Discovery of selective urokinase plasminogen activator (uPA) inhibitors as a potential treatment for multiple sclerosis

Article abstract of DOI:10.1016/j.bmcl.2018.09.001

We report here the design and synthesis of a novel series of benzylamines that are potent and selective inhibitors of uPA with promising oral availability in rat. Further evaluation of one representative (ZK824859) of the new structural class showed that

Full text of DOI:10.1016/j.bmcl.2018.09.001

Accepted Manuscript
Discovery of Selective Urokinase Plasminogen Activator (uPA) Inhibitors as a
Potential Treatment for Multiple Sclerosis
Imadul Islam, Shendong Yuan, Christopher W. West, Marc Adler, Ulrich Bothe,
Judi Bryant, Zheng Chang, Kieu Chu, Kumar Emayan, Giovanna Gualtieri,
Elena Ho, David Light, Cornell Mallari, John Morser, Gary Phillips, Caralee
Schaefer, Drew Sukovich, Marc Whitlow, Deborah Chen, Brad O. Buckman
PII:
S0960-894X(18)30732-7
https://doi.org/10.1016/j.bmcl.2018.09.001
BMCL 26022
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
25 June 2018
29 August 2018
1 September 2018
Please cite this article as: Islam, I., Yuan, S., West, C.W., Adler, M., Bothe, U., Bryant, J., Chang, Z., Chu, K.,
Emayan, K., Gualtieri, G., Ho, E., Light, D., Mallari, C., Morser, J., Phillips, G., Schaefer, C., Sukovich, D., Whitlow,
M., Chen, D., Buckman, B.O., Discovery of Selective Urokinase Plasminogen Activator (uPA) Inhibitors as a
Potential Treatment for Multiple Sclerosis, Bioorganic & Medicinal Chemistry Letters (2018), doi: https://doi.org/
10.1016/j.bmcl.2018.09.001
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.

Discovery of Selective Urokinase Plasminogen Activator (uPA) Inhibitors as a
Potential Treatment for Multiple Sclerosis
Imadul Islam,*a,b Shendong Yuan,b Christopher W. West,b Marc Adler,c Ulrich
Bothe,b,h Judi Bryant,b Zheng Chang,b Kieu Chu,d Kumar Emayan,b Giovanna
Gualtieri,b Elena Ho,e David Light,f Cornell Mallari,e John Morser,g Gary Phillips,b
Caralee Schaefer,e Drew Sukovich,d Marc Whitlow,c Deborah Chenb and Brad O.
Buckmanb
aMedical Core Facility and Research Platforms, King Abdullah International Medical Research Center/King
Saud bin
Abdulaziz University for Health Sciences, Ministry of National Guard Health Affairs, Riyadh 11426, Saudi
Arabia,
bDepartment of Medicinal Chemistry, Berlex Biosciences, 2600 Hilltop Drive, Richmond, CA 94804
cDepartment of Biophysics, Berlex Biosciences, 2600 Hilltop Drive, Richmond, CA 94804
dDepartment of Molecular Pharmacology, Berlex Biosciences, 2600 Hilltop Drive, Richmond, CA 94804
eDepartment of Animal Pharmacology, Berlex Biosciences, 2600 Hilltop Drive, Richmond, CA 94804
fDepartment of Antibody Technology, Berlex Biosciences, 2600 Hilltop Drive, Richmond, CA 94804
gCardiovascular Department, Berlex Biosciences, 2600 Hilltop Drive, Richmond, CA 94804
hResearch and Development Pharmaceutical, Bayer AG, 13342, Berlin, Germany
Keywords: uPA; Urokinase; serine protease; multiple sclerosis; uPA inhibitor
*Corresponding author. Imadul Islam, EMail ID: imadulislam@yahoo.com
Abstract: We report here the design and synthesis of a novel series of benzylamines that are potent and selective inhibitors of
uPA with promising oral availability in rat. Further evaluation of one representative (ZK824859) of the new structural class
showed that this compound lowered clinical scores when dosed in either acute or chronic mouse EAE models, suggesting that
uPA inhibitors of this type could be useful for the treatment of multiple sclerosis.
Urokinase plasminogen activator (uPA) is a serine protease belonging to the trypsin family that is involved in
various biological processes.1 As a key initiator in extracellular proteolytic cascades, uPA is part of the
fibrinolytic system, which includes tissue plasminogen activator (tPA), uPa receptor (uPAR), plasminogen
amongst others. The proteolytic activity of the fibrinolytic system is highly regulated through the secretion of
inactive precursors that are subsequently activated by initiating enzymes and by the presence of endogenous
inhibitors such as plasminogen activator inhibitor-1 (PAI-1). An imbalance between the regulation of
fibrinolysis and excess protease activity has been associated with several diseases characterized by low
fibrinolytic activity leading to thrombosis or high fibrinolytic activity resulting in abdominal aortic aneurysm
and cancer.
uPA inhibitors could especially be a effective therapeutic agents for the treatment of multiple sclerosis (MS).
Immunohistochemical analysis has shown that uPA level is significantly increased in MS lesions.2 On the
other hand, studies investigating the consequences of deficiencies in other components of the fibrinolytic
system have not demonstrated improvements in outcomes in EAE. For example, genetic deficiency in tPA3, 4

resulted in more severe disease in the EAE model than in wild type animals. While EAE differs from human
MS in several ways, it is widely studied as an animal model of human CNS demyelinating diseases such as
MS. Overall, targeting uPA by an inhibitor may be a unique method for the treatment of MS.
Many small molecule inhibitors of uPA have been described but most of the non-peptidic inhibitors have
compromised oral availability as they feature either a guanidine or amidine moiety forming a key interaction
with the Asp189 carboxylate in the S1 pocket of the active site of uPA.5 We have previously reported the use
of a benzylamine to pick up this crucial interaction6. Frederickson et. al. disclosed a series of weakly basic
orally bioavailable inhibitors of uPA showing the same interaction with Asp 1897. This change led to improved
oral bioavailability. Here we demonstrate that new orally bioavailable benzylamine derivatives of small
molecule uPA inhibitors can reduce disease scores in EAE models and provide a basis for development of
novel MS therapies.
O
OH
H2N
O
F
N
O
F
1
Figure 1. Benzylamine lead.
Literature known compound 1 displaying a benzylamine moiety was selected for further optimization.
Compound 1, showed potency against uPA (IC50 = 240 nM) and selectivity against two related trypsin family
serine proteases, tPA and plasmin (IC50 = 22 M and 5.0 M respectively). Additionally this compound has an
orally bioavailability of 37 % in rat7. However, improvements in potency and oral bioavailability were still
desired.
The structure of compound 1 displaying four aromatic moieties enables a modular modification strategy ideal
for SAR exploration.12 The synthesis shown in Scheme 1 is representative for the synthetic methodology used
to prepare the benzylamines shown in Table 1. Pyridine derivative 3 was synthesized via a regioselective
nucleophilic aromatic substitution (SNAr) reaction between tetrafluoropyridine 1 and tert-butyl (R)-2-
hydroxybutyrate 2 at 0oC. Synthesis of the left hand side was accomplished by forming triflate 5 from anisole
4 and then reaction of the triflate with 3-cyanophenylboronic acid by a Suzuki reaction to provide 6.
Demethylation using boron tribromide afforded phenol 7. The nitrile group of 7 was reduced with Raney
Nickel and selectively protected using Schotten-Bauman conditions to give BOC protected phenol 9. A
second SNAr reaction between pyridine derivative 3 and phenol 9 required heating in order to furnish fully
protected compound 10. Cleavage of both protecting groups using trifluoroacetic acid then provided the final
compound 824859.12
Scheme 1: Synthesis of ZK-824859

O
O
F
F
N
F
F
a
F
F
N
O
F
HO
OtBu
+
OtBu
1
2
3
HO
O
TfO
O
c
b
O
R
NC
6
7
R = Me
R = H
d
4
5
O
OH
BocHN
O
F
N
O
F
R
f
OtBu
10
8
R = CN
e
9
R = CH2NHBoc
O
H2N
O
F
N
O
g
OH
F
824859
Reagents and conditions: (a) NaH, THF, 0° C,80% (b) Tf2O, DMAP, DIEA, 0oC, CH2Cl2 88% (c) 3-Cyanophenyl boronic acid, Pd(PPh3)4, 1M
Na2CO3, EtOH, Toluene, 90°C 67%(d) BBr3, CH2Cl2 86%(e) i. RaNi, NH3, H2, IPA, MeOH 88% ii. Boc2O, NaOH, Dioxane, H2O 92%(f) 3,
Cs2CO3, DMSO, 85°C 88%(e) TFA, CH2Cl2 72%
Prior SAR work had demonstrated that both the amine and carboxylic acid were necessary for uPA inhibition.7
In addition, modifications at the pyridine ring decreased either the potency for uPA inhibition or the oral
bioavailability. Therefore, we focused on the biphenyl and salicylate units of Compound 1. Replacement of
the salicylate unit by a glycolate moiety originally resulted in significant loss of activity (824017). However,
with the introduction of some steric bulk at the R2 position, i.e. methyl (824216) and ethyl (824190), the
potency improved 10 fold. Comparison of enantiomers (824190 and 824184) showed that the R-enantiomer
was 30-fold more potent than its S counterpart. This trend was consistently seen throughout the glycolate
series. Increasing the steric bulk by introducing a tert-butyl group (825895) even further improved the potency
approximately 4 fold.
Table 1. SAR of Glycolate Series
O
H2N
O
F
N
O
F
OH
R2
R1
uPA IC50
(nM)
3280
360
237
7260
84
tPA IC50
(nM)
ND
Plasmin
Cmpd
R1
R2
IC50 (nM)
ND
5100
824017
824216
824190
824184
825895
824859
825095
825200
825184
H
H
H
H
H
Me
Me
OMe
CN
H
Mea
Eta
ND
1600
>30000
1450
1580
470
1850
Etb
>30000
5450
1330
1900
3325
5400
tBua
Eta
79
51
23
34
tBua
tBua
tBua
1850
5700

aR-enantiomer
bS-enantiomer
cValues are the mean of ≥ 2 experiments.
For assay descriptions see ref 8. ND = Not determined
Modifications at the position R1 were found to affect potency and selectivity. For instance, in the ethyl
glycolate series switching from R1 = hydrogen (824190) to methyl (824859) provided a 4 fold increase in uPA
potency without sacrificing selectivity. Unfortunately, the same modification in the tert-butyl series (825895
to 825095) gave only a slight increase in uPA potency but a loss in selectivity. However, when polarity was
introduced at R1 with either an electron donating methoxy group (825200) or a nitrile group (825184), uPA
potency improved and the selectivity was enhanced.
The pharmacokinetic profiles in rats of two uPA inhibitors are shown in Table 2. Beginning with moderately
potent ethyl glycolate 824190, we found that this compound exhibited a relatively high oral availability
F=55%, but with a lower level of exposure ( Cmax=0.64 g/mL and AUCall=2.2 h* g/mL) when dosed applying
2 mg/kg PO. Structural modifications at R1 did not only change potency and selectivity for uPA, but changes at
this position also affected the PK profile. This is exemplified in compound 824859, where incorporation of a
methyl group at the R1 position improved the overall exposure to Cmax=5.3 g/mL (AUCall=13.2 h* g/mL,)
while the oral availability F=30% remained acceptable.
Table 2. PK results for selected compounds
PK Data
Cmax
( g/mL)
T1/2
(h)
Cl
(mL/min/kg)
Vd
(L/kg)
Cmpd
%F
824190
824859
0.64
5.3
2.8
2.2
8.4
4
0.76
0.60
55
30
In the frame of lead optimization, compound design was guided by a combination of computational models
and X-ray crystallography. Figure 2 shows a model of the complex formed between compound 824859 and
human uPA. The biphenyl moiety binds in the S1 pocket with the amino group forming a salt bridge to
Asp189 (2.7Å) at the bottom of the pocket. In addition, a hydrogen-bond (H-bond) to the carbonyls of Ser190
(2.7Å) and Gly218 (2.8Å) is formed. The glycolate carboxylic acid mimics the cleavage product and forms an
H-bond to the catalytically active Ser195 (2.5Å). This close interaction with Ser195 has been observed in the
crystal structures of other serine proteases and is indicative of the acyl enzyme intermediate formed during
catalysis. In addition, the carboxylic acid moiety forms H-bonds to the side chain of His57 (2.7Å) and the
amide nitrogen of Gly193 (2.8Å).

Figure 2. A model of Compound 824859 bound to human uPA. The inhibitor is shown with orange carbon atoms and cyan fluorines. H-bonds
are shown with dashed lines. The model was derived from the published coordinates of homologous inhibitor, 3IG6.PDB,7 using MOE/CCG
software
The hydrophobic parts of the inhibitor play an important role explaining the selectivity of compound 824859.
The ethyl group extends into the S1’ site, packing against the Cys42-58 disulfide bond (3.6Å). The methyl
substituent on the biphenyl extends into the S1 pocket,9 interacting with the side chain of Cys220 (3.7Å). The
model also indicates a possible interaction between the methyl and Ser146 (6Å). In other uPA crystal
structures (4FUI.PDB) Ser146 is shifted and donates a side chain H-bond to the ligand. In our hands,
substituents with H-bond acceptors, 825184 (CN group) and 825200 (OMe group), showed improved
selectivity against proteins where the corresponding residue could not form this H-bond (table 1). Finally, the
difluoropyridine packs in the S2 pocket with one of the fluorines in close contact with His99 (3.3Å). Residue
99 varies significantly in other homologous proteases. For example, tPA has a tyrosine at this position which
would cause a steric clash with the fluorine. Contrary, plasmin has a 6 residue deletion from residues 95-100
which would cause the pyridine ring to loose contact with the protein.
In order to test the hypothesis that small molecule inhibitors of uPA could ameliorate MS, we tested the ability
of ZK 824859 to improve outcomes in both acute and chronic mouse EAE models10. We picked ZK 824859
due to its overall favorable pharmacokinetic profile. Unfortunately, due to sequence differences between
human and murine proteases, ZK 824859 is 5 fold less potent and has lost selectivity in mouse: uPA IC50
=
410 nM; tPA IC50 = 910 nM; plasmin IC50 = 1600 nM compared to human IC50 values of 79 nM; 1580 nM and
1330 nM respectively.
Figure 3. Efficacy of ZK-824859 in mouse chronic (adoptive) EAE. Female SJL mice (n=10 per experimental group) were immunized with
encephalitogenic T cells and treated twice daily with vehicle control, ZK-824859 at 3 doses, or Prednisolone. Clinical score was assessed daily.
In a chronic mouse EAE model (Figure 3) at a dose of 50 mg/kg b.i.d. ZK 824859 completely prevented the
development of disease. However, two lower doses (25 and 10 mg/kg) had no effect on clinical scores. In an
acute mouse EAE model (Figure 4), ZK 824859 was also efficacious. In this study the 50 and 25 mg/kg b.i.d.
doses showed a reduction in clinical scores compared to the control group, although only the 25 mg/kg
treatment regime was statistically significant compared to the control group.
We chose to test the compounds in both mouse models because those models allow a ready comparison of the
acute and chronic EAE models despite the known species differences in potency and selectivity of the
compounds between mouse and human uPA. Due to the poor selectivity of ZK 824859 for uPA against tPA

and plasmin in the mouse system, we cannot definitively conclude that the effects observed in these efficacy
studies are uPA-dependent. However, taking into the account that tPA deficiency leads to earlier onset and
increased severity of the disease,3, 4 it is not likely that tPA inhibition contributed to the efficacy observed. If
the pathophysiological mechanism by which uPA is causing or exacerbating EAE is activation of plasminogen,
then inhibition of uPA and plasmin will be expected to have similar outcomes in improving the disease. This is
confirmed by a rat model of EAE applying an inhibitor of plasmin, epsilon amino-caproic acid, showing that
the outcomes were improved11.
Figure 4. Efficacy of ZK-824859 in mouse acute (active) EAE. Female SJL mice (n=10 per experimental group) were immunized with PLP
139-151/FCA and treated twice daily with vehicle control, ZK-824859 at 3 doses, or Prednisolone. Clinical score was assessed daily.
In conclusion, we have optimized a series of benzylamines resulting in potent, selective and orally bioavailable
inhibitors of uPA. Remarkably, a member of this series, ZK824859, lowered clinical scores when dosed in
both acute and chronic mouse EAE models. Future efficacy testing will be in EAE models in primate species
such as marmosets. Thus, application of small molecule inhibitors of uPA may represent a new therapeutic
strategy for the treatment of multiple sclerosis.
References
1. Werb, Z.; Ashkenas, J.; MacAuley, A.; Wiesen, J. F. Braz. J. Med. Biol. Res. 1996, 29, 1087.
2. Gveric, D.; Hanemaaijer, R.; Newcombe, J.; van Lent, N. A.; Sier, C. F. M; Cuzner, M. L. Brain 2001, 124,
1978.
3. Lu, W.; Bhasin, M.; Tsirka, S. E. J. Neuroscience 2002, 22(24), 10781.
4. East, E.; Baker, D.; Gareth, P.; Lijnen, H. R.; Cuzner, M. L.; Gveric, D. Am. J. Pathology 2005, 167(2), 545.
5. a) Bruncko, M.; McClellan, W. J.; Wendt, M. D.; Sauer, D. R.; Geyer, A.; Dalton, C. R.; Kaminski, M. A.;
Weitzberg, M.; Gong, J.; Dellaria, J. F.; Mantei, R.; Zhao, X.; Nienaber, V. L.; Stewart, K.; Klinghofer, V.;
Bouska, J.; Rockway, T. W.; Giranda, V. L. Bioorg. Med. Chem. Lett. 2005, 15, 93. b) Fish, P. V.; Barber, C. G.;
Brown, D. G.; Butt, R.; Collis, M. G.; Dickinson, R. P.; Henry, B. T.; Horne, V. A.; Huggins, J. P.; King, E.;
O’Gara, M.; McCleverty, D.; McIntosh, F.; Phillips, C.; Webster, R. J. Med. Chem. 2007, 50, 2341. c) Mackman,
R. L.; Hui, H. C.; Breitenbucher, J. G.; Katz, B. A.; Luong, C.; Martelli, A.; McGee, D.; Radika, K.; Sendzik,
M.; Spencer, J. R.; Sprengeler, P. A.; Tario, J.; Verner, E.; Wang, J. Bioorg. Med. Chem. Lett. 2002, 12, 2019. d)
Joossens, J.; Ali, O. M.; El-Sayed, I.; Surpateanu, G.; Van der Veken, P.; Lambeir, A.; Setyono-Han, B.;
Foekens, J. A.; Schneider, A.; Schmalix, W.; Haemers, A.; Augustyns, K. J. Med. Chem. 2007, 50, 6638. e)
Botkjaer et al; molecular Cancer Research, 2012, 10, 1532 f) El-Nahas et al; Current Organic Chemistry, 2014,
18, 629
6. West, C. W.; Adler, M.; Arnaiz, D.; Chen, D.; Chu, K.; Gualtieri, G.; Ho, E.; Huwe, C.; Light, D.; Phillips, G.;
Pulk, R.; Sukovich, D.; Whitlow, M.; Yuan, S.; Bryant, J. Bioorg. Med. Chem. Lett. 2009, 19, 5712.
7. Frederickson, M.; Callaghan, O.; Chessari, G.; Congreve, M.; Cowan, S. R.; Matthews, J. E.; McMenamin, R.;
Smith, D.-M.; Vinkovic, M.; Wallis, N. G. J. Med. Chem. 2008, 51, 183.
8. uPA, tPA, and plasmin assays were performed in a 96-well format adapted from protocols described by: a)
Lottenberg, R.; Christensen, U.; Jackson, C. M.; Coleman, P. L. Methods in Enzymology, 1981, 80, 341. b)

Liang, A. M.; Light, D. R.; Kochanny, M.; Rumennik, G.; Trinh, L.; Lentz, D.; Post, J.; Morser, J.; Snider, M.
Biochemical Pharmacology 2003, 65, 1407.
9. Nienaber, V.; Davidson, D.; Edalji, R.; Giranda, V. L.; Klinghofer, V.; Henkin, J.; Magdalinos, P.; Mantei, R.;
Merrick, S.; Severin, J. M.; Smith, R. A.; Stewart, K.; Walter, K.; Wang, J.; Wendt, M.; Weitzberg, M.; Zhao,
X.; Rockway, T. Structure, 2000, 8, 553.
10. a) Lassman, H.; Bradl, M. Acta Neuropathol, 2017, 133, 222 b) Constantinescu, C. et al; British J. Pharmacol.
2011, 164, 1079
11. Sibley, W.T.; Kiernat, J.; Laguna, J. F.; Neurology. 1978, 28, 102.
12.All final compounds were characterized by NMR, HPLC, MS, and EA. Additional synthetic details will be
provided in future publications.

Graphic Abstract:
O
OH
O
H2N
O
F
N
O
F
H2N
O
F
N
O
F
OH
Benzylamine Lead 1
824859
uPA IC50 = 240 nM
tPA IC50 = 22 M
uPA IC50 = 79 nM
tPA IC50 = 1.58 M
Plasmin IC50 = 5 M
Plasmin IC50 = 1.33 M

Highlights:
Design and synthesis of benzylamine based selective and orally available inhibitors of uPA has
been described
Prior knowledge of SAR has been used to optimize the lead compounds
Structure based drug design and modeling afforded analogs with acceptable drug like properties
Compound 824859 shows efficacy in acute and chronic EAE model for multiple sclerosis
This may represents new therapeutic strategy to treat MS using small molecule uPA inhibitors