Discovery and SAR of novel and selective inhibitors of urokinase plasminogen activator (uPA) with an imidazo[1,2-A]pyridine scaffold

Article abstract of DOI:10.1021/acs.jmedchem.5b01171

Urokinase plasminogen activator (uPA) is a biomarker and therapeutic target for several cancer types. Its inhibition is regarded as a promising, noncytotoxic approach in cancer therapy by blocking growth and/or metastasis of solid tumors. Earlier, we reported the modified substrate activity screening (MSAS) approach and applied it for the identification of fragments with affinity for uPA's S1 pocket. Here, these fragments are transformed into a novel class of uPA inhibitors with an imidazo[1,2-A]pyridine scaffold. The SAR for uPA inhibition around this scaffold is explored, and the best compounds in the series have nanomolar uPA affinity and selectivity with respect to the related trypsin-like serine proteases (thrombin, tPA, FXa, plasmin, plasma kallikrein, trypsin, FVIIa). Finally, the approach followed for translating fragments into small molecules with a decorated scaffold architecture is conceptually straightforward and can be expected to be broadly applicable in fragment-based drug design.

Full text of DOI:10.1021/acs.jmedchem.5b01171

Subscriber access provided by UQ Library
Article
Discovery and SAR of novel and selective inhibitors of urokinase
plasminogen activator (uPA) with an imidazo[1,2-a]pyridine scaffold
Rafaela Gladysz, Yves Adriaenssens, Hans De Winter, Jurgen Joossens,
Anne-Marie Lambeir, Koen Augustyns, and Pieter Van der Veken
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01171 • Publication Date (Web): 17 Nov 2015
Downloaded from http://pubs.acs.org on November 18, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery and SAR of novel and selective inhibitors
of urokinase plasminogen activator (uPA) with an
imidazo[1,2ꢀa]pyridine scaffold
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
±
±
±
±
Rafaela Gladysz , Yves Adriaenssens , Hans De Winter , Jurgen Joossens , AnneꢀMarie
≠
±
±*
Lambeir , Koen Augustyns and Pieter Van der Veken
±
Medicinal Chemistry (UAMC), Department of Pharmaceutical Sciences, University of Antwerp
UA), Universiteitsplein 1, Bꢀ2610 Wilrijk (Antwerp), Belgium
(
≠
Medical Biochemistry, Department of Pharmaceutical Sciences, University of Antwerp,
Universiteitsplein 1, Bꢀ2610 Wilrijk (Antwerp), Belgium
Keywords: urokinase inhibitors, MSAS, imidazo[1,2ꢀa]ꢀpyridine scaffold, GBB reaction
Abstract
Urokinase plasminogen activator (uPA) is a biomarker and therapeutic target for several cancer
types. Its inhibition is regarded as a promising, nonꢀcytotoxic approach in cancer therapy by
blocking growth and/or metastasis of solid tumors. Earlier, we reported the Modified Substrate
Activity Screening (MSAS) approach and applied it for the identification of fragments with
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
affinity for uPA’s S1 pocket. Here, these fragments are transformed into a novel class of uPA
inhibitors with an imidazo[1,2ꢀa]pyridine scaffold. The SAR for uPA inhibition around this
scaffold is explored and the best compounds in the series have nanomolar uPA affinity and
selectivity with respect to the related trypsinꢀlike serine proteases (thrombin, tPA, FXa, plasmin,
plasma kallikrein, trypsin, FVIIa). Finally, the approach followed for translating fragments into
small molecules with a decorated scaffold architecture is conceptually straightforward and can be
expected to be broadly applicable in fragmentꢀbased drug design.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction
Urokinase plasminogen activator (uPA, urokinase) is a trypsinꢀlike serine protease and a
1
,2,3
therapeutical target for many cancer types, including breast, ovarian, and pancreatic cancer.
It
is part of an extracellular enzyme system overexpressed in metastasizing solid tumors,
comprising the urokinaseꢀtype plasminogen activator (uPA), the plasminogen activator inhibitors
1
(PAI’s), tissueꢀtype plasminogen activator (tPA) and the uPA receptor (uPAR). The uPAR is an
important regulator of extracellular matrix (ECM) proteolysis, cellꢀECM interactions and cell
4
signaling. Binding of uPA to its receptor activates the enzyme and triggers a proteolytic cascade
through which plasminogen is converted into plasmin. This in turn activates matrix
1,4
metalloproteases (MMPs) leading to proteolytic degradation of the ECM components. As a
result, tumor cells degrade the surrounding tissue, invade into healthy tissue and migrate with the
blood stream to form metastasized tumors at distant organs. The uPA system also interacts with a
number of relevant molecularꢀbiological systems promoting tumor growth: it can activate growth
ACS Paragon Plus Environment
2

Page 3 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
factors, and interacts with proteins involved in cell adhesion and signal transduction (vitronectin,
2
,4
integrins).
Although uPA is a valuable oncology target, clinical development of uPA inhibitors has been
problematic. This is most probably related to the doubtful biopharmaceutical performance of
compounds developed so far and their insufficient selectivity with respect to other,
phylogenetically related trypsinꢀlike proteases. Nonetheless, the field of urokinase inhibitor
discovery still produces significant numbers of relevant compounds, mostly small molecules
with a competitive, reversible inhibition profile. Wilex’s orally available drug candidate WXꢀ671
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(1a, Figure 1) is currently the most advanced product. This compound is a proꢀdrug of the
corresponding amidineꢀbased, inhibitor WXꢀUK1 (1b, Figure 1). Interestingly, 1a (K = 410
i
nM) and 1b are the first inhibitors of uPA in oncology trials worldwide. The latter has
successfully completed two clinical phase I trials, and recently showed favorable results in
randomized phase II trial in patients with locally advanced nonꢀmetastatic pancreatic cancer as
well as met its primary objective in phase II trial in patients with HER2 receptor negative
5
metastatic breast cancer (MBC). Next to compounds developed by Wilex (1aꢀb), a significant
number of optimized inhibitors of urokinase has been reported, but these have only been
investigated in preclinical studies. Pfizer reported 1ꢀisoquinolinylguanidine UKꢀ356,202 (2; K =
i
3
7 nM) and its sulfonamide derivative UKꢀ371,804 (3; K = 10 nM) as potent and selective
i
6
inhibitors of uPA. These compounds were selected for preclinical evaluation for the treatment of
chronic dermal ulcers, condition characterized by high levels of uPA promoting uncontrolled
matrix breakdown and inhibition of wound repair. As a result, compound 3 was demonstrated to
inhibit exogenous uPA activity both in human chronic wound fluid in vitro, and in the porcine
6c
acute excisional wound model. Besides, according to Barber et al. 2 has been selected for
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6a
human clinical trials, but apparently these plans have never materialized for unknown reasons.
An amidine based, peptideꢀderived uPA inhibitor CJꢀ463 (4; K = 20 nM) was found to reduce
i
the number of experimental lung metastases in a fibrosarcoma mouse model as well as primary
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
tumor growth and metastasis formation in murine lung carcinoma model. A fragmentꢀbased
approach by Astex led to the discovery of a mexiletineꢀderived, low basicity inhibitor of uPA (5;
IC50 = 72 nM) with moderate selectivity against closely related proteases and high oral
8
bioavailability. Abbott Laboratories developed a series of naphthamidine inhibitors of uPA
(compound 6 as a representative example; K = 263 nM) with improved pharmacokinetic
i
9
properties, as good oral bioavailability and extended half life in rats. Additionally, the 4ꢀ
oxazolidinone analogue UK122 (7) was reported by Zhu and coꢀworkers as a selective inhibitor
of uPA with significant potency to inhibit the migratory and invasive capacity of pancreatic
10
cancer cells in vitro.
Other approaches have been focusing on discovery of irreversible uPA ligands, antibodies or
peptideꢀbased molecules. Our group described series of highly potent and selective, irreversible
diaryl phosphonate inhibitors of uPA (8a; IC = 3.1 ± 0.5 nM, and 8b; IC = 3.4 ± 0.4 nM) with
5
0
50
1
1
significant antiꢀmetastatic activity in a rodent model of breast cancer. Mazar and coꢀworkers
invented a novel therapeutic uPAR antibody ATNꢀ658 demonstrating antitumor effects across a
variety of tumor models, including inhibition of invasion, metastasis and proliferation as well as
12
induction of apoptosis. Additionally, bicyclic peptide constructs were recently reported as
13
highly potent uPA inhibitors. Nonetheless, given all the preclinical evidence mentioned, it is
remarkable that none of the compounds were developed clinically.
ACS Paragon Plus Environment
4

Page 5 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Reported inhibitors of uPA.
This study focuses on the discovery of a novel class of reversible, nonꢀpeptide derived uPA
inhibitors with an imidazo[1,2ꢀa]pyridine scaffold. It directly furthers on an earlier publication in
which we reported an optimized version of Substrate Activity Screening (SAS), an approach
pioneered by Ellman and coꢀworkers for the discovery of fragments with affinity for the active
center of enzymes. Our “Modified Substrate Activity Screening” (MSAS) approach was
1
4
validated by identifying several fragmentꢀsized ligands of the uPA S1 pocket. As hypothesized
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
in this earlier report, an S1ꢀbinding fragment could be transformed into a druglike uPA inhibitor
by grafting it onto a suitable scaffold and by further decorating this scaffold with additional
affinityꢀconferring substituents. This hypothesis will be extensively investigated here and
corroborated with the preparation of potent uPA inhibitors. Since, to the best of our knowledge,
no earlier examples of imidazopyridine inhibitors of urokinase have been reported, evaluation of
the inhibitory potencies under the assay conditions of this manuscript will include the reference
compounds 7, gabexate (9) and amiloride (10) (Figure 2).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Inhibitory activities of the reference compounds against uPA determined under the
assay conditions of this manuscript.
The S1ꢀbinding fragments used in this study were demonstrated in our earlier report to possess
high micromolar uPA affinity. Their structures are summarized in Figure 3 (entry A).
Noteworthy, our earlier publication already included a proofꢀofꢀconcept example where fragment
1
b
R was grafted onto the 3ꢀposition of an imidazopyridine scaffold. A comparison of the IC ꢀ
5
0
ACS Paragon Plus Environment
6

Page 7 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
values of the separate and the scaffolded fragment indicated the imidazopyridine system to be a
novel, potentially useful scaffold for uPA inhibitors.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Overview of the strategy followed for the preparation of potent uPA inhibitors with an
imidazopyridine scaffold.
The general strategy that was followed in this study for obtaining potent uPA inhibitors,
consists of two parts. In the first part (Figure 3, entry B, Step1), a set of imidazo[1,2ꢀ
1
aꢀh
a]pyridines is prepared, each bearing one of the S1ꢀbinding fragments (R ). The choice to
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
introduce the S1ꢀbinders at the 3ꢀposition of the imidazopyridine is mainly governed by practical
synthetic considerations (vide infra), although our earlier proofꢀofꢀconcept work already
indicated this to be a viable approach for preparing compounds with appreciable uPAꢀaffinity.
Comparison of the target potencies of the monosubstituted scaffolds obtained in this manner, is
then used to identify the optimal S1ꢀbinding substituent. Further optimization (Figure 3, entry
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
B, Step 2) is achieved by combining the optimal S1ꢀsubstituent with additional substituents (R
3
and R ) on the imidazopyridine ring system. Although we extensively relied on molecular
2
3
modeling to guide the selection of the R and R substituents, diversification in terms of steric
and electronic parameters was an equally important goal.
The key step in the preparation of all target compounds of this study is the soꢀcalled Groebkeꢀ
BlackburnꢀBienaymé condensation, a variant of the Ugi reaction, reported for the first time in
15,16,17
1998.
It is based on the threeꢀcomponent coupling of (1) an isocyanide, (2) an aldehyde
and (3) a 2ꢀaminoazine in the presence of a suitable catalyst, usually a Lewis acid or Brønsted
acid, to generate fused imidazo[1,2ꢀa]heterocycles in a oneꢀpot transformation. Taking into
account the limited number of commercially available isocyanides and their often nonꢀ
straightforward synthesis, we considered the latter reaction component the least suited as a
source of molecular diversity. As mentioned, this practical consideration was instrumental to
1
reserve the isocyanideꢀderived R group for the S1ꢀbinding fragments identified in the MSAS
hits. Conversely, aldehydes are commercially available in abundant numbers. Therefore, the
2
aldehydeꢀderived R ꢀsubstituent was deemed an appropriate source of steric and electronic
diversity. Finally, we also found the commercial availability of decorated 2ꢀaminopyridines to be
rather limited. Nonetheless, a significant number of 2ꢀaminopyridines equipped with functional
groups (e.g. carboxylates, halides) that can easily be derivatized using standard chemical
ACS Paragon Plus Environment
8

Page 9 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
transformations were found to be available. This approach was used to generate sufficient
diversity in the produced compound series.
On a general level, imidazo[1,2ꢀa]pyridines have already been applied successfully as scaffold
moieties in drug discovery, and the GroebkeꢀBlackburnꢀBienaymé reaction represents one of the
simplest routes for the diversityꢀoriented synthesis of this pharmacophore. Compounds of this
type have been in clinical investigation for various therapeutic targets resulting in several drugs
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
entering the market, as for instance the hypnotic GABA receptor ligand zolpidem, the selective
A
phosphodiesterase 3 (PDE3) inhibitor olprinone, or the CXC chemokine receptor 4 (CXCR4)
18
antagonist GSK812397. Our group, to the best of our knowledge, is the first to report the
application of an imidazo[1,2ꢀa]pyridine scaffold for the construction of uPA inhibitors. Next to
producing novel and potent uPAꢀinhibitors, this study also discloses a significant amount of
structureꢀactivity relationship data for this class of compounds. Finally, the present study also
successfully exemplifies a novel strategy for transforming lowꢀaffinity fragments into potent,
druglike compounds with a decorated scaffold architecture. Theoretically, the same strategy
(fragment scaffolding, followed by optimization through diversityꢀoriented introduction of
additional substituents) could be applied for all targets where a fragmentꢀbased approach to drug
discovery is followed.
Results and discussion
1
aꢀh
First, the set of imidazo[1,2ꢀa]pyridines linked with S1ꢀbinding fragments R , was
synthesized relying on the GroebkeꢀBlackburnꢀBienaymé (GBB) reaction (Scheme 1).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 1. Synthetic steps leading to the monosubstituted scaffoldꢀbased inhibitors 14h,
a
15aꢀg.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) ethyl formate, TEA, 55 °C, 24 h, 55ꢀ90%; (b) POCl , DIPA, DCM,
3
60ꢀ85%; (c) pyridinꢀ2ꢀamine, glyoxylic acid monohydrate, HClO (cat), MeOH, rt, 24 h, 38ꢀ
4
66%; (d) TFA/DCM (1:1), rt, 1 h, 95ꢀ100%.
The isocyanides (13aꢀh) required for this reaction were synthesized from amines 11aꢀh by
consecutive formylation and dehydration. The formylation reaction was carried out by refluxing
the corresponding amines overnight in ethyl formate and in the presence of triethylamine as
1
9
previously described by Hartman et al., to afford the desired formamides (12aꢀh) in 55ꢀ90%
yield. Dehydration in the presence of POCl and triethylamine led to the formation of the
3
isocyanides (13aꢀh) in low to moderate yields. Careful optimization of the reaction conditions,
including the nature of base and the reaction medium, were therefore necessary. Following the
protocol described by Ugi et al., we replaced triethylamine by diisopropylamine, which
increased not only the average yields (60ꢀ85%), but also purity, making chromatographic
2
0
purification in some cases avoidable. Detailed experimental procedures and characterization
data for preparation of amines 11aꢀh, formamides 12aꢀh, and isocyanides 13aꢀh can be found in
the Experimental Section.
ACS Paragon Plus Environment
10

Page 11 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
3
For obtaining the monosubstituted imidazopyridines (R =R =H), the GBBꢀreaction requires
formaldehyde as the aldehyde component. However, the scope of this nonꢀconcerted [4+1]
2
1,22
cycloaddition is rather limited for formaldehyde.
This is due to formation of unstable imines,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
resulting in poor conversions. Our initial attempts to use formaldehyde hydrate and
paraformaldehyde as potential formaldehyde equivalents did not produce satisfactory results and
afforded the desired product in poor yields (< 30%). Following the protocol reported by Kercher
2
3
et al., we then applied glyoxylic acid as a formaldehyde equivalent, and reacted it with 2ꢀ
aminopyridine and the set of isocyanides. Glyoxylic acid was indeed found to be an efficient and
experimentally more convenient reagent. Further optimization of the original experimental
protocol was achieved by using HClO as the catalyst and MeOH as the solvent. This resulted in
4
conditions applicable to all isocyanides that were evaluated and yields for the desired products
(14h, 15aꢀg) ranging between 38 and 65% yield. All reactions were completed within 24 h.
Evaluation of the uPA inhibitory potency of these monosubstituted imidazo[1,2ꢀa]pyridines,
revealed that the most potent analogues were guanidinophenyl derivatives 15a and 15b (bearing
1a
1b
fragments R and R ) (Table 1). These displayed IC ꢀvalues of 9.04 ± 0.62 µM and 19.39 ±
5
0
1.22 µM, respectively. The hypothetical binding modes of these closely related molecules were
investigated by docking studies. Results suggested the guanidine side chain of both compounds
to interact through hydrogen bonds with Glyꢀ219, Serꢀ190 and, as reasonably expectable, with
the anionic carboxylate function of uPA’s Asp189 (Figure 4). In addition, a hydrogen bond
between the amine nitrogen of the imidazopyridine ring system and the catalytically important
Serꢀ195 is also predicted both in the case of compounds 15a and 15b (Figure 4). A helicopter
view of the predicted binding of compound 15a within the active site pocket of uPA is given in
Figure 5a, showing complementarity between ligand and pocket. The guanidine moiety of the
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
compound is deeply buried in the active site, while the imidazole nitrogen of the imidazopyridine
ring system is pointing into the solvent and not participating in any hydrogen bonding to the
protein. The imidazopyridine scaffold in these simulations was found not to contribute to target
affinity via directed interactions with the enzyme, although it also does not negatively interfere
with binding of 15a and 15b to the active center.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Predicted binding mode of compounds 15a (light brown) and 15b (yellow) in the
active site of uPA (PDBꢀcode 2O8W). Proposed hydrogen bonds between ligands and labeled
protein residues are indicated by yellow dashed lines. Both compounds overlay quite well and
show an almost identical binding pattern.
The significant potency of 15a and 15b also reflects the ranking of their corresponding S1ꢀ
1
a
1b
fragments R and R during the Modified SAS (MSAS) experiments mentioned earlier, where
both displayed among the highest affinities in the collection of „hits”. For the other inhibitors in
ACS Paragon Plus Environment
12

Page 13 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1 however, the translation of fragment ranking data into potency of the corresponding
imidazopyridine inhibitors, seems less straightforward. In general, affinities for compounds 14h,
15cꢀg are in the high micromolar range. Noteworthy, the presence of a guanidinobutyl fragment
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
in 15c did not lead to potent inhibition, despite this fragment’s resemblance to the arginine side
chain that is used by uPA for recognition of its natural substrates. Likewise, the
aminoethoxyphenyl fragments present in 15d and 15e displayed satisfactory potency during our
earlier MSAS experiments, but did not result in potent imidazopyridine inhibitors. In addition,
these fragments also occur in mexiletine analogue and/or its likes, and both of them were
8
confirmed as ligands of uPA’s S1 pocket using Xꢀray crystallography. Comparable discrepancy
1
a
is present for the dichlorophenoxy fragment in 14h; while this moiety was ranked close to R
1
b
and R in the MSAS experiments, grafting it onto an imidazopyridine scaffold leads to a
relatively poor inhibitor. Finally, evaluation data for the aminomethylꢀsubstituted 15f and 15g
also demonstrated these compounds to be weak uPA inhibitors. Taken together, these
biochemical evaluation data led to prioritization of 15a and 15b in the optimization effort. The
slightly higher uPA potency, lower molecular weight and reduced conformational flexibility of
15a were decisive for focusing on this compound during optimization by introducing additional
substituents on different positions of the monosubstituted scaffold.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Biochemical evaluation of the monosubstituted analogues 14h, 15aꢀg against uPA.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
Cpd R =
IC (uPA) [µM]
50
14h
~250
1
1
5a
5b
9.04 ± 0.62
19.39 ± 1.22
~200
1
1
5c
5d
>1000
500
15e
1
1
5f
>500
500
5g
ACS Paragon Plus Environment
14

Page 15 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Optimization of the initial hit 15a began with investigating the influence of an additional
substituent at the 2ꢀposition of the imidazo[1,2ꢀa]pyridine scaffold (compounds 17aꢀk, Table 2).
For preparing these compounds, we reacted 2ꢀaminopyridine and the corresponding isocyanide
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
3a with different commercially available aldehydes (Scheme 2). Selection of the latter was
mainly done in a nonꢀtargetꢀbiased manner, aiming to cover as much of druglike chemical space
as possible by taking steric, electronic and electrostatic parameters into account.
a
Scheme 2. Synthesis of analogues modified at the 2ꢀposition 17aꢀk.
a
Reagents and conditions: (a) HClO (cat), MeOH, rt, 24 h, 22ꢀ86%; (b) TFA/DCM (1:1), rt, 1 h,
4
9
4ꢀ100%.
Preparation of the 2ꢀsubstitutedꢀ3ꢀaminoꢀimidazo[1,2ꢀa]pyridines followed the classical GBB
17
reaction experimental protocol. Reactions were performed in MeOH, and included preꢀ
formation of the imine intermediate in the presence of a catalytic amount of HClO in order to
4
suppress byꢀproducts formation. Although the applied experimental protocol in most cases gave
satisfactory results (60ꢀ86% yield), reaction with some aliphatic aldehydes (e.g.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
propionaldehyde, 2ꢀaminoacetaldehyde, 3ꢀchlorophenylacetaldehyde) led to poor or moderate
yields for compounds 16dꢀf (49%, 32%, and 22%, respectively), while 2ꢀmethoxyaldehyde and
3ꢀ(2,4ꢀdichlorophenyl)propanal failed to react under these conditions. Additionally, two
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
analogues of 15b (19a and 19b) with a guanidinophenethyl substituent at the 3ꢀposition were
prepared for comparison of the inhibitory potency (Table 3). Compound 19a has an ethyl
substituent at the 2ꢀposition of the imidazo[1,2ꢀa]pyridine ring, whereas analogue 19b is
comprising of the imidazo[1,2ꢀa]pyrazine core and carries 3ꢀchlorophenyl substituent at the 2ꢀ
position. Preparation of compounds 19a and 19b involved the previously used GBB reaction
protocol affording intermediates 18a and 18b in moderate yields (54 and 49%, respectively).
Most of the performed GBB reactions were complete within 24 h. The final step involved a
simple Boc deprotection to afford the desired products in quantitative yields in the form of TFAꢀ
salts.
Table 2. Biochemical evaluation of the 2ꢀsubstitutedꢀ3ꢀaminoꢀimidazo[1,2ꢀa]pyridines set
17aꢀk against uPA.
2
Cpd R =
IC (uPA) [µM]
50
1
1
7a
7b
250 ± 2.23
63.92 ± 9.42
ACS Paragon Plus Environment
16

Page 17 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
7c
30.25 ± 4.71
1
1
7d
7e
48.47 ± 8.47
9.30 ± 2.67
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
1
7f
18.03 ± 0.95
~250
7g
1
1
7h
7i
99.16 ± 8.69
30.15 ± 1.92
~100
17j
1
7k
~125
The uPA potency of the prepared set of 2ꢀsubstituted analogues was then evaluated (Table 2).
In general, most of the compounds within the 2ꢀsubstituted imidazo[1,2ꢀa]pyridines set displayed
reduced uPA affinity. Especially aromatic substituents and sterically demanding aliphatic
substituents at the 2ꢀposition were not favourable. This detrimental effect is tentatively explained
by the computational model used (vide supra), suggesting that sterically demanding substituents
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
at the 2ꢀposition increase the internal conformational energy when the compound is in a
conformation that is optimal for binding to the uPA S1 pocket. For instance, the 3ꢀpyridyl (17a),
3
ꢀfluophenyl (17j), 4ꢀfluophenyl (17k), and 3,5ꢀdichlorophenyl (17h) analogues have IC ꢀ
50
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
values ≥100 µM, indicating at least 10ꢀfold decrease in potency in comparison to the initial hit
15a. The same is true for the 4ꢀpiperidinylꢀsubstituted compound 17g, displaying an IC of more
5
0
than 100 µM. Interestingly, the benzylic substituents in 17c and 17f caused a relatively less
pronounced affinity decrease in comparison to 15a, probably due to increased conformational
flexibility. Finally, only the aminomethylꢀsubstituted compound 17e retained the initial binding
affinity with an IC value of 9.30 ± 2.67 µM.
5
0
The guanidinophenethyl derivatives 19a, 19b followed a grossly comparable potency pattern
(Table 3), although 19a’s significant drop in potency compared to 15b cannot be readily
rationalized by taking only steric factors into account.
Table 3. Biochemical evaluation of the C3ꢀguanidinophenethylꢀsubstituted analogues 19a,
19b against uPA.
2
Cpd
R =
X=
IC (uPA) [µM]
50
1
1
9a
9b
C
~150
N
53.78 ± 3.72
ACS Paragon Plus Environment
18

Page 19 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Before evaluation data for the 2ꢀsubstituted imidazo[1,2ꢀa]pyridines set were obtained (Table
2
), we decided to explore substitution at the C6ꢀC8 position conserving the 2ꢀpyridyl substituent
(
Scheme 3). The reason to conserve the latter was inspired by the practically very
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
straightforward synthesis and purification of the compound.
Nine additional 2ꢀ(3ꢀpyridyl) analogues (22aꢀi) (Table 4) were prepared by varying the
aminopyridine reaction partner of the GBB reaction. Aminopyridine building blocks were either
commercially available or synthesized separately. Accordingly, 2ꢀaminoꢀNꢀbutylisonicotinamide
(20a) and 6ꢀaminoꢀNꢀbutylnicotinamide (20b) used for the preparation of compounds 22d and
22h, were synthesized by the aminolysis reaction in the presence of an organocatalyst (1,5,7ꢀ
2
4
triazabicyclo[4.4.0]decꢀ5ꢀene, TBD), as reported by Kiesewetter et al. (Table 5, entry 2).
Details on the preparation of the latter can be found in the Experimental Section.
a
Scheme 3. Synthesis of analogues 22aꢀi.
a
Reagents and conditions: (a) HClO (cat), MeOH, rt, 24 h, 25ꢀ68% (in case of compound 21d (i)
4
methyl 2ꢀaminopyridineꢀ4ꢀcarboxylate, TBD, DMF, 120 °C, 20 h, 52.5%; (ii) HClO (cat),
4
MeOH, rt, 24 h, 60%; compound 21h (i) methyl 6ꢀaminonicotinate, TBD, DMF, 120 °C, 24 h,
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
2%; (ii) HClO (cat), MeOH, rt, 24 h, 47%); (b) TFA/DCM (1:1), rt, 1 h, 94ꢀ100% (in case of
4
compound 22c (i) NaOH (2 M), DCM/MeOH (9:1), 12 h, rt, 69%; (ii) TFA/DCM (1:1), rt, 1 h,
9
3%).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Noteworthy, aminopyridines substituted at the 4ꢀ or 5ꢀposition (e.g. compounds 21d and 21i)
provided better yields in the GBB reaction (60 and 68%, respectively), whereas 2ꢀ
aminopyridines substituted at the 3ꢀ or 6ꢀposition performed worse or failed to react under these
conditions. For instance, reaction with 3ꢀmethylpyridinꢀ2ꢀamine provided 21a only in 25% yield,
and 3ꢀ(trifluoromethyl)pyridinꢀ2ꢀamine was found not to react at all under these conditions.
Replacing the reaction solvent (MeOH) by the nonꢀnucleophilic trifluoroethanol, as originally
1
7
proposed by Bienaymé et al., still did not afford the desired product.
Table 4. Biochemical evaluation of the C2 3ꢀpyridyl substituted analogues 22aꢀi against
uPA.
3
4
5
Cpd R =
R =
R =
IC (uPA) [µM]
50
22a
H
H
27.44 ± 1.34
ACS Paragon Plus Environment
20

Page 21 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
2b
H
H
H
H
14.95 ± 0.81
26.85 ± 1.57
22c
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
2d
H
H
6.89 ± 0.80
2
2
2
2e
H
H
H
H
H
H
23.89 ± 2.69
~200
2f
2g
57.55 ± 6.58
2
2h
2i
H
H
H
H
60.84 ± 3.39
27.26 ± 2.19
2
The uPAꢀevaluation results of inhibitors 22aꢀi are presented in Table 4. These demonstrate
that the detrimental effect of the 2ꢀpyridyl substituent on uPA potency, as observed with 17a, can
be significantly reduced by introducing additional affinityꢀconferring substituents (Table 4) on
the scaffold. Nonetheless, the position of a substituent was found to have a significant effect on
inhibitory activity. Generally, analogues substituted at the 7ꢀposition of the imidazopyridine
scaffold displayed a higher affinity than compounds bearing a substituent at the 6ꢀposition. For
instance, the C7 methyl carboxylate moiety in 22b caused an increase of binding affinity of
almost one order of magnitude relative to 17a. Replacement of the 7ꢀmethyl carboxylate group
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
by an Nꢀbutylcarboxamide group in 22d leads to a further increase in affinity. On the other hand,
introducing the Nꢀbutylcarboxamide at the 6ꢀposition resulted in analogue 22h with a
significantly reduced uPA affinity (60.84 ± 3.39 µM) compared to 22d. Also compounds with
C6 electronꢀwithdrawing groups like trifluoromethylꢀ and fluoroꢀsubstituted analogues 22f and
22g displayed reduced inhibitory potency (IC50 ~200 µM and IC50 = 57.55 ± 6.58 µM,
respectively).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Again, the more favourable results of the 7ꢀsubstituted congeners can be tentatively explained
by a molecular docking study. Figure 5 clearly indicates that the 6ꢀposition of the
imidazopyridine scaffold approaches the surface of the pocket and might therefore be compatible
with only very small substituent types. On the contrary, the 7ꢀposition is more accessible for
attaching alternative substituents.
After obtaining the evaluation results, we decided to synthesize a set imidazo[1,2ꢀa]pyridines
lacking the 2ꢀpyridyl substituent, but with diverse amide groups at the scaffold’s 7ꢀposition
(compounds 25aꢀf). The Nꢀbutylcarboxamide substituent that was already present in 22d was
included in this series.
The synthetic strategy for these molecules involved formation of the required amide
substituent via aminolysis of methyl 2ꢀaminoꢀpyridinꢀ4ꢀcarboxylate, prior to the GBB reaction
with glyoxylic acid and isocyanide 13a (Scheme 4). Although this strategy does not introduce
molecular diversity during the final synthetic step, it was deemed preferable for two reasons.
First, the aminolysis reaction was found to require harsh reaction conditions (high temperature
and pressure, long reaction times) that were not compatible with the labile diꢀBocꢀprotecting
groups of the guanidine function, resulting in poor yields of the desired product (Table 5, entry
1
). Second, performing the GBB further on in the synthesis allowed us to reduce the reaction
ACS Paragon Plus Environment
22

Page 23 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
scale, and therefore consume less of the synthetically more demanding guanidinophenylꢀderived
isocyanide 13a. Again, the aminolysis was carried out in the presence of TBD. Optimization of
the experimental conditions used for preparation of amides 20aꢀb (Table 5, entry 2) revealed
that by reacting methyl 2ꢀaminopyridineꢀ4ꢀcarboxylate with a twoꢀfold excess of the amine
component, increasing the catalyst amount (0.3 eq) and performing the reaction in dry toluene at
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
110 °C, the desired amides (23aꢀe) were obtained in very good to excellent yields and without
the need of chromatographic purification (exemplified in Table 5, entry 4). The devised
protocol does not require protection of the amino group in aminopyridine and was found to be
applicable with comparable outcome to methyl 2ꢀaminopyridineꢀ3ꢀcarboxylate, methyl 6ꢀ
aminopyridineꢀ3ꢀcarboxylate. Detailed experimental procedures and characterization data for
amides 23aꢀe can be found in the Experimental Section. Next to aminolysis, direct coupling of 2ꢀ
aminopyridineꢀ4ꢀcarboxylate with the corresponding amine in DMF in the presence of EDC and
HOBt was also tried. However, this also returned the desired amide only in 10% yield (Table 5,
entry 5).
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 5. Conditions applied to amide bond formation.
Entry Amine
Carbonyl compound Catalyst Solvent
Conditions Yield
%]
or
[
coupling
reagent
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
nꢀbutylamine
2 eq)
TBD
(0.2 eq)
DMF
pressure
tube,
<10
(
120 °C,
24 h
2
3
4
5
nꢀbutylamine
2 eq)
TBD
(0.2 eq)
DMF
DMF
pressure
tube,
52.5
17.2
89
(
120°C,
20 h
cyclopropyl
amine
TBD
(0.2 eq)
pressure
tube,
120°C,
(2.2 eq)
21 h
cyclopropyl
amine
(2.2 eq)
TBD
(0.3 eq)
Toluene pressure
tube,
110°C,
17 h
cyclopropyl
amine
EDC and DMF
HOBt
65 °C,
20 h
10
(1.5 eq)
The prepared amideꢀsubstituted aminopyridines were then reacted with glyoxylic acid and
isocyanide 13a following the previously used GBB reaction protocol (Scheme 4). Interestingly,
reactions involving 2ꢀaminopyridines with 4ꢀsubstituent proceeded faster and most of them were
complete within 6 h. Products of the threeꢀcomponent coupling (3CC) 24aꢀf were then subjected
to a simple Boc deprotection to afford the desired compounds in the form of TFAꢀsalts, or in
ACS Paragon Plus Environment
24

Page 25 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
case of compound 25e, HClꢀsalt. Finally, six analogues (25aꢀf) were synthesized in moderate
yields (42ꢀ60%) (Table 6).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 4. Synthesis of the amideꢀsubstituted analogues 25aꢀf.
a
Reagents and conditions: (a) TBD, toluene, 110 °C, 20 h, 86ꢀ97% (in case of compound 20a
TBD, DMF, 120 °C, 20 h, 52.5%); (b) glyoxylic acid monohydrate, isocyanide 13a, HClO (cat),
4
MeOH, rt, 6 h, 42ꢀ61%; (c) TFA/DCM (1:1), rt, 1 h, 96ꢀ100%.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 6. Biochemical evaluation of the amideꢀsubstituted analogues 25aꢀf against uPA.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
Cpd
R =
IC (uPA) [µM]
50
2
2
5a
5b
0.097 ± 0.010
0.184 ± 0.007
2
2
5c
0.254 ± 0.016
5d
0.174 ± 0.021
0.366 ± 0.017
0.404 ± 0.020
25e
25f
ACS Paragon Plus Environment
26

Page 27 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Affinities of compounds within this series were in the nanomolar range, indicating an optimal
substitution pattern for the scaffoldꢀbased uPA inhibitors. The most potent analogue 25a (IC =
50
9
7 ± 10 nM) displays an increase in uPA affinity of about two orders of magnitude relative to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
initial hit 15a. Similar potency was demonstrated by the cyclopropylamideꢀsubstituted analogue
25b (IC = 184 ± 7 nM), whereas the 2ꢀ(4ꢀmethylpiperazine) ethylamide group in compound
5
0
2
5f (IC = 404 ± 20 nM) caused a 4ꢀfold affinity decrease in comparison to 25a. The modeling
50
study of the Nꢀbutylamide fragment of compound 25a in the active site of uPA highlights the
available space in this region (Figure 5b). Besides, formation of an additional hydrogen bond
between the amide nitrogen of the Nꢀbutylamide fragment and the hydroxyl group of Tyrꢀ151
can be observed and could potentially explain the increase in binding affinity that is observed
when comparing compound 25a with compound 15a (Figure 5a,b). The modeling study of
compounds 25bꢀf revealed similar interactions occurring in the uPA’s active site (Figure 5c).
This additional interaction might also contribute favourably to the compounds’ selectivity with
respect to related proteases. (vide infra)
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 98
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Helicopter view of uPA’s active site (PDBꢀcode 2O8W) showing the proposed binding
mode of compound 15a (panel a), compound 25a (panel b), compounds 25bꢀf (panel c). The
protein surface is colored white, while the surface generated by residue Serꢀ195 is colored red
and the surface generated by residues Serꢀ190, Aspꢀ189 and Glyꢀ219 is colored green. Putative
hydrogen bonds are shown as dashed yellow lines. Docking of compound 25a (panel b) reveals
ACS Paragon Plus Environment
28

Page 29 of 98
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
an additional hydrogen bond between the ligand and the hydroxyl group of Tyrꢀ151 (colored
4
blue). The R group (Table 6) of compound 25a is colored green and has been docked using
idealized geometrical bond and torsion angles.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kinetic analysis of the identified inhibitors 25aꢀf indicated that these are reversible, tightꢀ
binding, slowꢀdissociating inhibitors of uPA. In addition, selectivity of the most potent analogues
for uPA was determined with respect to a panel of closely related proteases (tissueꢀtype
plasminogen activator (tPA), thrombin, Factor Xa (FXa), plasmin, plasma kallikrein, trypsin,
Factor VIIa (FVIIa)). The results are summarized in Table 7. Compound 15a was found to be at
least 10ꢀfold more selective for uPA than for the other assayed enzymes, while the amideꢀ
substituted compounds 25aꢀf were found selective over all specified enzymes. High specificity
of these compounds for uPA can probably be related to interacting with Tyr in position 151 of
uPA (Figure 5b, c). Compound 25a demonstrated the best affinity as well as selectivity profile
among this set: with an IC > 100 µM for all of the related enzymes, this compound has a uPAꢀ
50
3
selectivity index of more than 10 .
Besides, since the phenylguanidine (26) fragment is an essential part of the identified
inhibitors, we decided to check its contribution to selectivity. Given the fact that no inhibition of
thrombin, tPA, FXa, plasmin, plasma kallikrein and FVIIa was observed at 100 µM
concentration of compound 26, we can conclude that indeed the phenylguanidine fragment is
crucial for selectivity of the identified inhibitors for uPA.
ACS Paragon Plus Environment
29