Structure–activity relationship studies of 3-substituted pyrazoles as novel allosteric inhibitors of MALT1 protease

Article abstract of DOI:10.1016/j.bmcl.2021.127996

We report the discovery of a novel series of 1,5-bisphenylpyrazoles as potent MALT1 inhibitors. Structure–activity relationship exploration of a hit compound led to a potent MALT1 inhibitor. Compound 33 showed strong activity against MALT1 (IC₅₀

Full text of DOI:10.1016/j.bmcl.2021.127996

Bioorg. Med. Chem. Lett. 41 (2021) 127996
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure–activity relationship studies of 3-substituted pyrazoles as novel
allosteric inhibitors of MALT1 protease
Ken Nunettsu Asaba*, Yohei Adachi, Kazuyuki Tokumaru*, Akira Watanabe, Yasufumi Goto,
Takumi Aoki
Pharmaceutical Research Laboratories, Toray Industries, Inc., 6-10-1 Tebiro, Kamakura, Kanagawa 248-8555, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
We report the discovery of a novel series of 1,5-bisphenylpyrazoles as potent MALT1 inhibitors. Structur-
MALT1
e–activity relationship exploration of a hit compound led to a potent MALT1 inhibitor. Compound 33 showed
Paracaspase
strong activity against MALT1 (IC50: 0.49 μM), potent cellular activity (NF-κB inhibition and inhibition of IL2
Allosteric inhibitors
Structure activity relationship
production), and high selectivity against caspase-3, -8, and -9. The results of a kinetics study suggest that
compound 33 is a non-competitive inhibitor of MALT1 protein.
The nuclear factor-κB (NF-κB) family of transcription factors is
considered to play a central role in the immune systems, including
immune-cell development, homeostasis, survival, and function.1–4 NF-
κB is implicated in the pathogenesis of several autoimmune diseases,
such as rheumatoid arthritis, systemic lupus erythematosus, type I dia-
betes, multiple sclerosis, and inflammatory bowel disease.5–11 Although
NF-κB is activated transiently during the normal immune response, it is
chronically activated in tissues affected by autoimmune diseases and
induces excessive inflammatory cytokines and chemokines, which leads
to autoimmunity.
inhibiting proteolytic cleavage of the substrate (Table 1).17,18 None of
these phenothiazine derivatives significantly inhibits either caspase 3 or
caspase 8, which are structurally the closest relatives of MALT1 in
mammals.
We conducted molecular docking simulations with MALT1 to iden-
tify selective small inhibitors of MALT1. The model was constructed
from the crystal structure of MALT1 complexed with the allosteric in-
hibitor thioridazine (PDB ID: 4I1R), and we used thioridazine (1) as the
template compound for docking. Enzymatic biochemical assay of the top
ranked compounds identified compound 2, which showed weak activity
against MALT1 (Table 1). The calculated docking pose of 2 indicated
that the 1,5-bisphenylpyrazole moiety is located similarly to the
Mucosa-associated lymphoid tissue lymphoma translocation protein
1 (MALT1) is related to the activation of NF-κB signaling and the
negative feedback loop required for this signaling.12–16 MALT1 acts as a
scaffold protein, resulting in protein–protein interactions which lead to
the activation of the IκB kinase complex and subsequent activation of
NF-κB. In addition to this function, MALT1 is a protease. The para-
caspase (cysteine protease) domain of MALT1 cleaves RelB, CYLD, and
A20, all of which function as negative regulators of canonical NF-κB
signaling. Therefore, MALT1 is an important therapeutic target for the
treatment of immunomodulation disorders and lymphoma. In this study,
we describe the discovery and structure–activity relationship (SAR)
exploration of a series of allosteric small-molecule inhibitors of MALT1
protease.
Table 1
In vitro activity profiles of thioridazine (1) and compound 2.
2
Thioridazine (1)
Phenothiazine derivatives, including mepazine, thioridazine (1), and
promazine, target the allosteric site of MALT1, preventing rearrange-
ment of the protein from the inactive state to the activated state and
MALT1 IC50: 3.43
μM
MALT1 IC50: 19.79 μM
* Corresponding author.
E-mail addresses: kennunettsu.asaba.u8@mail.toray (K.N. Asaba), kazuyuki.tokumaru.n6@mail.toray (K. Tokumaru).
https://doi.org/10.1016/j.bmcl.2021.127996
Received 18 January 2021; Received in revised form 15 March 2021; Accepted 20 March 2021
Available online 26 March 2021
0960-894X/© 2021 Elsevier Ltd. All rights reserved.

K.N. Asaba et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 127996
to give ethyl 1H-pyrazole-3-carboxmides (III). Compounds III were
hydrolyzed to 1H-pyrazole-3-carboxylic acids (IV) using aqueous so-
dium hydroxide solution in ethanol. Target compounds V were synthe-
sized by the condensation of compounds IV with amines using amide
coupling reagents such as carbonyldiimidazole and 1-[bis(dimethyla-
mino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxide hexa-
fluorophosphate (HATU).
Substituents at the 5-position can be introduced by different
methods, as shown in Scheme 2. 1-(4-Chlorophenyl)-5-hydroxy-1H-
pyrazole-3-carboxylic acid (VIII) was synthesized using 4-chlorophenyl-
hydrazine hydrochloride and dimethyl acetylenedicarboxylate. Next,
VIII was converted to 1-(4-chlorophenyl)-3-(methylcarbamoyl)-1H-
pyrazol-5-yl trifluoromethanesulfonate ester (X) via amidation and
sulfonylation; then, X was coupled with boronic acid derivatives to
obtain 1,5-bissubstituted pyrazole-3-carboxamides (V).
We first introduced various substituents at the 1-position of 1H-
Table 2
Fig. 1. Docking pose of compound 2 (green) in the crystal structure of MALT1
bound with thioridazine (1, purple) (PDB ID: 4I1R).
SAR for the 1-position of 1H-pyrazoles.
phenothiazine moiety of thioridazine (1) bound in the hydrophobic
pocket of MALT1 (Fig. 1). The pose also showed that a hydrogen bond is
formed between the N-methylamide nitrogen of 2 and glutamic acid
E397, again similar to the crystal structure with thioridazine (1).17 We
selected this 3-substituted pyrazole compound as a starting point as it
has three sites for initial structure–activity studies, namely, two phenyl
rings and a substituted amide.
compound
R1
MALT1 inhibition
MALT1 inhibition
IC50 (µM)
at 25 µM
2
3
4
5
6
7
8
Ph
19.79
16.33
59%inh
57%inh
63%inh
7.7%inh
72%inh
57%inh
46%inh
Benzyl
Cyclohexyl
Me
not tested
>25
Scheme 1 illustrates the synthesis of 1,5-bissubstituted pyrazole-3-
carboxamides (V). Commercially available ketones (I) were condensed
with diethyl oxalate using sodium ethoxide in ethanol. The obtained
ethyl 2,4-dioxobutanoates (II) were treated with hydrazines in ethanol
4-Chlorophenyl
3-Chlorophenyl
2-Chlorophenyl
7.01
not tested
>25
Scheme 1. Synthesis of compounds V via a condensation reaction at the last step.
Scheme 2. Synthesis of compounds V via Suzuki coupling at the last step.
2

K.N. Asaba et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 127996
Table 3
Table 4
SAR for the 5-position of 1H-pyrazoles.
SAR for the 3-position of 1H-pyrazoles.
compound
R2
MALT1
inhibition
IC50 (µM)
MALT1
inhibitionat
25 µM
compound
MALT1 inhibition
IC50 (µM)
6
4-Chlorophenyl
3-Chlorophenyl
2-Chlorophenyl
Ph
7.01
>25
>25
16.79
6.55
>25
>25
>25
>25
>25
72%inh
43%inh
23%inh
61%inh
80%inh
47%inh
47%inh
19%inh
5.7%inh
38%inh
9
10
11
12
13
14
15
16
17
6
-CONHMe
-CONMe2
7.01
21.18
>25
23
24
4-Methoxyphenyl
4-Fluorophenyl
4-Trifluoromethylphenyl
4-Hydroxymethylphenyl
4-Methanesulfonylphenyl
4-
25
26
27
28
29
30
1.98
7.39
Trifluoromethoxyphenyl
4-Cyanophenyl
3-Thienyl
18
19
20
21
22
>25
>25
>25
>25
>25
31%inh
26%inh
20%inh
16%inh
21%inh
2-Furyl
3-Furyl
6.55
2-Methoxy-5-pyridyl
12.41
13.39
>25
pyrazole (Table 2). The benzyl (3) and cyclohexyl (4) derivatives
exhibited potency similar to that of compound 2, whereas methyl de-
rivative (5) was completely inactive, suggesting that large hydrophobic
substituents such as phenyl groups are required to fill the hydrophobic
pocket in MALT1. We performed chloride scanning to identify the
optimal substitution position and identified 4-chlorophenyl (6) as the
most potent of the chloride-substituted phenyls (7, 8). Substituents at
the 5-position of 1H-pyrazole were explored by fixing 4-chrolophenyl as
the 1-position substituent, then introducing substituents at the 5-posi-
tion to identify moieties more potent than phenyl (Table 3). In addi-
tion, we examined the effect of removing the chloride or changing the
substitution position of the chloride, and found that both attenuated
MALT1 inhibition (9, 10, 11). Other p-substituted phenyl groups were
explored. The potency of the 4-methoxyphenyl (12) derivative was
similar to that of 2, whereas substitution with other p-substituted phenyl
groups such as 4-fluorophenyl (13) or heteroaromatics such as 3-thienyl
(19) provided inactive compounds. In particular, compounds with polar
functional group substituents (15, 16, 20, 21, 22) had significantly less
potency compared with 2, consistent with the docking result (Fig. 1)
showing the phenyl ring located to the hydrophobic pocket of MALT1.
Next, we explored substitutes at the 3-position (Table 4). Comparing
secondary amide (6) to tertiary amides (23, 24), amide hydrogen is
essential for increased potency. This result was consistent with the
binding mode of compound 2 shown in Fig. 1, which implies that the
amide hydrogen of compound 6 interacts with E397. The potency of the
aminoethyl (25) and hydroxyethyl (27) derivatives was similar to or
higher than that of 6, whereas the aminopropyl (26), hydroxypropyl
(28), and aminocarbonylethyl (29) derivatives were less potent. The
phenylethyl (30) and cyclohexylmethyl (31) were inactive. These results
indicate that hydrophilic substituents such as amino and hydroxyl
groups are favorable substituents at the 3-position. Therefore, we next
focused on the amino groups. Introducing a methyl group into 25 pro-
vided more potent compounds (32, 33), with N,N-dimethylaminoethyl
(33) being 14-fold more potent than compound 6. The acetoamide (34)
and sulfonylamide (35) derivatives exhibited less potencies compared to
25 or 33, showing that the basicity of the amine is a key factor to
improving potency. Lengthening the carbon chain of compound 33 led
to the slightly less potent compounds 36 and 37, suggesting that optimal
31
>25
32
33
34
0.800
0.490
3.30
35
4.37
36
37
2.63
2.22
carbon chain length may be important for improving inhibitory activity.
For example, in a series of 1,5-bisphenylpyrazoles, the ethylene linker is
advantageous (33 vs. 36 vs. 37; 25 vs. 26; 27 vs. 28).
Cyclized substituents were introduced into N,N-dimethylaminoethyl
(33) (Table 5). N,N-Dimethylaminopiperidine (38) had weak potency
compared with 33 but was more potent than pyrrolidine (24) lacking an
N,N-dimethylamino moiety. Inhibitory potency against MALT1 was
enhanced by introducing amino groups at appropriate positions,
3

K.N. Asaba et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 127996
Table 5
Cyclized substituents
compound
MALT1 inhibition
IC50 (µM)
38
39
40
10.62
2.10
0.521
41
42
1.70
2.01
Fig. 2. Inhibition of MALT1-mediated CYLD cleavage by compound 33 by
western blotting. Relative Frag-CYLD/Full-CYLD values are the value relative to
the positive control (PMA/Ionomycin(+), 33(ꢀ )).
43
44
0.993
0.819
of thioridazine (1) (85%inh, 94%inh). Additionally, to confirm the
direct inhibition of MALT1 substrate cleavage, degradation of CYLD in
Jurkat cells was measured by western blotting (Fig. 2). Compound 33
dose-dependently decreased CYLD degradation product (Frag-CYLD)
induced by PMA/Ionomycin stimulation. The value of relative Frag-
CYLD/Full-CYLD ratio of 33 at 10 µM was same as that of unstimu-
lated cells without MALT1 activation.
We investigated the selectivity of 33 against caspase-3, ꢀ 8, and ꢀ 9,
which are structurally related to paracaspase, using a caspase drug
screening kit (Promokine) according to the manufacturer’s protocol
(Fig. 3). The pan-caspase inhibitor Z-VAD-fmk was used as a positive
control for each experiment. Compound 33 exhibited little activity
against caspase-3, ꢀ 8, and ꢀ 9, even at concentrations up to 100 µM,
showing it is a selective MALT1 paracaspase inhibitor.
Table 6
Cellular activities of thioridazine (1) and compounds 6 and 33.
Compounds
NF-κB inhibition IC50
Inhibition of IL2 production IC50
Thioridazine (1)
Compound 6
not tested (85%inh)
9.34 µM (25%inh)
5.41 µM (82%inh)
not tested (95%inh)
4.26 µM (54%inh)
1.79 µM (94%inh)
Compound 33
Numbers in parentheses represent the percent inhibition at 5 µM concentration.
A kinetics study was conducted to investigate the mode of action of
33 (Fig. 4). Compound 33 at a concentration around the IC50 (0.5 µM)
strongly decreased Vmax, from 35,612 to 19,979 RFU/s, while Km
remained essentially unchanged (246.9 to 234.2 µM), showing that
compound 33 is a non-competitive inhibitor. This suggests that 33 does
not bind to the paracaspase active site but rather to an allosteric pocket.
In summary, SAR exploration of hit compound 2 identified by
screening led to the discovery of 1,5-bisphenylpyrazoles as novel,
potent, and selective MALT1 inhibitors. Our exploration showed that
phenyl rings are required to maintain potency, and the secondary amide
with an aminoethyl moiety is an essential substructure for improved
potency. Derivatization of the 3-position of 1H-pyrazoles led to com-
pound 33, which exhibits cellular activity similar to that of thioridazine
(1). Investigation of the binding mode of 33 indicated a non-competitive
inhibition mechanism against MALT1 protein. Further studies on
structural optimization and identification of the binding site of this se-
ries of compounds is in progress.
suggesting that amino group interact with an amino acid of MALT1. The
potency of the N-methylpiperidin-2-ylmethyl derivative (40) was
similar to that of 33, whereas other cyclized substituents resulted in
decreased potency (39, 41). Lipophilic efficiency (LipE) is an important
metric for optimizing potency and ADME (absorption, distribution,
metabolism, and excretion) properties.19,20 To improve lipophilic effi-
ciency, an oxygen atom was introduced into 40 to provide the 4-hy-
droxy-1-methylpiperidin-2-ylmethyl derivative (43) and removal of
the methyl group from 42 to provide the morpholin-3-ylmethyl deriv-
ative (44). The LipE of 43 and 44 was improved compared to that of 33
and 6 (1.62, 1.50, 1.02, and 0.28 respectively).
The cellular activities of three compounds (thioridazine (1) and
compounds 6 and 33) were evaluated (Table 6). The inhibition of NF-κB
activity and IL2 production by compound 33 (82%inh, 94%inh) was
stronger than that of compound 6 (25%inh, 54%inh) and similar to that
4

K.N. Asaba et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 127996
Fig. 3. Selectivity of compound 33 against caspase-3, -8, and -9.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bmcl.2021.127996.
References
1 DiDonato JA, Mercurio F, Karin M. Immun Rev. 2012;246:379–400.
2 Hayden Matthew S, Ghosh Sankar. Cell Res. 2011;21:223–244.
3 Hayde MS, West AP, Ghosh S. Oncogene. 2006;25:6758–6780.
4 Liang Yan, Zhou Yang, Shen Pingping. Cell Mol Immunol. 2004;1(5):343–350.
5 Sun S-C, Chang J-H, Jin J. Trends Immunol. 2013;34(6):282–289.
6 Herrington Felicity D, Carmody Ruaidhrí J, Goodyear Carl S. J Biomol Screen. 2016;
21:223–242.
7 Pai S, Thomas R. J Autoimmun. 2008;31:245–251.
8 Greve B, Weissert R, Hamdi N, et al. J Immunol. 2007;179:179–185.
9 Han Z, Boyle DL, Manning AM, Firestein GS. Autoimmunity. 1998;28:197–208.
10 Isabelle Cleynen, Peter Jüni, GeertruidaBekkering E, et al. PLoS One. 2011;6(9),
e24106.
11 Liu Ting, Zhang Lingyun, Joo Donghyun, et al. Signal Transduct Target Ther. 2017;2:
17023.
12 Klein Theo, Fung Shan-Yu, Renner Florian, et al. Nat Commun. 2015;6:8777.
13 Rosebeck Shaun, Rehman Aasia O, Lucas Peter C, et al. Cell Cycle. 2011;10(15):
2485–2496.
14 Afonina Inna S, Elton Lynn, Carpentier Isabelle, et al. FEBS J. 2015;282(17):
3286–3297.
15 Coornaert Beatrice, Baens Mathijs, Heyninck Karen, et al. Nat Immunol. 2008;9:
263–271.
Fig. 4. Michaelis-Menten kinetics (determined by increasing the concentration
of the Ac-LRSR-AMC substrate in the absence or presence of 0.5 µM 33).
16 Hailfinger Stephan, Nogai Hendrik, Pelzer Christiane, et al. Proc Natl Acad Sci U S A.
2011;108(35):14596–14601.
17 Schlauderer Florian, Lammens Katja, Nagel Daniel, et al. Angew Chem Int Ed Engl.
2013;52:10384–10387.
Declaration of Competing Interest
18 Nagel Daniel, Spranger Stefani, Vincendeau Michelle, et al. Cancer Cell. 2012;22:
825–837.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
19 Ryckmans Thomas, Edwards Martin P, Horne Val A, et al. Bioorg Med Chem Lett.
2009;19:4406–4409.
20 Johnson Ted W, Gallego Rebecca A, Edwards Martin P. J Med Chem. 2018;61:
6401–6420.
5