaban.2 This enabled the discovery of neutral compounds
with enhanced oral bioavailability over earlier basic inhibitors.3
Whilst SAR studies4a for ligands binding in the S1 pocket have
shown limited scope for variations that maintain potency, plas-
ma-based activity and good DMPK properties, the back-up pro-
gramme identified a number of novel P4 motifs (replacing the
morpholine alanyl amide of 1, with the aim of improving pharma-
cokinetic half-life) through the exploitation of structure-based de-
a
-methyl benzylamine (a
MBA) series,5c exemplified by 2. Growth
and fusion of the benzylic amine motif into a cyclic structure rep-
resented an attractive strategic extension; three such constrained
series are generically represented as 3 A, B and C in Figure 1.
Aminoindanes and phenylpyrrolidines, represented by A and B,
respectively, are described herein; the tetrahydroisoquinolines
and related structures denoted by C are described in the accompa-
nying paper.
Computational docking studies using FLO6 indicated that each
of these structures was likely to maintain good contacts in the S4
pocket. Docking studies also provided the impetus to synthesise
6-indolylsulfonyl P1 analogues,7 which were predicted to maintain
Table 1
fXa inhibitory activities,15 anticoagulant potency16 and physical descriptors17 for compounds 1, 2 and 9–17, with P1 structures and abbreviations
S
Cl
R2
Cl
O
S
R1
N
1'
6ClNap
ClThE
N
N
R3
H
H
O
N
3S
O
X = H, 6-Ind
X= Cl, 3ClInd
8 or 9-17
X
Entry
R3
10 Stereo
R1
R2
FXa (Ki, nM)
1.5 ꢁ PT (
lM)
c log D7.4
MW
cmr
1
2
9
ClThE
6ClNap
ClThE
Ala-morpholine
R
R/S
R/S
S
4
2
4
3
2
9
1
4
1.2
7.0
1.4
4.7
3.9
7.4
0.9
2.6
6.5
2.8
13.3
2.7
1.6
2.0
2.2
2.2
2.2
1.9
0.9
1.2
0.1
0.9
447
476
466
484
484
484
473
452
470
440
456
10.9
12.5
12.3
13.2
13.2
13.2
12.6
11.8
12.7
12.6
12.3
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
H
Me
Me
Me
Me
Me
H
H
H
H
10
11
12
13
14
15
16
17
6ClNap
6ClNap
6ClNap
3ClInd
ClThE
6ClNap
3ClInd
6ClNap
R
R/S
R/S
R/S
R/S
R/S
5
0.3
5

1584
R. J. Young et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1582–1587
Table 2
series, which has been postulated to arise through the generation
of a reactive nitroso metabolite.9 With these modest structural
modifications, the desirable size (expressed as calculated molar
refraction, cmr) and hydrophobicity characteristics achieved with
2 would be maintained,5c while enhancing conformational re-
straint. Such properties were expected to facilitate the attainment
of good pharmacokinetic profiles and efficient translation of antic-
ipated fXa potency into good levels of anticoagulant activity. The
latter was anticipated due to expected modest levels of plasma
protein binding, with favourable binding kinetics,10 of relatively
hydrophilic compounds with only 2 or 3 aromatic rings.11
Targets in the aminoindane series were synthesised via homo-
chiral pyrrolidin-2-one ring construction (Scheme 1) using estab-
lished chemistry from commercial 5-amino indanone 4 and N-
Boc methionine.12 The amino functionality was introduced (as an
epimeric mixture) by reduction of the ketone 5 to the correspond-
ing alcohol, mesylation and displacement by either ammonia,
methylamine or dimethylamine. These intermediates 6 were N-
Rat DMPK parameters18 for 2 and aminoindanes 9, 11, 12, 14 and 15
a
b
c
Entry
Men
Stereo-chem
Clp
T1/2 (h)
Vss
Fd (%)
(mL/min/kg)
(L/kg)
2
9
11
12
14
15
1
2
2
2
1
1
R
R/S
S
R
R/S
R/S
1.6
9.5
6
7.6
2.2
2.5
18
2.4
1.1
2.0
1.7
0.9
2.4
>10
57
61
27
nd
25
1.7
4.4
3.4
5.4
12.4
a
b
c
Clp, plasma clearance expressed as mL/min/kg.
T1/2, half-life of the test compound expressed in h.
Vss, steady state volume of distribution expressed as L/kg.
F, oral bioavailability expressed as %.
d
protected as the trifluoroacetamide
7
in the amino (6,
R1 = R2 = H) or methylamino examples (6, R1 = H, R2 = Me). Acidic
deprotection of the 3-N-Boc group and sulfonylation furnished
the final N,N-dimethylamino (R1 = R2 = Me) analogues 8; trifluoro-
acetamide protection was removed using sodium carbonate in
aqueous methanol to furnish the other variants of 8. The indolyl
sulphonyl chlorides13 were N-triisopropylsilyl (TIPS) protected,
thus an additional fluoride deprotection step was required to fur-
nish these P1 analogues. To access homochiral compounds, the dia-
stereomers of dimethylamino intermediate 6 were separated by
preparative chiral HPLC,14 and then processed in the same manner
as the racemate.
The fXa potencies15 achieved in the aminoindane series were at
least comparable to the
aMBA series, exemplified by 2, which
translated into useful levels of anticoagulant activity16 as expected
from the predicted physical properties17 (Table 1). These data were
complemented by encouraging rat pharmacokinetics (Table 2),18
with trends and values similar to the
aMBA series; the N-mono-
methyl analogues again displayed extended half-lives in compari-
son to the N,N-dimethyl compounds. Particularly encouraging
activity data were secured with the 3-chloroindol-6-yl derivates
13 and 16. However, full evaluation of this aminoindane P4 motif
was curtailed by Cyp3A4 studies with 10, 15 and 17.19 Whilst
the levels of TDI observed were lower than those observed in the
Figure 2. Overlay of observed binding modes of 11 (green) and 12 (cyan) showing
contacts in the fXa S4 pocket: hydrogen bonds to Glu97 and proximity of Tyr99,
Phe174 and Trp215.
good physical properties and overlaid effectively with alternative
P1 structures in the sulfonamidopyrrolidin-2-one series4a and
other inhibitors exploiting indolyl P1 motifs.8 Importantly, the P4
targets offered potential alternative routes of metabolism (via,
for example, conjugation into indene, pyrrolene or dihydroquino-
aMBA series, they remained significant. As this series retained a
pendant amino group, it was postulated that metabolism of the
nitrogen via a likely nitroso species, as proposed for the aMBA ser-
ies,5c was the most relevant path, rather than elimination via an
indene.
line ring systems) which might address the TDI seen in the aMBA
F
F
H
N
b
a
N
Ar
BocHN
BocHN
Br
19
18
20
R2
d, e, f, g
R1
c, f
F
(for R1 = Me)
O
S
O
R1
(for R1 = Me)
F
O
N
HN
N
H2N
N
(h), f, i
j, k. l, g, (m)
(for R1 = Boc)
(for R1 = Boc)
21
22
R1 = H or Me
Scheme 2. Reagents and conditions: (a) nBuLi, pyrrolidinone-OTMS, THF, ꢀ78 °C to rt, 60%; (b) NaBH4, MeOH, rt, 90%; (c) HCO2H, H2C@O, CHCl3, reflux, 95%; (d) N-Boc
homoserine lactone, Me3Al, DCM, 0 °C to rt; (e) DtBAD, Bu3P, THF, rt, 49% (d and e); (f) HCl, MeOH, rt, 95%; (g) R2SO2Cl, pyridine, MeCN, rt, 75–95%; (h) optional chiral hplc; (i)
Boc2O, DCM, rt, 90%; (j) Cbz-Met-OH, HATU, DIPEA, DCM, rt, 95%; (k) MeI, MeCN, 60 °C, then Cs2CO3, 79%; (l) H2, Pd(OH)2–C, EtOH, rt, quant.; (m) optional for N-TIPS-protected
indole-P1; Et4NF, THF, rt, 90%.

R. J. Young et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1582–1587
1585
Table 3
fXa inhibitory activities,15 anticoagulant potency16 and physical descriptors17 for compounds 23–32
O
N
S
R2
N
R1
N
N
H
O
3S
O
F
H
F
2R shown
22 or 23-32
Entry
R2
Stereo
R1
FXa (Ki, nM)
1.5 ꢁ PT (lM)
c log D7.4
MW
cmr
23
24
25
26
27
28
29
30
31
32
ClThE
6ClNap
ClThE
6ClNap
3ClInd
6-Ind
ClThE
ClThE
6ClNap
6ClNap
R/S
R/S
R/S
R/S
R/S
R/S
R
S
R
S
Me
Me
H
H
H
H
H
H
H
15
6
5
2
2
100
3
6
2
4
1.6
8.6
3.2
8.0
2.9
9.9
0.7
1.8
4.5
7.7
0.9
1.1
1.1
1.4
1
0.1
1.1
1.1
1.4
1.4
484
502
470
488
477
442
470
470
488
488
12.3
13.2
11.9
12.8
12.2
11.7
11.9
11.9
12.8
12.8
H
Table 4
Rat DMPK parameters18 for phenylpyrrolidines
R2
Clp (ml/min/kg)
T1/2 (h)
Vss (L/kg)
Fd (%)
a
b
c
P1
Stereochem
23
25
27
26
31
32
ClThE
ClThE
3ClInd
6ClNap
6ClNap
6ClNap
R/S
R/S
R/S
R/S
R
Me
11
1.4
3.4
11
9.0
10.2
7.5
0.8
0.7
0.8
1.9
1.9
1.7
41
52
<1
24
25
38
H
H
H
H
H
2.8
1.3
2.7
2.6
2.8
S
a–d
As Table 2.
Effective binding, leading to maintenance of fXa activity was
(vide supra). Good oral bioavailability was achieved with (E)-2-(5-
chlorothien-2-yl)ethenyl (ClThE) 25 and 6-chloronaphth-2-yl
(6ClNap) 26 motifs, however the otherwise attractive profile of
the 3-chloroindol-6-yl 27 was undermined by the lack of oral
exposure. The R-ClThE 29 was one of the most potent anticoagu-
lants with the sulfonamidopyrrolidin-2-one template, whilst R-6-
Chloronaphth-2-yl 31 showed the potential for once daily dosing
based on the rat studies. No selectivity issues were apparent
against our panel of trypsin-like serine proteases25 and, more sig-
nificantly, the risk of time dependant inhibition of Cyp3A4 was
predicted by modelling for each of the 10-epimeric pair of indanes
11 and 12. This was substantiated by X-ray crystal structures of 11
and 12 bound into fXa; the whole indane rotates through 180° to
maintain the contact between the terminal amine and the back-
bone carbonyl of Glu97, with each of these maintaining stacking
and edge on hydrophobic contacts with the box resides Tyr99,
Phe174 and Trp215 (Fig. 2).20
The key bond forming reaction towards the 2-phenylpyrrolidine
series was construction of pyrroline 18 by reaction of lithiated N-
Boc-2-fluoro-4-bromoaniline 19 with TMS pyrrolidinone21
(Scheme 2). Following sodium borohydride reduction, Eschweil-
er–Clarke N-methylation of 20, followed by acidic deprotection
gave an intermediate (21, R1 = Me) for pyrrolidinone construction
using N-Boc-homoserine lactone and trimethylaluminium meth-
odology;22 this was progressed towards targets 22 as described
in Scheme 1. The activity of these compounds provided an impetus
to design a route to access the des-methyl analogues; moreover
preparative chiral hplc of pyrrolidine 20 facilitated access its
homochiral enantiomers in multi-gram quantities.23 Subsequently,
the position of N-Boc protection was switched by deprotection of
homochiral (or racemic) 20 under acidic conditions and selective
reaction of the more basic aliphatic nitrogen with Boc anhydride,
furnishing 21, R1 = Boc. These intermediates were converted into
homochiral des-methyl target compounds 22 exploiting construc-
tion of the pyrrolidinone from Cbz-methionine and elaboration as
described in Scheme 1.24
The fXa potencies achieved with N-methyl 23 and 24 translated
into useful anticoagulant activity (Table 3), which, combined with
encouraging oral exposure for 23 (Table 4) provided the impetus to
pursue the series further. In the des-methyl series, low nM potency
was achieved in molecules with similar anticoagulant activities
and longer pharmacokinetic half-lives, again driven by lower clear-
ance associated with this tertiary to secondary amine modification
Figure 3. Overlay of the X-ray crystal structures of 31 (green) and 32 (yellow)
showing contacts in the fXa S4 pocket: water mediated hydrogen bonding to Lys 96/
Glu97 with 31 and proximity of Tyr99, Phe174 and Trp215.

1586
R. J. Young et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1582–1587
effectively discharged.26 Taken together these data formed a pack-
age warranting further investigation of this series.
9. (a) Kalgutkar, A. S.; Gardner, I.; Obach, R. S.; Shaffer, C. L.; Callegari, E.; Henne,
K. R.; Mutlib, A. E.; Dalvie, D. K.; Lee, J. S.; Nakai, Y.; O’Donnell, J. P.; Boer, J.;
Harriman, S. P. Curr. Drug Metab. 2005, 6, 161; (b) Kalgutkar, A. S.; Obach, R. S.;
Maurer, T. S. Curr. Drug Metab. 2007, 8, 407.
10. We have previously reported on our understanding of the role of plasma
protein binding in the relationship between intrinsic factor Xa activity and
anticoagulant activity; the extent of binding gives only a partial explanation,
due to the multiple variables inherent in the process (see note 15 of Ref.5a
above).
11. (a) Ritchie, T. J.; MacDonald, S. J. F. Drug Discovery Today 2009, 14, 1011; The
recognition of an additive contribution of hydrophobicity plus number of
aromatic rings in impacting solubility was noted in: (b) Hill, A. P.; Young, R. J.
Drug Discovery Today 2010, 15, 648–655; This observation also holds for %HSA
binding, to be published shortly in Young, R. J.; Green, D. V. S.; Luscombe, C. N.;
Hill, A. P. Drug Discovery Today, 2011 invited submission, manuscript under
review.
X-ray structures27 of 31 and 32 bound into fXa showed the dia-
stereomeric sub-units in the P4 pocket (Fig. 3). The S1 pocket is
occupied by the aromatic sulfonamide group as seen throughout
previously reported series.4,5 The distal pyrrolidine ring of S,R-31
makes hydrophobic contacts with the Trp215 residue though a
3.5 Å CH2ꢂ ꢂ ꢂ
p
interaction.28 The likely protonated ring nitrogen
makes water-mediated hydrogen bonds to the backbone carbonyls
of Lys96 and Glu97. In epimeric S,S-32 the confirmation of the pyr-
rolidine ring required to allow orientation of the ring nitrogen
atom towards Glu97 meant that it did not reach quite as far into
the hydrophobic cleft of the pocket, whilst maintaining contacts
with Trp215 through the equivalent methylene in a similar posi-
tion and without the water mediated hydrogen bond from the
nitrogen.
12. By analogy to our previous publications4a,5 these transformations were
expected to have produced homochirality at N-3; findings of chiral hplc
studies14 of final compounds were consistent with this notion.
13. Harling, J. D.; Watson, N. S.; Young, R. J. WO108709 A1, 2006.
14. An analytical separation was achieved on a Chiracel OJ column, eluted with
15% EtOH in heptanes; the isomers had retention times of 8.7 min and
11.2 min for S,S and S,R, respectively. This was scaled to a preparative column
to deliver compounds 11 and 12 with ee and de of >96%.
In conclusion, two approaches have been described to expand
upon the SAR of fXa inhibitors incorporating the
amine ( MBA) P4 motif, which were designed to constrain the
molecules and to address the risk of Cyp3A4 TDI in the MBA ser-
a-methyl benzyl-
a
15. Factor Xa inhibitory activities were determined using Rhodamine 110, bis-
(CBZ-glycylglycyl)-L-arginine amide as the fluorogenic substrate; details are
a
described in Ref.4a
.
ies. These studies have led to the discovery of potent and long-last-
ing oral inhibitors of fXa incorporating aminoindane and
phenylpyrrolidine P4 groups. Importantly, the latter series effec-
tively addressed the key issue of TDI and provided candidate qual-
ity molecules; these have been progressed into further studies that
will be reported in due course.
16. Anticoagulant activities were determined in the prothrombin time (PT) assay;
see Ref.4a, expressed as the concentration required to extend the control
coagulation time by 50% (1.5 ꢁ PT).
17. Hydrophobicity predictions, expressed as c log D7.4, were all re-calculated
using Advanced Chemistry Development software v11.0 to ensure consistency
in this paper; calculated molecular refractivity (cmr) was derived from
Daylight software v4.9. The paradoxical advantage of these c log D7.4
predictions over actual octanol–water measurements is noted in Ref.11b; we
have quoted these through all of our publications for consistency.
18. Pharmacokinetics measured in male Sprague–Dawley rats following
intravenous and oral administration. The formulation used for both iv and po
dosing was a 5:95% (v/v) mixture of DMSO and 50:50 PEG-200: sterile water.
Serial blood samples were collected into heparinised containers at various
time-points and blood centrifuged to yield plasma. These studies used at least
three animals for each (iv/po) leg.
19. Amino indene TDI experiments were carried out at Biodynamics
(www.biodynamics.co.uk) and were reported as fold shifts in the observed
Cyp3A4 (BFC substrate) IC50 after 25 min incubation versus the initial IC50
value. Apparent TDI was detected for compounds 10, 15 and 17 with 2.2, 2.5
and 2.1-fold shifts (respectively) in the IC50; compared with 5.8-fold for 2 in a
parallel assay.
References and notes
1. (a) Eikelboom, J. E.; Weitz, J. I. Thromb. Haemost. 2009, 101, 2; (b) Sanford, M.;
Plosker, G. L. Drugs 2008, 68, 1699; (c) Sobera, L. A.; Bozzo, J.; Castaner, J. Drugs
Future 2005, 30, 877.
2. (a) Abrams, P. J.; Emerson, C. R. Pharmacotherapy 2009, 29, 167; (b) Escolar, G.;
Villalta, J.; Casals, F.; Bozzo, J.; Serradell, N.; Bolos, J. Drugs Future 2006, 31, 484;
(c) Perzborn, E.; Roehrig, S.; Straub, A.; Kubitza, D.; Misselwitz, F. Nat. Rev. Drug
Disc. 2011, 10, 61.
3. (a) Sobieraj-Teague, M.; O’Donnell, M.; Eikelboom, J. Semin. Thromb. Hemost.
2009, 35, 515; (b) Harenberg, J.; Wehling, M. Semin. Thromb. Hemost. 2008, 34,
39; (c) Lassen, M. R. Expert Opin. Pharmacother. 2009, 10, 1769; (d) Gomez-
Outes, A.; Lecumberri, R.; Pozo, C.; Rocha, E. Curr. Vasc. Pharmacol. 2009, 7, 309;
(e) Eriksson, B. I.; Quinlan, D. J.; Weitz, J. I. Clin. Pharmacokinet. 2009, 48, 1; (f)
Pinto, D. J. P.; Smallheer, J. M.; Cheney, D. L.; Knabb, R. M.; Wexler, R. R. J. Med.
Chem. 2010, 53, 6243.
4. (a) Chan, C.; Borthwick, A. D.; Brown, D.; Burns-Kurtis, C. L.; Campbell, M.;
Chaudry, L.; Chung, C.-W.; Convery, M. A.; Hamblin, J. N.; Johnstone, L.; Kelly, H.
A.; Kleanthous, S.; Patikis, A.; Patel, C.; Pateman, A. J.; Senger, S.; Shah, G. P.;
Toomey, J. R.; Watson, N. S.; Weston, H. E.; Whitworth, C.; Young, R. J.; Zhou, P.
J. Med. Chem. 2007, 50, 1546–1557; (b) Abboud, M. A.; Needle, S. J.; Burns-
Kurtis, C. L.; Valocik, R. E.; Koster, P. F.; Amour, A. J.; Chan, C.; Brown, D.;
Chaudry, L.; Zhou, P.; Patikis, A.; Patel, C.; Pateman, A. J.; Young, R. J.; Watson,
N. S.; Toomey, J. R. J. Cardiovasc. Pharmacol. 2008, 52, 66.
5. (a) Young, R. J.; Borthwick, A. D.; Brown, D.; Burns-Kurtis, C. L.; Campbell, M.;
Chan, C.; Charbaut, M.; Chung, C.-w.; Convery, M. A.; Kelly, H. A.; King, N. P.;
Kleanthous, S.; Mason, A. M.; Pateman, A. J.; Patikis, A. N.; Pinto, I. L.; Pollard, D.
R.; Senger, S.; Shah, G. P.; Toomey, J. R.; Watson, N. S.; Weston, H. E. Bioorg. Med.
Chem. Lett. 2008, 18, 23; (b) Young, R. J.; Borthwick, A. D.; Brown, D.; Burns-
Kurtis, C. L.; Campbell, M.; Chan, C.; Charbaut, M.; Convery, M. A.; Diallo, H.;
Hortense, E.; Irving, W. R.; Kelly, H. A.; King, N. P.; Kleanthous, S.; Mason, A. M.;
Pateman, A. J.; Patikis, A. N.; Pinto, I. L.; Pollard, D. R.; Senger, S.; Shah, G. P.;
Toomey, J. R.; Watson, N. S.; Weston, H. E.; Zhou, P. Bioorg. Med. Chem. Lett.
2008, 18, 28; (c) Kleanthous, S.; Borthwick, A. D.; Brown, D.; Burns-Kurtis, C. L.;
Campbell, M.; Chan, C.; Chaudry, L.; Clarte, M. O.; Convery, M. A.; Harling, J. D.;
Hortense, E.; Irving, W. R.; Irvine, S.; Pateman, A. J.; Patikis, A. N.; Pinto, I. L.;
Pollard, D. R.; Roethka, T. J.; Senger, S.; Shah, G. P.; Stelman, G. J.; Toomey, J. R.;
Watson, N. S.; West, R. I.; Whittaker, C.; Zhou, P.; Young, R. J. Bioorg. Med. Chem.
Lett. 2010, 20, 618.
20. The structure for 11 was refined at 1.85 Å (overall Rmerge is 0.16) in Refmac5 to
a final Rfactor of 0.196 and Rfree of 0.246 and 12 was refined at 1.9 Å (overall
Rmerge is 0.061) in Refmac5 to a final Rfactor of 0.189 and Rfree of 0.241 using
procedures described in Ref.4a
. Co-ordinates are deposited in the protein
databank with codes 2y7z 11 and 2y80 12. These structures and many others
unpublished from our factor Xa series will be made available via the
Community Structure–Activity Resource at the University of Michigan
(www.csardock.org).
21. Feringa, B. L.; Jansen, J. F. G. A. Tetrahedron Lett. 1986, 27, 507.
22. Bell, I. M.; Beshore, D. C.; Gallicchio, S. N.; Williams, T. M. Tetrahedron Lett.
2000, 41, 1141.
23. An analytical separation was achieved on a Chiralpak AD column, eluted with
15% EtOH in heptanes with 0.1% TFA; the isomers had retention times of
4.8 min and 6.7 min for S and R, respectively. This was scaled to a preparative
column to deliver both enantiomers of 20 with ee of >96% in multi-gram
quantities.
24. 19F NMR of Mosher amides of the intermediate amines between 21 and 22
indicated that these transformations gave homochirality at this centre. The
diastereomeric excess of final compounds 31 and 32 was shown to be >96% by
chiral hplc on a Chiralpak AD column, eluted with 50% EtOH in heptanes with
0.1% TFA; the isomers had retention times of 4.6 min and 8.6 min S,R-31 and
S,S-32, respectively.
25. Selectivity data versus a panel of trypsin-like serine proteases, expressed as
fold selectivity by ratio of Ki values.
fIIa
tPA
Kallikrein
APC
Plasmin
Trypsin
29
31
500
3200
1000
160
320
80
10,000
10,000
8000
2000
>10,000
>10,000
6. QXP/FLO, ThistleSoft Inc.: Colebrook, C. T.; McMartin, C.; Bohacek, R. S. J.
Comput. Aided Mol. Des. 1997, 11, 333.
7. Shi, Y.; Sitkoff, D.; Zhang, J.; Klei, H. E.; Kish, K.; Liu, E. C.-K.; Hartl, K. S.; Seiler, S.
M.; Chang, M.; Huang, C.; Youssef, S.; Steinbacher, T. E.; Schumacher, W. A.;
Grazier, N.; Pudzianowski, A.; Apedo, A.; Discenza, L.; Yanchunas, J.; Stein, P. D.;
Atwal, K. S. J. Med. Chem. 2008, 51, 7541.
8. Franciskovicha, J. B.; Masters, J. J.; Weber, W. W.; Klimkowski, V. J.; Chouinard,
M.; Sipes, P. R.; Johnson, L. M.; Snyder, D. W.; Chastain, M. K.; Craft, T. J.;
Towner, R. D.; Gifford-Moore, D. S.; Froelich, L. L.; Smallwood, J. K.; Foster, R. S.;
Smith, G. F.; Liebeschuetz, J. W.; Murray, C. W.; Young, S. C. Bioorg. Med. Chem.
Lett. 2007, 17, 6910.
fIIa = thrombin; tPA = tissue plasminogen activator; APC = activated protein C.
26. Phenylpyrrolidine TDI assays were carried out in house and expressed as fold
decrease in enzyme activity after 30 min incubation. 31 did not demonstrate
time-dependent inhibition of CYP1A2, CYP2C9, CYP2C19 or CYP2D6, or of
CYP3A4 in the DEF assay, over
However, it exhibited apparent time-dependent inhibition of CYP3A4 activity
in the 7BQ assay with a 1.7-fold decrease at 33
M (1.4ꢁ the IC50) and 2.5-fold
a concentration range of 0.33–100 lM.
l

R. J. Young et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1582–1587
1587
at 100
l
M (4ꢁ the IC50).The observed decreases in activity should be compared
Rmerge is 0.122) in Refmac5 to a final Rfactor of 0.178 and Rfree of 0.233 using
procedures described in Ref. 4a. Coordinates are deposited in the protein data
bank with codes 2y81 31 and 2y82 32.
with the 17-fold shift for 2 at 10
l
M, its IC50, in this assay.
27. The structure for 31 was refined at 1.69 Å (overall Rmerge is 0.083) in Refmac5 to
a final Rfactor of 0.186 and Rfree of 0.225 and 32 was refined at 2.19 Å (overall
28. Bissantz, C.; Kuhn, B.; Stahl, M. J. Med. Chem. 2010, 53, 506.