Highly Selective Sub-Nanomolar Cathepsin S Inhibitors by Merging Fragment Binders with Nitrile Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.0c00949

Pharmacological inhibition of cathepsin S (CatS) allows for a specific modulation of the adaptive immune system and many major diseases. Here, we used NMR fragment screening and crystal structure-aided merging to synthesize novel, highly selective CatS inhibitors with picomolar enzymatic Ki values and nanomolar functional activity in human Raji cells. Noncovalent fragment hits revealed binding hotspots, while the covalent inhibitor structure-activity relationship enabled efficient potency optimization.

Full text of DOI:10.1021/acs.jmedchem.0c00949

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Highly selective sub-nanomolar cathepsin S inhibitors by
merging fragment binders with covalent nitrile inhibitors
Markus Schade, Beatrix Merla, Bernhard Lesch, Markus Wagener,
Simone Timmermanns, Katrien Pletinckx, and Torsten Hertrampf
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00949 • Publication Date (Web): 03 Sep 2020
Downloaded from pubs.acs.org on September 3, 2020
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 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 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.
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 10
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
Highly selective sub-nanomolar cathepsin S inhibitors by merging
fragment binders with covalent nitrile inhibitors
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
Markus Schade,*† Beatrix Merla, Bernhard Lesch, Markus Wagener, Simone Timmermanns,
Katrien Pletinckx, Torsten Hertrampf
Grünenthal GmbH, Zieglerstr. 6, 52078 Aachen, Germany.
KEYWORDS cathepsin S, fragment merging, FBDD, NMR fragment screening, co-crystal structures, SAR analysis, inhibitor
ABSTRACT: Pharmacological inhibition of cathepsin S (CatS) allows for a specific modulation of the adaptive immune system
and many major diseases. Here we used NMR fragment screening and crystal structure-aided merging to synthesize novel,
highly selective CatS inhibitors with picomolar enzymatic Ki values and nanomolar functional activity in human Raji cells.
Non-covalent fragment hits revealed binding hotspots, while covalent inhibitor SAR enabled efficient potency optimization.
Introduction
Concomitant inhibition of CatK can lead to cardiovascular
events and stroke, as was demonstrated for the CatK
inhibitor odanacatib in clinical phase 3 studies.12 Subtype
selectivity over CatB and CatL was deemed beneficial due to
their ubiquitous expression pattern and role in various
cellular processes, such as thyroglobulin processing.3
Earlier clinical studies with basic CatK inhibitors revealed
adverse effects due to lysosomotropism in which inhibitors,
highly selective in vitro, are enriched so much in lysosomes
that selectivity is lost in vivo.4,13,14 Therefore basic nitrogens
must be avoided for cathepsin inhibitors.
The cysteine protease cathepsin S (CatS) has been
implicated as a causative enzyme in several wide-spread
diseases, including cancer, atherosclerosis, abdominal
aortic aneurysm (AAA),1 Sjögren’s syndrome,2 diabetes type
II, asthma, allograft rejections, neuropathic pain and
obesity.3 Accordingly, inhibitors of CatS have shown efficacy
in preclinical rodent models of antigen-induced arthritis,
ovalbumin-challenged allergic lung inflammation, high-fat
diet induced type II diabetes, colorectal cancer, peripheral
nerve injury-induced neuropathic pain.4,5 In patients the
expression of CatS protein or CatS proteolytic activity was
found to be elevated in Sjögren’s syndrome,6 rheumatoid
Here we discovered non-basic, picomolar CatS inhibitors
by merging novel NMR fragment screening hits with N-(1-
cyanocyclopropyl)amide inhibitors. 3D crystals structures
made this merging approach very efficient. Our top
compound is >10,000-fold selective over CatB, CatK and
CatL and inhibits Lip10 cleavage in human Raji cells with an
EC50 of 580 nM.
arthritis,
relapsing-remitting
multiple
sclerosis,
atherosclerosis, AAA, myocardial infarction, colorectal
cancer and obesity.4,7 These comprehensive data triggered
interventional clinical phase 2 trials with CatS inhibitors in
rheumatoid arthritis, psoriasis and Sjögren’s Syndrome and
clinical phase 1 trials in further indications. Furthermore,
many industry and academic groups published on
nanomolar CatS inhibitors of variable selectivity and
functional cell activity.8-11 However, no CatS inhibitor has
reached clinical phase III or approval status yet.
Results and Discussion
Thorough analysis of previous cathepsin inhibitor
chemical series showed that reversible or irreversible
covalent binding to the active site cysteine of cathepsins
provides excellent inhibitor potency, but non-covalent
interactions in the S2 and S3 pocket of cathepsins are
required for inhibitor selectivity.3 To arrive at a highly
selective CatS substrate, the active site cysteine must not be
irreversibly covalently captured. Hence both potency as
well as selectivity must originate from S2 and/or S3 pocket
interactions. For this reason, we decided against
synthetically elaborating a covalent CatS inhibitor series
from the literature. Rather we searched for non-covalent,
CatS belongs to a family of 11 cysteine proteases,
designated cathepsin B, C, F, H, K, L, O, V, W and X.4 In
contrast to other family members CatS expression is
restricted to antigen presenting cells, where CatS-catalyzed
cleavage of the invariant chain (li)-derived propeptide
Lip10 is essential for antigen presentation to CD4+ T cells.
In patients CatS inhibition leads to Lip10 accumulation in B
cells and myeloid dendritic cells, but not monocytes.6
Therefore inactivation of CatS allows for a highly CD4+ T
cell-targeted pharmacological intervention, provided that
highly CatS-selective inhibitors are used.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 10
1
2
3
4
5
6
7
8
synthetic binders to either the S2 or S3 pocket of CatS by
using fragment-based drug discovery.
By comparison, the most potent non-covalent CatS
inhibitors published by researchers from JnJ16, Eli Lilly1 and
Novartis17 show LE values of 0.25, 0.30 and 0.30
kcal/(#HA), respectively, as calculated from their in vitro
IC50 values.
Given its comparatively small size of 24 kDa and
favorable stability for biophysical techniques, CatS is well
suited for fragment screening by protein-observed NMR
spectroscopy.15 This screening technique brings along the
advantage that fragment-size binders to the S2 or S3 pocket
can be detected that do not touch the catalytic center of the
enzyme, which would be a very unlikely event in an
enzymatic activity-based screen of CatS. Consequently, we
screened our in-house library of 1858 chemically diverse
fragments against uniformly 15N-labeled CatS using 2D 15N-
HSQC NMR detection, which monitors direct binding of a
test substance to the mature, catalytically active human
CatS protein. We identified a total of 18 hits, all of which
show a similar pattern of chemical shift perturbations
(CSPs), suggesting binding to the same site of CatS (Supp.
Info_Fig. S1A). Comparison with a non-covalent CatS
inhibitor, which reached clinical phase 1 trials,1 shows that
several but not all CSPs are shared between this inhibitor
and our fragment binders, suggesting that their binding
sites partially overlap (Supp. Info Fig. S1B).
Next we asked whether those fragment binders inhibit
the enzymatic activity of human CatS and whether they
show selectivity against the anti-target human CatK using
fluorogenic substrate assays (Supp. Info). Six fragment hits
block CatS activity by more than 50% at a fragment
concentration of 500 M, whereas CatK activity was
blocked by less than 35% except for one fragment hit with
51% blockade (Supp. Info Tab. S1). We considered CatS
inhibition values of 50% or less within the noise area of this
fluorogenic enzyme assay.
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
As CatS has been very well enabled for structure-based
drug design,3 we next submitted all fragment hits for a
soaking protein crystallography assay. We obtained co-
crystal structures for the three chemically diverse fragment
hits 1, 2 and 3, all of which bind to S2 pocket of CatS, where
they occupy a narrow channel between residues Phe70 and
Phe211 of CatS (Fig. 1) (residue numbering of mature
human CatS). The piperazine of 1, the benzimidazole of 2
and the pyrimidine of 3 interact face-to-edge with the -
system of Phe70, while the [1,2,4]triazolo[4,3-b]pyridazine
of 1, the 2-aminoimidazole of 2 and the cyclohexene of 3
sandwich face-to-face with the -system of Phe211.
N
O
S
N
N
N
N
N
O
1
H2N
N
F
F
N
N
F
N
O
F
2
3
O
F
F
N
N
H2N
N
N
N
(A)
HN
(A)
S.Cys25 Cpd46
F70
G69
(B)
(B)
F70
S3
S1
3.9Å
S1‘
C25.S
4.1Å
F211
C25.S
S2
F211
Figure 2. Crystal structure-aided merging via mutual -CF3. (A)
Covalent carbonitrile CatS inhibitor Cpd46 with an IC50 of
10nM.18 (B) The X-ray structure superimposition of 2 with
Cpd46 shows that the trifluoromethyls of 2 (olive) and Cpd46
(green) fill the same binding hotspot on the S2 substrate
binding surface of CatS (same orientation as in Fig. 1).
Figure 1. (A) Fragment screening hits 1, 2 and 3. (B) 3D
crystal structures of 1 (green), 2 (olive) and 3 (red)
complexed with human CatS (semi-transparent hydrophobic
surface) shows binding to the extended S2 pocket between
residues F70 and F211 (distances labeled in Å). The catalytic
cysteine C25 and the substrate pockets S1’, S1, S2 and S3 are
labeled. (PDB-IDs: 1 6YYO, 2 6YYP, 3 6YYQ).
Comparison with the crystal structures of other human
catalytic cysteine cathepsins reveals that this narrow
channel is unique to CatS. In CatK a much bulkier Leu213
side chain in the place of Phe211 partially occludes the
narrow binding channel. Similarly, the cysteine proteases
cathepsin B, C, F, H, L and V all contain a non-aromatic
Using the NMR assay we measured dose-response curves
for our 18 fragments hits yielding Kd values between 2 and
740 µM and ligand efficiencies (LE) between 0.22 and 0.50
kcal/(no. of heavy atoms) (Supp. Info Fig. S1C and Tab. S1).
ACS Paragon Plus Environment

Page 3 of 10
1
Journal of Medicinal Chemistry
residue in position 211 of CatS. Cathepsin X has a His in 211
In
2006
covalent
N-(1-cyanocyclopropyl)amide
2
3
4
5
6
7
of CatS, but also a Asn in position 70 of CatS. These
mutations of Phe211 or Phe70 in the other cathepsins are
expected to abrogate the face-to-face or edge-to-face
interactions formed by 1, 2 and 3, suggesting the fragment
hits bind to an attractive, selectivity-conferring site in the
S2 pocket of CatS.
inhibitors were disclosed which carried aryl- and alkyl-
methanesulfonyl substituents in the S2-pocket of CatS,21 as
evidenced by the crystal structure of compound_16 (Cpd16)
bound to CatS (Fig. 3). This was followed by another
publication of picomolar N-(1-cyanocyclopropyl)amide
inhibitors with aryl- and alkyl-methanesulfonyl
substituents in the S2-pocket of CatS, designating the
sulfonyl moiety the key structural feature, although crystal
structures were not disclosed.22 Interestingly, when bound
to CatS the sulfonyl of these crystal-defined covalent
inhibitor series occupies virtually the same position as the
sulfonyl of 1 with their outgoing SO2-CH2 bond vectors
being well aligned (Fig. 3B). This opens up the opportunity
for merging 1 with these covalent inhibitor series.
8
9
Merging fragment binders with inhibitor series whose
structure-activity-relationships (SARs) are well explored, is
one of the most efficient strategies for fragment growing,
provided that 3D protein-ligand structures are available for
at least one member each. We found an intriguing 3D
structural overlap between the -CF3 of 2 and the -CF3 of a
potent series of covalent 1H-imidazo[4,5-c]pyridine-4-
carbonitrile inhibitors, represented by the X-ray structure
of compound 46 (Cpd46) bound to CatS with an IC50 value
of 10 nM (Fig. 2).12 These imidazo[4,5-c]pyridine inhibitors
were derived from a related series of (3-trifluoromethyl-
phenyl)pyrimidine-2-carbonitriles and 6-cyano-2-(3’-
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
Our crystal structure of 1 reveals van-der-Waals
distances of the one SO2 oxygen to atoms V162.O and
N163.O, and of the other SO2 oxygen to M71.S, but no H
bond contacts to CatS. Therefore we tested the importance
of the methanesulfonyl of 1 experimentally. 19 lacking this
methanesulfonyl does not show any detectable binding to
CatS in our NMR binding assay (Supp. Info Tab. 2A). This
suggests that our fragment hit 1 utilizes a sulfonyl binding
hotspot in the S2 pocket of CatS, identified previously by
classical SAR exploration of covalent CatS inhibitors.
trifluoromethyl-phenyl)-3,5-dioxo-1,2,4-triazines,
for
which the -CF3 group was essential for potency.19,20 This
suggests that -CF3 is a privileged interaction partner for this
hotspot of the CatS substrate binding surface and that our
non-covalent fragment hit 2 picked up this hotspot. While
the non-peptidic composition of these potent carbonitriles
is attractive, we had safety concerns about their potentially
reactive warhead and did not follow up on this merging
opportunity.
Knowing that the class of N-(1-cyanocyclopropyl)amide
inhibitors of cathepsins has completed clinical phase 2 and
3 trials with chronic oral administration,2,7,23 we considered
this warhead the safest and most attractive for fragment
elaboration. We first extended the sulfonyl of 1 with an N-
methylpropanamide to yield 37, which showed a 100-fold
lower Ki value of 0.7 M than 1 and does not inhibit CatK at
concentrations up to 100 M (Fig. 4A).
F
CF3
HN
O
O
S
N
H
N
N
H
N
O
O
N
N
O
H
(A)
Cpd16
N
N
N
S
N
O
(B)
O
(A)
37
F70
G69
(B)
F70
G69
3.0Å
2.8Å
3.8Å
3.0Å
Å
2.9Å
C25.S
Å
3.7Å
3.0Å
F211
C25.S
N163.O
V162.O
Figure 3. Crystal structure-aided merging via mutual -SO2-.
(A) Covalent CatS inhibitor Cpd16 with a Ki <=100 nM. (B)
The superimposition of the X-ray structures of 1 (green) and
Cpd16 (olive, PDB code=2FQ9) indicates a shared sulfonyl
binding hotspot in the S2-pocket of CatS. Distances between
the -SO2- oxygens of 1 and M71.S and V162.O are labeled in
blue. The H bonds of Cpd16 to G69.O, G69.N and N163.O are
shown in red.
F211
(B)
N163.O
Figure 4. Crystal structure of non-covalent S1-S2 pocket
inhibitor 37 (green, PDB-ID 6YYN) superimposed with Cpd16
(olive) from Fig. 3B. The H bonds of 37 to G69.N and N163.O
are shown in red.
The crystal structure of 37 complexed with CatS shows
that the amide of 37 forms the protein backbone H bonds in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 10
1
2
3
4
5
6
7
8
the S1 pocket of CatS (Fig. 4B), which are well known for N-
(1-cyanocyclopropyl)amide CatS inhibitors. The
substructure originating from 1 including the sulfonyl has
not changed its pose compared with the crystal structure of
1, indicating that the merging approach has worked as
planned.
The crystal structure of 133 bound to CatS shows that the
covalent bond to the catalytic Cys25 as well as the
conserved backbone H bonds between the S1 amide and the
NH of Gly69 and the carbonyl of Asn163 as well as the H
bond between the S3 amide and the carbonyl of Gly69 are
all in place (Fig. 5D). Furthermore the sulfonylpiperazine of
133 shows virtually the same pose as in 37, indicating a
successful merging of the S3-amide substituent onto 39b.
However, the 3-methylisoxazol of 133 does not occupy the
upper S3 pocket as observed for Cpd16 (Fig. 3B) or
odanacatib,12 but rather packs against the surface-exposed
edge of F70. It is unknown whether this surface-exposed
binding pose is the rule for all of our S3-pocket variants, but
it is certainly consistent with tolerating a diverse range of
aryl- and alkylamides with small changes in potency. Taken
together the >1000-fold boost in potency may originate
from adding that conserved H bond between the S3-pocket
NH of 133 and the carbonyl of Gly69, but conformational
restriction of the propanamide of 133 may also contribute
to potency.
We next focused on reducing the heteroatom count of the
6-(piperazin-1-yl)-[1,2,4]triazolo[4,3-b]pyridazine of 1. We
knew from our fragment hits 2 and 3 that diverse ring
systems are tolerated at this far end of the S2 pocket (Fig.
1). For ease of synthetic access we firstly varied 37 without
its N-methylpropanamide extension. Of the 17 analogues
tested, six analogues with piperazine-to-benzene
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
substitutions
and
1-(6-aminopyridin-2-yl)-4-
(methylsulfonyl)piperazine 23 showed affinity scores equal
or better than 1 in our NMR assay (Supp. Info Tab. 2A).
Since a piperazine-to-benzene substitution alters the
angle between the ring axis and the SO2-CH3 bond vector,24
we tested the favorable abovementioned analogues with
their N-methylpropanamide attachment. The (6-
aminopyridin-2-yl)piperazine analogue 38 and the
(pyridin-2-yl)piperazine analogue 39 bind to CatS almost as
tightly as 37, whereas all sulfonylbenzenes show much
weaker binding in our NMR assay (Supp. Info Tab. 2B).
O
S
N
N
N
N
N
N
H
N
N
N
N
N
N
N
O
O
O
O
(A)
(B)
39b
Keeping the simpler (pyridin-2-yl)-piperazine of 39, we
next replaced its N-methyl group with an N-(1-
cyanocyclopropyl), as present in the reversible covalent
CatS inhibitors for this merging approach. The resulting
nitrile 39b blocked CatS activity with a Ki value of 0.8 M
(Fig. 5A), which surprisingly is no improvement over the Ki
value of the non-covalent inhibitor 37. We concluded that a
covalent nitrile warhead with merely substituents in the S1
and S2 pocket of CatS is insufficient for high potency.
N
O
O
O
S
HN
H
N
O
54
N
Therefore we next aimed at filling the S3 pocket of CatS.
Superimposition of the crystal structures of 37 and the
covalent nitrile inhibitor Cpd16 shows that the C2-H bond
vector of the propenamide of 37 overlaps well with the C2-
NHX vector of Cpd16 with (R) stereochemistry at the entry
point into the S3 pocket (Fig. 4B). The NH of this amide S3
pocket substituent forms a H bond with the carbonyl of
residue Gly69. This H bond is conserved in the crystal
structures of several chemically diverse covalent cathepsin
inhibitors. To design in this privileged amide, we
synthesized the (R)-2-aminopropanamide derivate of 39b
and used automated amide coupling parallel synthesis to
make a total of 50 S3-pocket amides (Supp. Info Tab. 3A).
Compared with 39b, the S3-pocket amides showed an up to
5000-fold jump in potency, yielding Ki values as low as 90
pM. Equally surprising was the broad tolerance for diverse
alkyl- and arylamides in the S3 pocket. The 3,3-
dimethylbutanamide of 44 and the 2-(3-methylisoxazol-5-
yl)acetamide of 54 (Fig. 5B) produced similarly potent Ki
values of 90 and 345 pM, respectively, even though they
differ in shape and polarity. 2-aryl-acetamide S3-
substituents showed universally higher potencies than aryl-
carboxamides, but even the least potent example 93 with a
O
O
O
S
HN
H
N
O
(C)
133
F70
(B)
G69
2.9Å
2.9Å
3.8Å
2.8Å
C25.S
F211
N163.O
Figure 5. Potency jump by S2-pocket substituents. (A) S1-S3
pocket inhibitor 39b with a Ki value of 800 nM. (B) S3-pocket
amide 54 with a Ki value of 0.345 nM. (C) S1-S2-S3 pocket
inhibitor 133 with a Ki value of 1.4 nM. (D) The crystal
structure of 133 (green, PDB-ID 6YYR) bound to CatS shows
4-(trifluoromethoxy)benzamide
showed a Ki of 16 nM.
S3-substituent
still
ACS Paragon Plus Environment

Page 5 of 10
1
Journal of Medicinal Chemistry
the covalent thioimidate bond between C25.S and the nitrile
M 54 showed negligible inhibition levels of <8%, <16%
2
3
of 133. Intermolecular H bonds of 133 are shown in red.
and <26% for CatB, CatK and CatL, respectively, equaling a
selectivity for CatS of more than 30000-fold (Supp. Info Tab.
3). This exquisite selectivity for CatS holds for all of our
potent S3-pocket amides. Given the surface-exposed
positioning of the S3-substituent of 133 and the similar
potencies of the other variants, we conclude that this
outstanding CatS selectivity stems from the fragment-
derived chemical matter in the S2-pocket, where CatS is
most divergent from other cathepsin family members.
To investigate the role of rotational freedom next to the
covalent bond with CatS, we also explored S3-pocket
amides with the N-(cyanomethyl) warhead, for which we
synthesized 33 examples with (R) stereochemistry (Supp.
Info Tab. 3B). The Ki values of matched-pair N-(1-
cyanocyclopropyl) versus N-(cyanomethyl) warhead
inhibitors generally vary by less than a 2-fold in both
directions, as exemplified by the N-(1-cyanocyclopropyl)
48 (Fig. 6) and its N-(cyanomethyl) matched analogue 96
with Ki values of 200 and 220 pM, respectively. The 4-fold
difference between the Ki values of 54 and 133 is rather an
exception (Fig. 5). This indicates that the exchange of
cyclopropyl for methyl in the S1-pocket of CatS does not
alter potency much.
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
The best established functional target engagement assay
for CatS inhibitors is monitoring the accumulation of the
CatS substrate Lip10 in human cells or in peripheral blood
mononuclear cells (PBMCs) of patients ex vivo.6,2 As
representatives we tested 44, 45 and 48 for Lip10
accumulation activity in human Raji cells, which is an
antigen presenting human B-lymphocyte cell line. 44, 45
and 48 showed cellular EC50 values of 580, 1550 and 1100
nM, respectively, which is more than 5000-fold less potent
than their in vitro enzymatic Ki values of 0.09, 0.18 and 0.20
nM, respectively. With polar surface areas (PSA) between
144 and 172 Å2, we wondered whether poor cell membrane
permeability could cause this harsh drop in cellular activity.
Indeed our in-house Caco-2 assay revealed that 45 and 48
have apparent apical to basolateral (A-B) cell permeabilities
of merely 0.3 and 0.4 *10-6 cm/sec, which is characteristic
of poorly cell permeable substances.25
Given the ease of automated amide bond coupling
synthesis, we additionally synthesized 36 N-(cyanomethyl)
S3-pocket variants with (S) stereochemistry. For many
matched-pair variants, potency did not drop much between
(R) and (S) stereochemistry, as evidenced by the Ki values
of 220 and 580 pM for the (R) 96 and the (S) 116,
respectively.
O
O
S
HN
N
N
H
N
N
N
Scheme 1. Synthesis of 48
44
O
O
Ki = 0.09 nM
Cbz
O
NH
R
O
NH2
b
S
S
OH
S
OH
a
R
R
HO
Bn
S
HO
S
R
O
HN
O
NH2
O
Cbz
48-1
Cbz
48-2
O
S
HN
N
O
S
N
H
N
N
N
O
R
NH
O O
Cl
S
O
c
Bn
d
R
48
O
O
S
R
Bn
O
O
Ki = 0.20 nM
HN
O
N
HN
Cbz
Cbz
N
N
48-3
48-4
O
O
S
N
O O
R
S
N
N
O O
N
O
S
HN
Bn
e
f
N
N
R
O
HO
N
H
N
N
N
N
N
HN
NH2
Cbz
N
48-5
48-6
163
Ki = 0.25 nM
O
O
O
O
O
N
N
O O
S
N
O O
S
N
N
N
R
R
g
HO
N
H
h
N
NHBoc
NHBoc
48-7
48-8
O
S
N
H
N
O
O
S
N
N
N
O
O
S
N
N
O
O
169
Ki = 290 nM
i
O
R
R
N
H
N
H
N
N
NH2
S
HN
Figure 6. 2-Aryl- and 2-Alkylacetamides(e.g. 44, 48) and
arylamines (163) are favorable S3-pocket substituents with
picomolar Ki values, while benzyl (169) shows a 1000-fold
potency drop. This highlights the importance of the H-bonding
between the NH (red circles) and the carbonyl of Gly69.
O
48-9
48
Reagents and conditions: (a) Cbz-OSu, NaOH, H2O, Acetone,
0°C to r.t.; (b) BnBr, K2CO3, DMF, r.t.; (c) NCS, 2 M HCl, CH3CN,
r.t.; (d) NEt3, DCM, r.t.; (e) Pd/C, HCOOH, HCOONH4, MeOH, r.t.;
(f) (Boc)2O, NEt3, THF, r.t.; (g) HOBt, EDCI, NEt3, DMF, r.t.; (h)
TFA, DCM, r.t.; (i) HATU, DIPEA, DMF, 50°C.
Next we tested these S3-pocket amides for selectivity
over CatB, CatK and CatL. At inhibitor concentrations of 10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 10
1
2
3
4
5
6
7
8
Accordingly we investigated whether the polar S3-pocket
Conclusion
amide can be replaced while preserving the pivotal H bond
between the NH and the carbonyl of Gly69. In order not to
introduce a basic nitrogen, we synthesized the (pyridin-3-
yl)amino analogue 163, which retains a potent Ki value of
0.25 nM and has a PSA of 140 Å2, which is 32 Å2 less than 48.
But disappointingly, 163 showed a Caco-2 A-B permeability
of 0.33 *10-6 cm/sec, which is no improvement over 45 and
48.
We started our inhibitor discovery with the fragment
screening hit 1, which binds to a hotspot of the CatS
substrate pocket with a Kd of merely 96 M. Further
fragment hits taught us that attractive LE values of >0.3 are
achievable in this pocket. By 3D crystal structure-assisted
merging, we efficiently elaborated 1 into >50,000-fold
selective, human cell active inhibitors with Ki values as low
as 90 pM and excellent LE values of 0.42 kcal/(#HA). Our
fragment-screen and merge approach is applicable to many
other targets and may help making drug discovery more
economical.
9
Dropping the polar S3-pocket -NH- entirely, which
reduces the PSA to 115 Å2, abrogates potency, as learned
from substances 165, 167 and 169 with a methyl, phenyl
and benzyl in this position, respectively (Supp. Info Tab.
3C). We conclude that both amido and amino aryl or alkyl
substituents are favored in the S3 pocket, but more
synthetic effort will be required to increase the cell
membrane permeability of our inhibitor series. An
interesting alternative to the S3-pocket -NH- is (2S,4R)-1-
acetyl-4-(2-chlorophenyl)sulfonyl-N-(1-
cyanocyclopropyl)pyrrolidine-2-carboxamide published by
Hilpert et al.,24 which places a di-substituted amide nitrogen
in this very position, which might be protonated when
bound to CatS. This proline scaffold, which reduces the H
bond donor count by 1 and rigidifies the molecule, inhibits
human CatS with an IC50 of 1 nM and showed an EC50 of 5
mg/kg in a transgenic DO10.11 mouse model of antigen
presentation.24
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
Experimental Section
Substance purity was analyzed by LC-MS and 1H-NMR and
was ≥95% for all substances.
Automated Parallel Amide Bond Coupling Synthesis.
In a dry vial the amine derivative (0.5 ml, 100 µM, 0.2 M in
DMF), the acid derivative (1.0 ml, 400 µmol, 0.4 M in DMF),
HATU (0.5 ml, 125 µmol, 0.25
M in DMF) and
diisopropylethylamine (0.087 ml, 500 µmol) were added at
r.t. and then incubated at 50°C o/n. After cooling to r.t. the
solvent was evaporated and DCM (3 ml) and sodium
carbonate (2 ml, 1 M in water) were added. The mixture was
shaken for 30 min at r.t. The organic layer was separated
and the aqueous layer was washed with DCM (2 ml) twice.
The solvent was evaporated and the product was purified
by HPLC.
Scheme 2. Synthesis of 169 and 170
Synthesis
of
(2R)-N-(1-cyanocyclopropyl)-3-((4-
O
O
O
O
O
(pyridin-2-yl)piperazin-1-yl)sulfonyl)-2-(2-(thiophen-2-
yl)acetamido)propanamide (48). NaOH (4N, 250 ml) was
dropwise added to (2R, 2'R)-3,3'-disulfanediylbis(2-
amino)propanoic acid (40 g, 0.167 mol) in acetone (400 ml)
under N2 atmosphere with stirring at 0°C. After 10 min Cbz-
OSu (100 g, 0.40 mol) was added in portions at 0°C and then
stirred o/n at r.t. The solvent was removed, the pH was
adjusted to 11 with NaOH (aq), and the product was
extracted with ether (3 x 100 ml). Then the pH was adjusted
to 2-3 with HCl (2N) and the solid was collected and dried
under infrared light to yield the Cbz-protected product.
Next the Cbz-protected product (50 g, 0.098 mol), K2CO3 (68
g, 0.49 mol), BnBr (67 g, 0.39 mol) and DMF (600 ml), were
stirred o/n at r.t., poured into 1 l of crushed ice and
extracted with EtOAc (3 x 1 l). The organic layer was washed
with brine (3 x 500 ml), dried over anhydrous sodium
sulfate, concentrated under vacuum and purified by silica
gel column chromatography. Next it was mixed with NCS
(11.5 g, 86.4 mmol), 2 N HCl (20 ml dropwise slowly),
CH3CN (100 ml) and stirred for 10 min at r.t. Afterwards
benzyl ester (10 g, 14.5 mmol in CH3CN) was dropwise
added with stirring at r.t., stirred for another 10 min and
then diluted with EtOAc (250 ml). The organic layer was
dried over anhydrous sodium sulfate and concentrated
under vacuum. Next sulfonyl chloride (20 g, 29.0 mmol,
crude) and DCM (700 ml) were added to the pellet under N2
atmosphere, followed by dropwise addition of 2-pyridino-
piperazine (6 g, 36.8 mmol), Et3N (5.87 g, 58.1 mmol) at 0°C.
The mixture was stirred o/n at r.t., concentrated under
vacuum and purified by silica gel chromatography. The
MeO
OMe
HO
OH
OH
b
a
c
169-2
169-1
169-3
O
O
d
OEt
SAc
e
OEt
169-4
169-5
O
O
O
O
S
H
N
S
N
OEt
N
N
f
g
N
N
O
N
N
N
O
O
169-6
169-7
O
O
O
O
S
H
N
H
N
N
S
N
(S)
(R)
N
N
N
N
N
O
170
169
Reagents and conditions: (a) NaOH, MeOH, H2O, 100°C 4h.
(b) Formaline, Et2NH, reflux 16h. (c) SOCl2, EtOH, r.t. 16h. (d)
thioacetic acid, benzene, 70°C 16h. (e) (1) NCS, HCl, 20°C 4h (2)
1-(pyridin-2-yl) piperazine, TEA, -15°C 20 min. (f) 1-
aminocyclopropanecarbonitrile hydrochloride, TEA, trimethyl
aluminum, 110°C 8h. (g) chiral HPLC.
ACS Paragon Plus Environment

Page 7 of 10
1
Journal of Medicinal Chemistry
protected sulfone amide (16 g, 29.7 mmol), HCOOH (6.8 g,
mixture was refluxed for 16 h. The reaction mixture was
2
147 mmol), ammonium formate (9.2 g, 147 mmol), Pd/C
(1.6 g) and MeOH (300 ml) were stirred for 1h at r.t. The
solid was filtered out and the filtrate was concentrated
under vacuum and the residue was purified by reversed
phase flash chromatography. The collect aqueous layer was
dried by lyophilization to give the amino acid. The amino
acid (10g, 31.8 mmol), THF (200 ml), Boc2O (10.3 g, 47.6
mmol), Et3N (6.4 g, 63.3 mmol) were stirred o/n at r.t., the
solvent evaporated and the pellet was purified by silica gel
chromatography.
cooled, concentrated under reduced pressure. The left
behind viscous mass was acidified (pH~4) with 2N HCl and
extracted with dichloromethane (3 X 20 mL). Combined
organic layers were washed with brine solution, dried over
Na2SO4 and evaporated to get 2-benzylacrylic acid (169-3)
(12 g, 85%) as a pale yellow liquid.
3
4
5
6
7
8
9
LC-MS: m/z [M+H]+ calc.: 163.1; found: 163.1
1H NMR (400 MHz, DMSO-d6): δ 7.32 -7.27 (m, 2H), 7.26
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
-7.20 (m, 3H), 6.37 (s, 1H), 5.58 (s, 1H), 3.63 (s, 2H).
Step 3: To a stirred solution of compound 169-3 (2 g,
12.33 mmol, 1.0 eq) in ethanol (20 mL) SOCl2(2.8 mL, 37.28
mmol, 3.0 eq) was added drop wise at 0-50C and stirred at
rt for 16h. The reaction mixture was cooled and
The Boc-protected amino acid (8 g, 19.3 mmol), HOBT
(3.65 g, 27.0 mmol), EDCI (5.14 g, 27.0 mmol), the nitrile
hydrochloride (2.7 g, 22.8 mmol), Et3N (5.07 g, 50.1 mmol)
and DMF (150 ml) were stirred o/n at r.t., poured into 200
g of crushed ice and extracted with EtOAc (3 x 150 ml). The
combined organic layers were washed with brine (3 x 100
ml), dried over anhydrous sodium sulfate, concentrated
under vacuum and purified by silica gel chromatography.
The Boc-protected product (5 g, 10.4 mmol), TFA (30 ml)
and DCM (70 ml) were stirred o/n at r.t., concentrated
under vacuum. Then 20 ml of water were added, the pH
adjust to 7-8 with NaHCO3, and the product was purified by
reversed phase Flash chromatography. Amide bond
coupling was performed as described above.
concentrated,
basified
with
saturated
NaHCO3
solution(pH~8) and extracted with EtOAc (3 X 10 mL).
Combined organic layers were washed with brine solution,
dried over Na2SO4 and evaporated to get ethyl 2-
benzylacrylate (169-4) (2 g, 86%) as a pale yellow liquid.
LC-MS: m/z [M+H]+ calc.: 191.1; found: 191.1
1H NMR (400 MHz, DMSO-d6): δ 7.31 - 7.26 (m, 2H), 7.22
- 7.19 (m, 3H), 6.23 - 6.22 (m, 1H), 5.45 - 5.44 (m, 1H), 4.21
- 4.15 (m, 2H), 3.63 (s, 2H), 1.34 (t, J = 7.2 Hz, 3H).
Step 4: To a stirred solution of compound 169-4 (12 g,
63.15 mmol, 1.0 eq) in benzene (120 mL) was added
thioacetic acid (12 mL) and stirred at 70oC for 16h. The
reaction mixture was cooled, solvent was distilled off under
reduced pressure. Obtained crude was purified by column
chromatography over silica gel (100-200 mesh size) using
20 % EtOAc in pet ether as eluent to get ethyl 3-(acetylthio)-
2-benzylpropanoate (169-5) (6.4 g, ~40%) as a pale yellow
liquid.
Analytical Data: 48-2: white solid; LC-MS m/z [M+H]+
calculated: 509.1; found: 509.2. 48-3: light yellow oil; LC-
MS m/z [M+H]+ calc.: 689.2; found: 689.2. 48-4: white solid;
LC-MS m/z [M+H]+ calc.: 411.1; found: 411.0. 48-5: white
solide; LC-MS m/z [M+H]+ calc.: 539.2; found: 539.2. 48-6:
white solid; LC-MS m/z [M+H]+ calc.: 315.1; found: 315.0.
48-7: yellow oil; LC-MS m/z [M+H]+ calc.: 415.2; found:
415.2. 48-8: white solid; LC-MS m/z [M+H]+ calc.: 479.2;
found: 479.2. 48-9: white solid; LC-MS m/z [M+H]+
calculated: 379.1; found: 379.2. 48: white solid; LC-MS: m/z
LC-MS: m/z [M+H]+ calc.: 267.0; found: 267.1
1H NMR (400 MHz, DMSO-d6): δ 7.30 - 7.26 (m, 2H), 7.23
- 7.19 (m, 1H), 7.17 - 7.15 (m, 2H), 4.11 - 4.06 (m, 2H), 3.16
- 3.11 (m, 1H), 3.05 - 2.86 (m, 4H), 2.51 (s, 3H), 1.14 (t, J =
7.2 Hz, 3H).
1
[M+H]+ calc.: 503.2; found: 503.1 H NMR (600 MHz, DMSO-
d6, 30°C) δ 9.07 (1H, s), 8.60 (1H, d), 8.13 (1H, ddt), 7.52 –
7.6 (1H, m), 7.33 (1H, dt), 6.92 (1H, ddd), 6.90 (1H, dt), 6.86
(1H, dd), 6.65 – 6.71 (1H, m), 4.62 (1H, td), 3.63 – 3.77 (2H,
m), 3.56 (4H, t), 3.50 (1H, dd), 3.19 – 3.28 (5H, m), 1.4 – 1.52
(2H, m), 1.08 – 1.17 (2H, m).
Step 5: To a stirred solution of NCS (1.0g, 7.48 mmol, 4.0
eq) in ACN (5 mL) were added 2N HCl (5 mL) and compound
169-5 (500 mg, 1.87 mmol, 1.0 eq) in ACN (5 mL) as slowly
drop wise at 0-50C and stirred at 20oC for 1h. The reaction
mixture was extracted with dichloromethane (3 X 10 mL).
Combined organic layers were washed with brine solution,
dried over Na2SO4 and evaporated under reduced pressure
left behind viscous mass was dissolved in dichloromethane
(10 mL), were added 1-(pyridin-2-yl) piperazine (0.168g,
1.034 mmol, 1.25eq), TEA (0.2 mL, 1.40 mmol, 1.7 eq) at -
150C and stirred for 20 min. The reaction mixture was
extracted with dichloromethane (3 X 10 mL), combined
organic layers were washed with brine solution, dried over
Na2SO4 and evaporated to get crude compound. Obtained
crude compound was purified by using column
chromatography over silica gel (100-200 mesh size) using 5
% methanol in dichloromethane as eluent to get ethyl-2-
benzyl-3-((4-(pyridin-2-yl)piperazin-1-
Synthesis of (2R)-2-benzyl-N-(1-cyanocyclopropyl)-3-
((4-(pyridin-2-yl)piperazin-1-yl)sulfonyl)propanamide
(169) and its enantiomer 170. Step 1: To a stirred solution
of dimethyl-2-benzylmalonate (169-1) (10.0 g, 39.95 mmol,
1.0 eq) in methanol (20 mL), water (20 mL) and NaOH (3.5
g, 87.89 mmol, 2.2 eq) were added at r.t. and heated to reflux
for 4 h. The reaction mixture was cooled and concentrated
to get crude. Obtained crude reaction mixture was acidified
(pH~4) with 2N HCl and extracted with EtOAc (3 X 20 mL).
The combined organic layers were washed with brine
solution, dried over Na2SO4 and evaporated to get 2-
benzylmalonic acid (169-2) (7 g, 90%) as a white solid.
LC-MS: m/z [M-H]- calc.: 193.1; found: 193.1
1H NMR (400 MHz, DMSO-d6): δ 12.76 (brs, 2H), 7.29-
7.17 (m, 5H), 3.57 (t, J = 7.6 Hz, 1H), 3.03 (d, J = 7.6 Hz, 2H).
yl)sulfonyl)propanoate (169-6) (100 mg, ~25% over 2
steps) as a pale yellow solid.
LC-MS: m/z [M+H]+ calc.: 418.2; found: 418.2
Step 2: To a stirred solution of compound 169-2 (17 g,
87.5 mmol, 1.0 eq) in formaline (36 mL) diethyl amine (8.5
mL) was added drop wise at 0-50C. The resulting reaction
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 10
1
2
3
4
5
6
7
8
1H NMR (400 MHz, CDCl3): δ 8.20 - 8.19 (m, 1H), 7.51 -
7.50 (m, 1H), 7.30 - 7.26 (m, 2H), 7.25 - 7.23 (m, 1H), 7.17 -
7.15 (m, 2H), 6.69 - 6.62 (m, 2H), 4.13 - 4.08 (m, 2H), 3.60 -
3.56 (m, 4H), 3.50 - 3.44 (m, 1H), 3.30 - 3.24 (m, 5H), 3.05 -
3.03 (m, 1H), 2.96 - 2.89 (m, 2H), 1.16 (t, J =7.2 Hz, 3H).
† AstraZeneca, Chesterford Science Park, Cambridge, CB10
1XL, UK. Email: Markus.Schade1@astrazeneca.com
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Step 6: To a stirred solution of compound 169-6 (0.5 g,
1.19 mmol, 1.0 eq) in toluene (30 mL) were added 1-
aminocyclopropanecarbonitrile hydrochloride (0.168g,
1.034 mmol, 1.25eq), TEA (0.2 mL, 1.40 mmol, 1.2 eq) and
trimethyl aluminum (1.2 mL, 2.398 mmol, 2.0eq) (Note: For
every 2h again added 1.0 eq of amine and 1.0 eq of trimethyl
aluminum due to non-consumption of SM) at r.t. and heated
to 110oC for 8h. The reaction mixture was quenched with
NH4Cl solution and extracted with EtOAc (3 X 10 mL). The
combined organic layers were washed with brine solution,
dried over Na2SO4 and evaporated under reduced
pressure. The obtained crude product was purified by
preparative HPLC to get 2-benzyl-N-(1-cyanocyclopropyl)-
3-((4-(pyridin-2-yl)piperazin-1-yl)sulfonyl)propenamide
(169-7) as a white solid.
Funding Sources
This work was funded by Grünenthal GmbH.
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
ACKNOWLEDGMENT
The authors thank Florian Jakob for institutional and scientific
support, and Stefanie Peters and Utz Jagusch for LogD
measurements.
ABBREVIATIONS
Cat, cathepsin; LE, ligand efficiency; o/n, overnight; PSA, polar
surface area; r.t., room temperature; SAR, structure-activity
relationship
LC-MS: m/z [M+H]+ calc.: 454.2; found: 454.0
REFERENCES
Step 7: Compound 169-7 was purified by preparative
normal phase chiral HPLC to get substances 169 and 170.
(1) Payne, C.D.; Deeg, M.A.; Chan, M.; Tan, L.H.; LaBell, E.S.; Shen, T.;
DeBrota, D.J. Pharmacokinetics and pharmacodynamics of the
cathepsin S inhibitor, LY3000328, in healthy subjects. Br. J. Clin.
Pharmacol. 2014, 78, 1334-1342.
(2) Hargreaves, P.; Daoudlarian, D.; Theron, M.; Kolb, F.A.;
Manchester Young, M.; Reis, B.; Tiaden, A.; Bannert, B.; Kyburz, D.;
Manigold, T. Differential effects of specific cathepsin S inhibition in
biocompartments from patients with primary Sjögren syndrome.
Arthritis Res. Ther. 2019, 21, 175.
(3) Kramer L, Turk D, Turk B. The future of cysteine cathepsins in
disease management. Trends Pharmacol. Sci. 2017, 38, 873-898.
(4) Wilkinson RD, Williams R, Scott CJ, Burden RE. Cathepsin S:
therapeutic, diagnostic, and prognostic potential. Biol. Chem. 2015,
396, 867–882.
(5) Tato, M.; Kumar, S.V.; Liu, Y.; Mulay, S.R.; Moll, S.; Popper, B.;
Eberhard, J.N.; Thomasova, D.; Rufer, A.C.; Gruner, S.; Haap, W.;
Hartmann, G.; Anders, H.J. Cathepsin S inhibition combines control
of systemic and peripheral pathomechanisms of autoimmune
tissue injury. Sci. Rep. 2017, 7, 2775.
(6) Hamm-Alvarez, S.F.; Janga, S.R.; Edman, M.C.; Madrigal, S.; Shah,
M.; Frousiakis, S.E.; Renduchintala, K.; Zhu, J.; Bricel, S.; Silka, K.;
Bach, D.; Heur, M.; Christianakis, S.; Arkfeld, D.G.; Irvine, J.; Mack,
W.J.; Stohl, W. Tear cathepsin S - a candidate biomarker for
Sjögren's syndrome. Arthritis Rheumatol. 2014, 66, 1872–1881.
(7) Theron, M.; Bentley, D.; Nagel, S.; Manchester, M.; Gerg, M.;
Schindler, T.; Silva, A.; Ecabert, B.; Teixeira, P.; Perret, C.; Reis, B.
Pharmacodynamic monitoring of RO5459072, a small molecule
inhibitor of cathepsin S. Front Immunol. 2017, 8, 806.
(8) Katunuma, N.; Sekiya, T. Structure-based development of
specific inhibitors for individual cathepsins and their medical
applications. Proc. Japan Acad. Ser. B, Phys. Biol. Sci. 2011, 87, 29-
39.
(9) Moss, N.; Xiong, Z.; Burke, M.; Cogan, D.; Gao, D.A.; Haverty, K.;
Heim-Riether, A.; Hickey, E.R.; Nagaraja, R.; Netherton, M.; O’Shea,
K.; Ramsden, P.; Schwartz, R.; Shih, D.; Ward, Y.; Young, E.; Zhang,
Q. Exploration of cathepsin S inhibitors characterized by a triazole
P1-P2 amide replacement. Bioorg. Med. Chem. Lett. 2012, 22, 7189-
7193.
(10) Tber, Z.; Wartenberg, M.; Jacques, J.; Roy, V.; Lecaille, F.;
Warszycki, D.; Bojarski, A.J.; Lalmanach, G.; Agrofoglio, L.A.
Selective inhibition of human cathepsin S by 2,4,6-trisubstituted
1,3,5-triazine analogs. Bioorg. Med. Chem. 2018, 26, 4310-4319.
Substance 169. LC-MS: m/z [M+H]+ calc.: 454.2; found:
454.2 1H NMR (500 MHz, DMSO-d6): δ 8.83 (s, 1H), 8.13-
8.12 (m, 1H), 7.56 (t, J = 8.0 Hz, 1H), 7.27-7.24 (m, 2H), 7.21-
7.15 (m, 3H), 6.86 (d, J = 8.5 Hz, 1H), 6.68 (t, J = 6.5 Hz, 1H),
3.55 - 3.54 (m, 4H), 3.47 - 3.42 (m, 1H), 3.40 - 3.22 (m, 4H),
3.04 - 3.01 (m, 1H), 2.89 - 2.87 (m, 1H), 2.79 - 2.76 (m, 2H),
1.36-1.33 (m, 2H), 0.89-0.81 (m, 2H).
Substance 170. LC-MS: m/z [M+H]+ calc.: 454.1; found:
454.2. 1H NMR (500 MHz, DMSO-d6): δ 8.83 (s, 1H), 8.13-
8.12 (m, 1H), 7.56 (t, J = 8.0 Hz, 1H), 7.27-7.24 (m, 2H), 7.21-
7.15 (m, 3H), 6.86 (d, J = 8.5 Hz, 1H), 6.68 (t, J = 6.5 Hz, 1H),
3.55-3.54 (m, 4H), 3.47-3.42 (m, 1H), 3.40-3.22 (m, 4H),
3.04-3.01 (m, 1H), 2.89-2.87 (m, 1H), 2.82-2.73 (m, 2H),
1.36-1.33 (m, 2H), 0.89-0.81 (m, 2H).
ASSOCIATED CONTENT
Supporting Information. NMR binding scores, NMR-detected
dose-response curves, enzyme inhibition values, Ki and EC50
values, crystal structure determination details, procedures for
enzymatic and cellular assays, chemical synthesis, analytical
characterization and Molecular Formula Strings of all
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
PDB codes. Atomic coordinates for the X-ray structures of
compounds 1, 2, 3, 37 and 133 (PDB codes 6YYO, 6YYP, 6YYQ,
6YYN and 6YYR, respectively) are available from the RCSB
Protein Data Bank (www.rcsb.org). Authors will release the
atomic coordinates and experimental data upon article
publication.
AUTHOR INFORMATION
Corresponding Author
* M. S.: e-mail, Markus.Schade@grunenthal.com
Present Addresses
ACS Paragon Plus Environment

Page 9 of 10
1
Journal of Medicinal Chemistry
(11) Plebanek, E.; Chevrier, F.; Roy, V.; Garenne, T.; Lecaille, F.;
Bruin, J., Cameron, K.S., Cumming, I., Hamilton, W., Jones, P.S.,
2
3
4
5
Warszycki, D.; Bojarski, A.J.; Lalmanach, G.; Agrofoglio, L.A.
Straightforward synthesis of 2,4,6-trisubstituted 1,3,5-triazine
compounds targeting cysteine cathepsins K and S. Eur. J. Med.
Chem. 2016, 121, 12-20.
Kaptein, A., Kinghorn, E., Maidment, M., Martin, I., Mitchell, A.,
Rankovic, Z., Robinson, J., Scullion, P., Uitdehaag, J.C., Vink, P.,
Westwood, P., van Zeeland, M., van Berkom, L., Bastiani, M.,
Meulemans, T. 6-Phenyl-1H-imidazo[4,5-c]pyridine-4-carbonitrile
as cathepsin S inhibitors. Bioorg. Med. Chem. Lett. 2010, 20, 4350-
4354.
(19) Cai, J.; Bennett, D.J.; Rankovic, Z.; Dempster, M.; Fradera, X.;
Gillespie, J.; Cumming, I.; Finlay, W.; Baugh, M.; Boucharens, S.;
Bruin, J.; Cameron, K.S.; Hamilton, W.; Kerr, J.; Kinghorn, E.;
McGarry, G.; Robinson, J.; Scullion, P.; Uitdehaag, J.C.; van Zeeland,
M.; Potin, D.; Saniere, L.; Fouquet, A.; Chevallier, F.; Deronzier, H.;
Dorleans, C.; Nicolai, E. 2-Phenyl-9H-purine-6-carbonitrile
derivatives as selective cathepsin S inhibitors. Bioorg. Med. Chem.
Lett. 2010, 20, 4447-4450.
(12) McClung, M.R.; O'Donoghue, M.L.; Papapoulos, S.E.; Bone, H.;
Langdahl, B.; Saag, K.G.; Reid, I.R.; Kiel, D.P.; Cavallari, I.; Bonaca,
M.P.; Wiviott, S.D.; de Villiers, T.; Ling, X.; Lippuner, K.; Nakamura,
T.; Reginster, J.Y.; Rodriguez-Portales, J.A.; Roux, C.; Zanchetta, J.;
Zerbini, C.A.F.; Park, J.G.; Im, K.; Cange, A.; Grip, L.T.; Heyden, N.;
DaSilva, C.; Cohn, D.; Massaad, R.; Scott, B.B.; Verbruggen, N.;
Gurner, D.; Miller, D.L.; Blair, M.L.; Polis, A.B.; Stoch, S.A.; Santora,
A.; Lombardi, A.; Leung, A.T.; Kaufman, K.D.; Sabatine, M.S.;
Odanacatib for the treatment of postmenopausal osteoporosis:
results of the LOFT multicenter, randomized, double-blind,
placebo-controlled trial and LOFT Extension study. Lancet Diabetes
Endocrinol. 2019, 7, 899-911.
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
(20) Rankovic, Z.; Cai, J.; Fradera, X.; Dempster, M.; Mistry, A.;
Mitchell, A.; Long, C.; Hamilton, E.; King, A.; Boucharens, S.;
Jamieson, C.; Gillespie, J.; Cumming, I.; Uitdehaag, J.; van Zeeland, M.
Dioxo-triazines as a novel series of cathepsin K inhibitors. Bioorg.
Med. Chem. Lett. 2010, 20, 1488-1490.
(13) Falgueyret, J.; Desmarais, S.; Oballa, R.; Black, W.C.; Cromlish,
W.; Khougaz, K.; Lamontagne, S.; Massé, F.; Riendeau, D.;
Toulmond, S.; Percival, M.D. Lysosomotropism of basic cathepsin K
inhibitors contributes to increased cellular potencies against off-
target cathepsins and reduced functional selectivity. J. Med. Chem.
2005, 48, 7535-7543.
(14) Borišek, J.; Vizovišek, M.; Sosnowski, P.; Turk, B.; Turk, D.;
Mohar, B.; Novič, M. Development of N-(functionalized benzoyl)-
homocycloleucyl-glycinonitriles as potent cathepsin K inhibitors. J.
Med. Chem. 2015, 58, 6928-6937.
(21) Mossman, C.J. Sulfonyl Containing Compounds as Cysteine
Protease Inhibitors. WO Pat. Appl. 102535 A2, 2006.
(22) Gauthier, J.Y.; Black, W.C.; Courchesne, I.; Cromlish, W.;
Desmarais, S.; Houle, R.; Lamontagne, S.; Li, C.S.; Massé, F.; McKay,
D.J.; Ouellet, M.; Robichaud, J.; Truchon, J.F.; Truong, V.L.; Wang, Q.;
Percival, M.D. The identification of potent, selective, and
bioavailable cathepsin S inhibitors. Bioorg. Med. Chem. Lett. 2007,
17, 4929-4933.
(15) Schade, M. NMR fragment screening: advantages and
applications. IDrugs 2006, 9, 110-113.
(23) https://pubchem.ncbi.nlm.nih.gov/compound/Petesicatib
(16) Wei, J.; Pio, B.A.; Cai, H.; Meduna, S.P.; Sun, S.; Gu, Y.; Jiang, W.;
Thurmond, R.L.; Karlsson, L.; Edwards, J.P. Pyrazole-based
cathepsin S inhibitors with improved cellular potency. Bioorg. Med.
Chem. Lett. 2007, 17, 5525-5528.
(17) Tully, D.C.; Liu, H.; Chatterjee, A.K.; Alper, P.B.; Williams, J.A.;
Roberts, M.J.; Mutnick, D.; Woodmansee, D.H.; Hollenbeck, T.;
Gordon, P.; Chang, J.; Tuntland, T.; Tumanut, C.; Li, J.; Harris, J.L.;
Karanewsky, D.S. Arylaminoethyl carbamates as a novel series of
potent and selective cathepsin S inhibitors. Bioorg. Med. Chem. Lett.
2006, 16, 5107-5111.
(accessed Jul 1, 2020).
(24) Hilpert, H.; Mauser, H.; Humm, R.; Anselm, L.; Kuehne, H.;
Hartmann, G.; Gruener, S.; Banner, D.W.; Benz, J.; Gsell, B.;
Kuglstatter, A.; Stihle, M.; Thoma, R.; Sanchez, R.A.; Iding, H.; Wirz,
B.; Haap, W. Identification of potent and selective cathepsin S
inhibitors containing different central cyclic scaffolds. J. Med. Chem.
2013, 56, 9789-9801.
(25) Tavelin, S.; Taipalensuu, J.; Söderberg, L.; Morrison, R.; Chong,
S.; Artursson, P. Prediction of the oral absorption of low-
permeability drugs using small intestine-like 2/4/A1 cell
monolayers. Pharm Res. 2003, 20, 397-405.
(18) Cai, J., Baugh, M., Black, D., Long, C., Jonathan Bennett, D.,
Dempster, M., Fradera, X., Gillespie, J., Andrews, F., Boucharens, S.,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 10
Insert Table of Contents artwork here
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
10
ACS Paragon Plus Environment