Discovery of N-{N-[(3-cyanobenzene) sulfonyl]-4(R)-(3,3-difluoropiperidin-1-yl)-(l)-prolyl}-4-[(3′,5′-dichloro-isonicotinoyl) amino]-(l)-phenylalanine (MK-0617), a highly potent and orally active VLA-4 antagonist

Article abstract of DOI:10.1016/j.bmcl.2009.07.111

A series of prolyl-N-isonicotinoyl-(L)-4-aminophenylalanine derivatives substituted at the proline 4-position with cyclic amines was evaluated as VLA-4 antagonists. The ring size and presence or absence of fluorine affected potency and receptor occupancy.

Full text of DOI:10.1016/j.bmcl.2009.07.111

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5803–5806
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of N-{N-[(3-cyanobenzene) sulfonyl]-4(R)-(3,3-difluoropiperidin-1-yl)-
(L)-prolyl}-4-[(30,50-dichloro-isonicotinoyl) amino]-(L)-phenylalanine (MK-0617),
a highly potent and orally active VLA-4 antagonist
a,
Shankar Venkatraman a, , Alec D. Lebsack , Kenneth Alves c, Michael F. Gardner b,à, Joyce James b,à
,
*
Russell B. Lingham c, Salony Maniar c, Richard A. Mumford c, Qian Si c, Nicholas Stock a,§, Kelly M. Treonze c,
Bowei Wang a,§, Jasmine Zunic a,§, Benito Munoz a,–
a Department of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
b Department of Drug Metabolism, Merck Research Laboratories, San Diego, CA 92121, USA
c Department of Inflammation and Rheumatology Research, Merck Research Laboratories, Rahway, NJ 07065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of prolyl-N-isonicotinoyl-(L)-4-aminophenylalanine derivatives substituted at the proline 4-posi-
tion with cyclic amines was evaluated as VLA-4 antagonists. The ring size and presence or absence of
fluorine affected potency and receptor occupancy. The analog with 3,3-difluoropiperidine at the proline
4-position (13) was the most potent compound and had very good duration of receptor occupancy
in vitro. The ethyl ester prodrug of 13 demonstrated excellent receptor occupancy after oral dosing in
rats.
Received 9 April 2009
Revised 14 July 2009
Accepted 15 July 2009
Available online 28 July 2009
Keywords:
VLA-4
Ó 2009 Elsevier Ltd. All rights reserved.
Antagonist
Multiple scelerosis
The integrin very late antigen-4 (VLA-4,
a
4b1) belongs to the
pancy = 64% bound @ 3 h, Fig. 1).5 The lead compound 1 did not
achieve desirable levels of receptor occupancy in vivo. Herein, we
report the synthesis and SAR of analogs of 1 substituted with cyclic
tertiary amines at the proline 4-position as a means to obtain im-
proved potency and in vivo properties.
heterodimeric integrin receptor superfamily.1 VLA-4 is expressed
on the surface of most leukocytes and binds to vascular cell adhe-
sion molecule-1 (VCAM-1) and the CS-1 domain of fibronectin.
VCAM-1 is expressed on activated endothelial cells in response to
inflammatory cytokines. Adhesion interactions between VLA-4 and
VCAM-1 may be important for activation, migration, proliferation,
and differentiation of leukocytes during normal and pathophysio-
logical processes.2 Tysabri, a humanized monoclonal antibody
against VLA-4, is approved for patients suffering from multiple
sclerosis and Crohn’s disease.3 This has generated significant inter-
est in the development of small molecule antagonists of VLA-4 for
the treatment of diseases where cell-trafficking and activation are
important, such as asthma, rheumatoid arthritis and multiple
sclerosis.4
Inhibition potency was measured using a competition binding
assay with a 35S labeled ligand and VLA-4 in its non-activated state
HN
O
H
N
OH
N
O
SO2
In our laboratory, earlier work on VLA-4 antagonists identified
potent prolyl-dipeptides containing secondary amines with slow
O
Cl
N
H
relative off-rates such as
1 (IC50 = 0.05 nM; receptor occu-
N
CN
Cl
1
* Corresponding author. Tel.: +1 2153532900.
E-mail address: p3ptide@yahoo.com (S. Venkatraman).
Present address: J& J San Diego, CA, United States.
α4β1 (Ca2+, Mg2+) IC50 = 0.05 nM
Receptor occupancy (Ca2+, Mg2+ ) =
64%bound @ 3 h
à
Present address: Ambit Biosciences, San Diego, CA, United States.
Present address: Amira Pharmaceuticals, San Diego, CA, United States.
§
–
Present address: Broad Institute, Cambridge, MA, United States.
Figure 1. Lead compound for the program.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.111

5804
S. Venkatraman et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5803–5806
Table 1
bound after three hours were utilized to further distinguish these
compounds.7–9 As shown in Table 1, a variety of fluorinated cyclic
tertiary amine-dipeptide derivatives were evaluated using these
assays in tandem.
Binding to VLA-4 by 4-substituted prolyl-N-isonicotinoyl-(L)-4-aminophenylalanine
derivatives
R
The structure–activity trends in Table 1 suggest that both ring
size and presence or absence of fluorine at various positions are
important in optimizing potency and receptor occupancy. Increas-
ing the ring size from 4 through 6 provided greater receptor occu-
pancy; optimal results were obtained with piperdines 10 and 13. In
general, 3,3-difluoroazacycles 5, 9, and 13 were more potent
(IC50 = 0.03–0.10 nM) and afforded higher receptor occupancy
(70–90% bound @ 3 h) than the corresponding des-fluorine ana-
logues 3, 6, and 10. Additionally, most compounds presented in Ta-
ble 1 showed higher levels of receptor occupancy (>58% bound @
3 h) when compared with the initial cyclobutyl amine lead 1. The
3,3-difluoropiperdine 13 afforded the best combination of binding
affinity (IC50 = 0.03 nM) and receptor occupancy (87% bound @
3 h).
Interestingly, 4,4-difluoropiperdine 14 was sevenfold less
potent and exhibited a 2.5-fold lower receptor occupancy than
3,3-difluoropiperdine 13, suggesting that location of fluorine
groups is critical. Only small differences in potency and receptor
occupancy could be observed with enantiomers of mono-fluori-
nated compounds such as 7 and 8. Differences in the pKa values
of fluorinated verses des-fluorinated cyclic tertiary amines did
not explain the differences in binding affinities and receptor occu-
pancy. For example, 10 and 13 exhibit nearly identical binding
affinities and receptor occupancies, while their measured pKa val-
ues for the amine were 8.4 and 4.6, respectively.10 Introduction
of 2,2,2-trifluoroethyl in the lead compound provided 2 with
improved potency, but 2 exhibited a significant shift to lower
potency in the human whole blood assay, (IC50 = 3.1 nM) compared
to cyclic amines such as 13 which exhibited little shift in the same
assay (IC50 = 0.07 nM).
Since monoclonal antibodies have the ability to recognize both
the activated and non-activated states of VLA-4,2 we explored the
binding and receptor occupancy for 3,3-difluoropiperdine 13 using
the activated state of VLA-4 (Mn2+, Fig. 2). The potency of 3,3-diflu-
oropiperdine 13 with the activated form of VLA-4 afforded an IC50
value of 0.02 nM, similar to that of the resting state
(IC50 = 0.03 nM). In addition, 13 exhibited similar receptor occu-
pancies at 3 h using the activated and non-activated forms of
VLA-4. Interestingly, when the receptor occupancies at 3 h were
compared for 13 and the initial lead 1, we could only differentiate
the two compounds using the non-activated state of VLA-4. The
results demonstrated utility of the receptor occupancy assay to
better distinguish between compounds in the non-activated state
and highlighted the improved receptor occupancy of 3,3-difluoro-
piperdine 13 compared to 1.
O
H
N
OH
N
O
SO2
O
Cl
N
H
N
CN
Cl
Compdsa
R
a
4b1 Ca2+
Mg2+ IC50 (nM) (37 °C) @ 3 h (% bound @ 3 h)
,
In vitro relative off-rate
H
N
1
0.15 0.01
0.02 0.02
0.15 0.03
0.27 0.06
0.90 0.12
0.22 0.07
0.12 0.03
64 10
80 12
F3C
2
N
N
3
56
6
4
60 6
65
F
N
N
F
F
5
6
65
N
F
F
(R)-7
80
N
N
(S)-8
9
0.24 0.07
0.10 0.03
0.11 0.03
0.21 0.06
0.35 0.13
0.03 0.01
0.22 0.05
75
F
F
81 12
N
10
80
70
55
90
37
8
N
F
F
11b
12
N
N
100
95%
95%
F
F
90
87%
13
6
4
N
80
1 off-rate (Mn2+
13 off-rate (Mn2+
13 off-rate (Ca2+, Mg2+
1 off-rate (Ca2+, Mg2+
)
70
60
50
F
F
)
14
N
)
58%
a
)
All data is an average of three independent runs or n = 1.
b
Mixture of diastereomers.
0
30 60 90 120 150 180
Time (min)
(Ca2+, Mg2+ buffer) in Jurkat cells.6 In addition, competitive binding
assay in whole blood and dissociation kinetics expressed as %
Figure 2. Receptor occupancy versus time curves for 1 and 13 under activating
(Mn2+) and non-activating (Ca2+, Mg2+) conditions.

S. Venkatraman et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5803–5806
5805
SO2Cl
100
R
100
75
50
25
0
HO
OR2
16
N
CN
a,b
OMe
N
O
SO2
10
1
H
O
60-80%
15
CN
17; R2 = Me
18; R2 =H
O
c
13 Receptor Occupancy
13 Portal Plasma Levels
H2N
OEt
O
Cl
24
0
3
6
9
12
Time (h)
N
H
N
Cl
Figure 3. Portal plasma levels and VLA-4 receptor occupancy in rats following oral
dosing of 13 at 5 mpk.
19
d,e
60-70%
R
As a prelude to evaluating 3,3-difluoropiperdine 13 in animal
models of autoimmune and allergic diseases, the receptor occu-
pancy and pharmacokinetics were measured (Fig. 3). While the
systemic pharmacokinetics were poor for 13 (AUC and Cmax could
not be determined; Clp >100 mL/min/kg; F <5%), portal plasma lev-
els provided an AUC of 50 nM h and a Cmax of 27 nM. Although the
drug levels were low, the Cmax was 75-fold over the average con-
centration of VLA-4 receptors in the rat (300 pM). Furthermore,
in the receptor occupancy assay, 13 achieved 70% occupancy at
12 h and 40% occupancy at 24 h (Fig. 2). The initial exposure of
13 to circulating leukocytes bearing VLA-4 in the portal vein and
slow off-rate likely compensate for the poor systemic pharmacoki-
netics to provide receptor occupancy coverage even after the drug
levels can no longer be detected. The correlation between receptor
occupancy and efficacy in animal models and human is well
documented.8
Furthermore, when 13 was dosed as the ethyl ester prodrug at
1.5 mpk in rats, it yielded receptor occupancy values of 88% and
60% at 12 h and 24 h, respectively. This is in contrast to the parent
acid which, when dosed at 5 mpk orally in rats, gave receptor occu-
pancy values of 65% and 22% at 12 h and 24 h, respectively. The
ethyl ester prodrug of 13 also had higher receptor occupancy com-
pared to the ethyl ester prodrug of 1 at 1.5 mpk over 24 h as shown
in Figure 4.
O
H
N
OH
N
O
SO2
O
Cl
N
H
N
CN
Cl
2-14
Scheme 1. Reagents and conditions: (a) 15, Et3N, CH2Cl2; (b) TfO2, iPrEtN, CH2Cl2,
ꢀ60 °C, 2 h then R, iPrEtN ꢀ20 °C to rt, 12 h; (c) LiOH, CH3CN/H2O; (d) 18, EDC,
HOBT, Et3 N, DMF; (e) LiOH, CH3CN/H2O.
provided a triflate. This was displaced with a variety of commercially
available cyclic tertiary amines (R) to provide 17 in one pot.9 Hydro-
lysis of the methyl ester 17 by reaction with lithium hydroxide affor-
ded the corresponding acids 18. Standard peptide coupling of these
acids 18 with N-isonicotinoyl-(L)-4-aminophenylalanine (19) using
1-(3-dimethylamino-propyl)-3-ethylcarbodiimide hydrochloride
(EDC) and 1-hydroxybenzotriazole hydrate (HOBT) followed by
hydrolysis of the ethyl ester in the presence of lithium hydroxide
delivered the final acids 2–14.11
In summary, we have developed a novel series of potent prolyl
dipeptide VLA-4 antagonists containing fluorinated cyclic tertiary
amines at the proline 4-position. In general, the fluorinated cyclic
tertiary amines provided improved potency and receptor occu-
pancy levels when compared with their des-fluoro analogues and
the initial cyclobutyl amine lead compound. An optimal compound
from this work is 3,3-difluoropiperdine 13 (MK-0617) with an IC50
value of 0.03 nM and a receptor occupancy on both activated and
non-activated VLA-4 receptors of ꢁ90% at 3 h. Dosed orally as its
ethyl ester prodrug to rats, 13 had low systemic exposure but
excellent receptor occupancy at 24 h due to its slow off-rate.
The fluorinated and non-fluorinated cyclic tertiary amine-dipep-
tide derivatives were prepared as outlined in Scheme 1. The synthe-
sis began with commercially available cis-4-hydroxy-L-proline
methyl ester (15) which was sulfonylated with 3-cyano-sulfonyl-
chloride (16). Exposure of the resulting sulfonamide to triflic anhy-
dride (Tf2O) in the presence of diisopropylethyl amine at ꢀ60 °C
100
88%
Acknowledgments
75
60%
77%
We acknowledge the contributions of many members of the la-
beled synthesis group and the Laboratory of Animal Resources of
Merck Research Laboratories.
50
25%
25
References and notes
Ethyl ester prodrug of 1
0
Ethyl ester prodrug of 13
1. Hemler, M. E.; Elices, M. J.; Parker, C.; Takada, Y. Immunol. Rev. 1990, 114, 45.
2. Elices, M. J. Curr. Opin. Anti-infamm. Immunol. 1999, 1, 15.
3. (a) Kent, S. J.; Karlik, S. J.; Cannon, C.; Hines, D. K.; Yednock, T. A.; Fritz, L. C.;
Horner, H. C. J. Neuroimmunol. 1995, 58, 1; (b) Honey, K. J. Clin. Invest. 2008, 118,
825.
3
6
9
12
24
Time (h)
4. For Asthma see: (a) Sielszak, M. W.; Ahmed, A.; Cortes, A.; Lauredo, I. T.;
Kim, J.; Pepinsky, B.; Beenjamin, C. D.; Leone, D. R.; Lobb, R. R.; Weller,
Figure 4. Rat Receptor occupancy of 1 and 13 following oral dosing of ester
prodrugs at 1.5 mpk.

5806
S. Venkatraman et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5803–5806
P. F. J. Clin. Invest. 1994, 93, 776; (b) Milne, A. A. Y.; Piper, P. J. Eur. J.
Pharmacol. 1995, 282, 243; For Rheumatoid Arthritis see: (c) Seiffge, D. J.
the radioactivity associated with the cells was measured in a Packard TopCount
NXTTM.
Rheumatol. 1996, 23, 2086; For multiple sclerosis see: (d) Yednock, T. A.;
Cannon, D.; Fritz, L. C. Nature 1992, 356, 63; (e) Keszthelyi, E.; Karlik, S.;
Hyduk, S.; Rice, A.; Gordon, G.; Yednock, T.; Horner, H. Neurology 1996,
47, 1053.
7. Hagmann, W. K. J. Med. Chem. 2008, 51, 4359.
8. Sandborn, W. J.; Colombel, J. F.; Enns, R.; Feagan, B. G.; Hanauer, S. B.; Lawrance,
I. C.; Panaccione, R.; Sanders, M.; Schreiber, S.; Targan, S.; Deventer, S. V.;
Goldblum, R.; Despain, D.; Hogge, G. S.; Rutgeerts, R. NEJM 2005, 353, 1912.
9. (a) Doherty, G. A.; Yang, G. X.; Borges, E.; Tong, S.; McCauley, E. D.; Treonze, K.
M.; Riper, G. V.; Pacholok, S.; Si, Q.; Koo, G. C.; Shah, K.; Mumford, R. A.;
Hagmann, W. K. Bioorg. Med. Chem. Lett. 2003, 13, 1891; (b) See Ref. 6a.; (c)
Leone, D. R.; Giza, K.; Gill, A.; Dolinski, B. M.; Yang, W.; Perper, S.; Scott, D. M.;
Lee, W.-C.; Cornebise, M.; Wortham, K.; Nickerson-Nutter, C.; Chen, L. L.;
LePage, D.; Spell, J. C.; Whalley, E. T.; Petter, R. C.; Adams, S. P.; Lobb, R. R.;
Pepinsky, R. B. J. Pharmacol. Exp. Ther. 2003, 305, 1150.
5. Lin, L. S.; Lanza, T.; Jewell, J. P.; Liu, P.; Jones, C.; Kieczykowski, G. R.; Treonze,
K.; Si, Q.; Manior, S.; Koo, G.; Tong, X.; Wang, J.; Schuelke, A.; Pivnichny, J.;
Wang, R.; Raab, C.; Vincent, S.; Davies, P.; MacCoss, M.; Mumford, R. A.;
Hagmann, W. K. J. Med. Chem. 2009, 52, 3449.
6. (a) Hagmann, W. Curr. Top. Med. Chem. 2004, 4, 1461; (b) Hagmann, William K.;
Lin, L. S.; Liu, P.; Mumford, R. A.; Reger, T. S.; Smith, N. D.; Stock, N. S.; Zunic, J.
PCT Int. Appl. WO 2005087760 A1, 2008.; (c) For kinetic off rate measurements,
Jurkat cells were first treated with saturating levels of compound at 37 °C. After
one hour incubation, the cells were collected by filtration and washed once
with ice-cold binding buffer. Subsequently, cells were re-suspended in the
binding buffer containing an excess of [125I]-tracer, and further incubated for
increasing periods of time. The cells were collected at different time points, and
10. HPLC log D also did not account for differences in binding affinity (all had
similar values of 2 at pH 7.3).
11. All the final compounds were greater than 95% ee as measured by chiral HPLC
using analytical column (CHIRALCEL OD 10
isoproponol/triethylamine as mobile system.
l, 4.6 ꢂ 250 mm) with heptane and