Design, synthesis and discovery of 2(1H)-quinolone derivatives for the treatment of pulmonary fibrosis through inhibition of TGF-β/smad dependent and independent pathway

Article abstract of DOI:10.1016/j.ejmech.2020.112259

Idiopathic pulmonary fibrosis (IPF) is a progressive, life-threatening and interstitial lung disease with the median survival of only 3–5 years. However, due to the unclear etiology and problems in accurate diagnosis, up to now only two drugs were approved by FDA for the treatment of IPF and their outcome responses are limited. Numerous studies have shown that TGF-β is the most important cytokine in the development of pulmonary fibrosis and plays a role through its downstream signaling molecule TGF-binding receptor Smads protein. In this paper, compounds bearing 2(1H)-quinolone scaffold were designed and their anti-fibrosis effects were evaluated. Of these compounds, 20f was identified as the most active one and could inhibit TGF-β-induced collagen deposition of NRK-49F cells and mouse fibroblasts migration with comparable activity and lower cytotoxicity than nintedanib in vitro. Further mechanism studies indicated that 20f reduced the expression of fibrogenic phenotypic protein α-SMA and collagen Ⅰ by inhibiting the TGF-β/Smad dependent pathways and ERK1/2 and p38 pathways. Moreover, compared with the nintedanib, 20f (100 mg/kg/day, p.o) more effectively alleviated collagen deposition in lung tissue and delayed the destruction of lung tissue structure both in bleomycin-induced prevention and treatment mice pulmonary fibrosis models. The immunohistochemical experiments further showed that 20f could block the expression level of phosphorylated Smad3 in the lung tissue cells, which resulted in its anti-fibrosis effects in vivo. In addition, 20f demonstrated good bioavailability (F = 41.55% vs 12%, compare with nintedanib) and an appropriate elimination half-life (T_1/2 = 3.5 h), suggesting that 20f may be a potential drug candidate for the treatment of pulmonary fibrosis.

Full text of DOI:10.1016/j.ejmech.2020.112259

Journal Pre-proof
Design, synthesis and discovery of 2(1H)-quinolone derivatives for the treatment of
pulmonary fibrosis through inhibition of TGF-β/smad dependent and independent
pathway
Linlin Xue, Dexin Deng, Shoujun Zheng, Minghai Tang, Zhuang Yang, Heying Pei,
Yong Chen, Tao Yang, Kongjun Liu, Haoyu Ye, Lijuan Chen
PII:
S0223-5234(20)30226-9
DOI:
https://doi.org/10.1016/j.ejmech.2020.112259
EJMECH 112259
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 21 January 2020
Revised Date: 18 March 2020
Accepted Date: 18 March 2020
Please cite this article as: L. Xue, D. Deng, S. Zheng, M. Tang, Z. Yang, H. Pei, Y. Chen, T. Yang, K.
Liu, H. Ye, L. Chen, Design, synthesis and discovery of 2(1H)-quinolone derivatives for the treatment of
pulmonary fibrosis through inhibition of TGF-β/smad dependent and independent pathway, European
Journal of Medicinal Chemistry (2020), doi: https://doi.org/10.1016/j.ejmech.2020.112259.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Masson SAS.

Graphical abstract

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
Design, synthesis and discovery of 2(1H)-quinolone derivatives for
the treatment of pulmonary fibrosis through inhibition of
TGF-β/Smad dependent and independent pathway
1
1
Linlin Xue , Dexin Deng , Shoujun Zheng, Minghai Tang, Zhuang Yang，Heying Pei,
*
*
Yong Chen, Tao Yang, Kongjun Liu, Haoyu Ye , Lijuan Chen
State Key Laboratory of Biotherapy and Cancer Center, West China Hospital, Sichuan
University and Collaborative Innovation Center, Chengdu, 610041, China
* Corresponding authors. Lab of Natural Product Drugs, Cancer Center, West China
Medical School, West China Hospital, Sichuan University, Chengdu 610041, PR
China.
1
1
1
1
1
1
1
1
1
1
2
2
2
E-mail addresses: haoyu_ye@scu.edu.cn (H. Ye), chenlijuan125@163.com (L. Chen).
1
Both authors contributed equally to this study.
ABSTRACT:
Idiopathic pulmonary fibrosis (IPF) is a progressive, life-threatening and
interstitial lung disease with the median survival of only 3 to 5 years. However, due to
the unclear etiology and problems in accurate diagnosis, up to now only two drugs
were approved by FDA for the treatment of IPF and their outcome responses are
limited. Numerous studies have shown that TGF-β is the most important cytokine in
the development of pulmonary fibrosis and plays a role through its downstream

2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
signaling molecule TGF-binding receptor Smads protein. In this paper, compounds
bearing 2(1H)-quinolone scaffold were designed and their anti-fibrosis effects were
evaluated. Of these compounds, 20f was identified as the most active one and could
inhibit TGF-β-induced collagen deposition of NRK-49F cells and mouse fibroblasts
migration with comparable activity and lower cytotoxicity than nintedanib in vitro.
Further mechanism studies indicated that 20f reduced the expression of fibrogenic
phenotypic protein α-SMA and collagen Ⅰ by inhibiting the TGF-β/Smad dependent
pathways and ERK1/2 and p38 pathways. Moreover, compared with the nintedanib,
20f (100 mg/kg/day, p.o) more effectively alleviated collagen deposition in lung tissue
and delayed the destruction of lung tissue structure both in bleomycin-induced
prevention and treatment mice pulmonary fibrosis models. The immunohistochemical
experiments further showed that 20f could block the expression level of
phosphorylated Smad3 in the lung tissue cells, which resulted in its anti-fibrosis
effects in vivo. In addition, 20f demonstrated good bioavailability (F = 41.55% vs
12%, compare with nintedanib) and an appropriate elimination half-life (T1/2 = 3.5 h),
suggesting that 20f may be a potential drug candidate for the treatment of pulmonary
fibrosis.
Keywords:
Pulmonary fibrosis; Collagen accumulation; Bioisosteres; TGF-β/Smad pathway;
Anti-fibrosis effects.

4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
6
6
6
6
6
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
1. Introduction
Idiopathic pulmonary fibrosis (IPF) is the most common idiopathic interstitial
pneumonia [1]. Like other interstitial lung diseases, the characteristic of idiopathic
pulmonary fibrosis is progressive scarring or fibrosis that is distributed and deposited
between the interstitial spaces of the lungs [2-3], hindering normal gas exchange and
decreased lung capacity and leading to clinical symptoms such as unexplained
exertional dyspnea, chronic dry cough, or Velcro like crackles on examination with
the median survival time of three to five years from diagnosis [4]. Although the
pathogenesis of IPF has not been elucidated, it is generally believed that early stage is
pneumonia and lung injury, and the late stage is the deposition of extracellular matrix
(collagen fiber).
The treatment of pulmonary fibrosis progresses really slowly due to the unclear
etiology and problems in accurate diagnosis. Except lung tissue transplantation,
considering the characteristics and possible pathogenesis of IPF, drugs with various
mechanisms had also been applied to the clinical treatment, such as antifibrotic agents
(nintedanib, 1, and pirfenidone 2), antioxidants (N-Acetylcysteine, 3 [5]),
corticosteroids (Prednisone, 4 [6]), immunomodulatory cytokines (Interferon
gamma-1β [7]), anticoagulants (Warfarin, 5 [8]), anti-gastroesophageal reflux agent
(Proton pump inhibitors and antacid medication [9]), and anti-pulmonary
hypertension drug (Sildenafil, 6 [10]) et al. (Figure 1). Up to now, only nintedanib (1)
and pirfenidone (2) were approved by FDA for the treatment of IPF. The former was a
multi-tyrosine kinase inhibitor mainly targeting VEGFR, FGFR and PDGFR, and

6
6
6
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
could consistently and significantly slow disease progression by reducing the annual
rate of decline in forced vital capacity by approximately 50% versus placebo in IPF
patients [11-13]. However, due to its poor oral bioavailability and metabolic
instability [14], as well as side effects caused by off-target effects, nintedanib in
clinical application also has limitations. Similar things happened to pirfenidone, an
inhibitor for TGF-β production and TGF-β stimulated collagen production. Although
clinical trials have demonstrated that it could reduce the decline in lung function in
IPF patients, there are concerns about the adverse drug reactions. Post-marketing
surveillance in Japan revealed that 24.3% of patients discontinued pirfenidone therapy
because of adverse drug reactions [15]. Hence, efficient and also safety drugs were
urgently needed for IPF treatment.
In recent years, the study of multiple cytokines in the lung has made important
progress in the study of the mechanism of pulmonary fibrosis. Among the cytokines
involved in pulmonary fibrosis, transforming growth factor TGF-β is the most deeply
studied and plays an important role in the onset and progression of the disease [16].
TGF-β1 plays an important role in regulation of inflammatory processes, ECM
production, and stem cell differentiation as well as T-cell regulation and
differentiation. Therefore, it is considered to be directly related to fibrosis. Inhibition
of the binding of TGF-β to its receptor and the function of related Smad proteins
become a critical strategy for anti-fibrosis recently. Except pirfenidone, the safety and
scientific validity of Epigallocatechin-3-gallate (EGCG, 7), a fibroblast specific
inhibitor of LOXL2 and TGF-β1 signaling, has now been evaluated in patients with

8
9
9
9
9
9
9
9
0
1
2
3
4
5
pulmonary fibrosis in the phase III clinical study in 2019 (NCT03928847).
Isoalatolactone derivative 8 exhibited promising efficacy in a bleomycin-induced
pulmonary fibrosis mice model through inhibition of TGF-β/Smad3 pathway [17].
Cynnamoyl anthranilate analogue FT011 (9) inhibited TGF-β1 and PDGF-BB
induced collagen production in vitro, and subsequent research showed that it also
reduced the fibrotic scar in the diseased kidney [18].
9
9
6
7
Figure 1. Compounds reported for the treatment of idiopathic pulmonary fibrosis.
9
8

9
9
In our previous reports, compound 10, a 4-substituated coumarins derivative,
was found to be a promising, potential, orally active candidate for the treatment of
fibrotic disease by its inhibition of TGF-β/Smad3 pathway and anti-inflammation
efficacy [19]. However, the anticoagulant effect of coumarin scaffold may increase
patient mortality during the treatment of pulmonary fibrosis, such as warfarin (5) [20].
Herein, considering that quinolone scaffold is widely used in the pharmaceutical field
[21, 22], as well as the poor druggability of nintedanib due to its indolinone scaffold,
bioisosteres and scaffold hopping [23] strategies were applied to design a new series
of compounds bearing 2(1H)-quinolone scaffold (Figure 2). Hydrophobic
heterocycles were also introduced to the terminal of 2(1H)-quinolone scaffold to
explore its effects on biological activity. Finally, both in vivo and in vitro experiments
were conducted to verify the anti-fibrotic effect and toxicity of these compounds, as
well as the underlying mechanism.
1
1
1
1
1
1
1
1
1
1
1
1
1
1
00
01
02
03
04
05
06
07
08
09
10
11
12
13

1
1
14
15
1
1
1
1
1
1
1
1
1
1
1
1
16
17
18
19
20
21
22
23
24
25
26
27
Figure 2. Design a novel series of compounds with anti-fibrotic effects by bioisosteres and
scaffold hopping.
2. Chemistry
Considering the poor druggability of nintedanib and disadvantages of coumarin
scaffold, we designed a series of compounds bearing 2(H)-quinolinone scaffold with
different terminal substituent to improve its druggability. To be specific, we
considered the influence of several factors on its in vitro activity: different
hydrophobic groups at the terminal, different linker atoms between 2(H)-quinolinone
and benzene ring, electron-withdrawing substituents in benzene ring, and different
substitution locations in 2(H)-quinolinone.
4-Hydroxy-2(H)-quinolinone
(1)
was
halogenated
to
obtain
2,

1
1
1
1
1
1
1
1
28
29
30
31
32
33
34
35
4-dibromoquinoline (2), and 4-bromoquinolin-2(1H)-one (3) was got by hydrolysis
reaction. The important intermediate 4 comprised NH-linker was synthesized though
Buchwald–Hartwig
reaction
catalyzed
by
Xantphos
and
Pd (dba) .
2 3
T-butyloxycarboryl was removed under acidic conditions, and then the exposed amino
group was reacted with chloroacetyl chloride to obtain important intermediate 6.
Specific methods were shown in Scheme 1.
Scheme 1. General procedure for the synthesis of compounds 7a-7l.
1
1
36
37
Reagents and conditions: (a) TBAB, P O , toluene, 100Ⅰ , 6 h, 42%; (b) hydrobromic acid,
2
5
1
1
1
1
1
1
38
39
40
41
42
43
1,4-dioxane, 90Ⅰ , 4 h, 75%; (c) 4-(tert-butoxycarbonylamino)aniline, Xantphos, Pd (dba) ,
2 3
t-BuOK, dioxane, 130Ⅰ , 24 h, 66%; (d) trifluoroacetic acid, rt, overnight, 92%; (e) chloroacetyl
chloride, Et N, DMF, 0-25Ⅰ , 4 h, 95%; (f) RNH , Et N, DMF, rt, overnight, 45-84%.
3
2
3
It was simpler to synthesize compound 12a-12l when the linker atom was sulfur
atom. 4-Hydroxy-2(H)-quinolinone (1) was reacted with POCl and heated to 100Ⅰ to
3

1
1
1
1
1
1
1
1
1
1
1
44
45
46
47
48
49
50
51
52
53
54
get 2,4-dichloroquinoline (8), and 4-chloroquinolin-2(1H)-one (9) was obtained by
hydrolysis reaction as the same method for 4-bromoquinolin-2(1H)-one (3). Lower
reactivity intermediate 9 was stirred with K CO and 4-aminobenzenethiol in DMF at
2
3
130Ⅰ to obtain 4-((4-aminophenyl)thio)quinolin-2(1H)-one (10). Compound 16a-16o
with oxygen atom as the linker atom also started from 4-Hydroxy-2(H)-quinolinone
(1), which was then stirred with 1-fluoro-4-nitrobenzene and K CO in DMF at 100Ⅰ
2
3
to get 4-(4-nitrophenoxy) quinolin-2(1H)-one (13). Under the catalysis of iron powder
and concentrated hydrochloric acid, nitro group was reduced to get intermediate 14
for subsequent reaction. Specific methods were shown in Scheme 2.
Scheme 2. General procedure for the synthesis of compounds 12a-12l and 16a-16o.
1
1
55
56
Reagents and conditions: (a) POCl , 100Ⅰ , 6 h, 65%; (b) hydrochloric acid, 1,4-dioxane, 90Ⅰ , 4 h,
3
1
1
1
1
1
57
58
59
60
61
73%; (c) 4-aminothiopenenol, K CO , DMF, 130Ⅰ , 6 h, 58%; (d) chloroacetyl chloride, Et N,
2
3
3
DMF, 0Ⅰ , 4 h, 0-25Ⅰ ; (e) RNH , Et N, DMF, rt, overnight, 62-90%; (f) 4-fluoronitrobenzene,
2
3
K CO , DMF, 100Ⅰ , 6 h, 72%; (g) Fe, HCl, MeOH / H O = 9/1, 85Ⅰ , 4 h, 90%.
2
3
2
The introduction of fluorine atom to the drug structure has been reported to

1
1
1
1
1
1
1
1
1
1
1
1
1
1
62
63
64
65
66
67
68
69
70
71
72
73
74
75
beneficially change its activities. A series of compounds containing fluorine atom on
the benzene ring were synthesized to verify its biological activity in vitro. Due to the
strong electrophilicity of 3,4-difluoronitrobenzene, it was stirred with
4-Hydroxy-2(H)-quinolinone (1) and K CO in DMF at room temperature without
2
3
heating to obtain 4-(2-fluoro-4-nitrophenoxy)quinolin-2(1H)-one (17). Subsequent
route to synthesize compounds 20a-20o (Scheme 3) was similar to that of compound
16a-16o.
In order to explore the effect of the substituent position of the
2-hydroxyquinoline skeleton on its activity, we then tried to connect the substituents
at the C-6 and C-7 positions and synthesized compounds 21a-21c and 22a-22c.
Specific and subsequent routes to synthesize these compounds are shown in
Supporting Information (Scheme S1).
Scheme 3. General procedure for the synthesis of compounds 20a-20o, 21a-12c, and 22a-22c.
H
H
NO2
NH2
N
N
a
b
Cl
R
c
d
A
OH
O
O
O
A
O
A
O
A
O
A
F
F
F
F
A = 2-quinolinone 17a: 4-substituted-2-quinolinone 18a: 4-substituted-2-quinolinone 19a: 4-substituted-2-quinolinone 20a - 20o: 4-substituted-2-quinolinone
1
1
7b: 6-substituted-2-quinolinone 18b: 6-substituted-2-quinolinone 19b: 6-substituted-2-quinolinone 21a - 21c: 6-substituted-2-quinolinone
7c: 7-substituted-2-quinolinone 18c: 7-substituted-2-quinolinone 19c: 7-substituted-2-quinolinone 22a - 22c: 7-substituted-2-quinolinone
N
O
1
1
76
77
H
Reagents and conditions: (a) 3,4-fifluoronitrobenzene, K CO , DMF, rt, overnight; (b) Fe, HCl,
2
3
1
1
1
78
79
80
MeOH/H O = 9/1, 85Ⅰ , 4 h; (c) chloroacetyl chloride, Et N, DMF, 0Ⅰ , 4 h, 95%; (d) RNH , Et N,
2
3
2
3
DMF, rt, overnight.
1
1
81 3. Results and discussion
82
3.1 Anti-fibrotic activity of compounds on NRK-49F cells in vitro.

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
Fibrosis is characterized by excessive collagen deposition between the cells of
diseased tissues. Thus, the in vitro screening cell model was established to determine
the deposited collagen amount between cells. According to our previous published
and other reports [19, 24, 25], TGF-β-induced NRK-49F cells (rat fibroblast cells)
produce a large amount of intercellular collagen deposition that is similar to the
characteristic of fibrosis. Thus, we adopted it as an effective and convenient in vitro
screening model for anti-fibrotic evaluation. The inhibition rates of collagen
deposition for all synthesized compounds were subsequently tested on TGF-β-induced
NRK-49F cells at a concentration of 10 μM. Simultaneously, the survival rates (SR)
of NRK-49F cells were also tested by MTT assay to verify their in vitro toxicities.
Compounds 7a-7l with different terminal substituent were synthesized initially
under the inspiring of nintedanib, which own same nitrogen atom as linker atom. As
shown in Table 1, most of compounds had weaker collagen deposition inhibition
effects in compared with nintedanib. However, we found that nintedanib exhibited
significantly toxicity to NRK-49F cells judging from low survival rate of NRK-49F
cells (40% survival rate (SR)). Interestingly, 7i showed relatively strong collagen
deposition inhibition effect and low toxicity to NRK-49F cells among them (73%
inhibition rate (IR)). Whether the Boc group with a large steric hindrance at the
terminal of its structure resulted in its activity improvement still needed further
confirmation.
Table 1. Collagen accumulation IR and cell SR of 7a-7l.

2
05
Cpd
R
IR (%)
SR (%)
Cpd
7g
7h
7i
R
IR (%)
SR (%)
7
7
a
59.91 ± 7.61
24.42 ± 2.68
40.86 ± 2.99
20.08 ± 6.43
38.93 ± 1.31
72.49 ± 3.82
75.40 ± 1.12
81.91 ± 3.66
82.55 ± 0.33
78.04 ± 3.33
38.97 ± 0.31
50.20 ± 2.12
73.34 ± 2.85
49.99 ± 1.47
44.24 ± 6.94
72.55 ± 3.47
83.49 ± 1.65
67.52 ± 4.96
83.55 ± 6.48
82.58 ± 2.22
b
7
c
7
d
7j
7
e
7k
7
f
45.95 ± 5.34
64.78 ± 2.66
7l
28.71 ± 5.19
95.47 ± 2.65
83.71 ± 0.94
40.02 ± 2.46
nintedanib
2
2
2
2
2
2
2
2
06
07
08
09
10
11
12
13
Inhibitory effects against TGF-β-induced total collagen accumulation in NRK-49F cells at a
concentration of 10 μM. Cell survival rate is calculated by MTT assay. The results are the means ±
SD of at least three independent experiments.
Bioisosteric replacement and scaffold hopping are two techniques widely applied
in structural optimization of lead compounds to reduce toxicity and improve activity.
To be specific, when the linker atoms nitrogen atom in compounds 7a-7l was
substituted by sulfur atom or oxygen atom, the effects of linker atoms on biological
activity were evaluated and discussed. The results were listed in Table 2.
2
2
14
15

2
16
Table 2. Collagen accumulation IR and cell SR of 12a-12l and 16a-16o.
2
17
Cpd
R
IR (%)
SR (%)
Cpd
16a
16b
R
IR (%)
SR (%)
1
2a
2b
80.88 ± 2.39
28.37 ± 7.65
87.05 ± 5.13
93.46 ± 2.63
68.05 ± 6.38
27.42 ± 5.41
82.21 ± 2.74
78.46 ± 1.95
1
1
2c
53.38 ± 4.78
77.44 ± 3.37
16c
59.27 ± 3.89
95.74 ± 4.54
1
2d
28.59 ± 8.17
39.97 ± 7.03
115.78 ± 4.55
86.69 ± 3.93
16d
16e
31.52 ± 7.86
43.66 ± 4.42
79.92 ± 6.73
80.18 ± 1.33
1
2e
1
2f
56.57 ± 0.88
77.95 ± 1.93
16f
78.33 ± 3.57
92.17 ± 2.11
1
2g
2h
39.38 ± 3.02
51.14 ± 2.05
72.37 ± 6.87
52.97 ± 2.21
58.62 ± 9.83
51.09 ± 7.84
58.11 ± 2.23
46.02 ± 5.45
101.31 ± 4.3
80.31 ± 4.37
71.25 ± 4.26
83.92 ± 3.41
85.49 ± 5.88
87.69 ± 8.85
95.14 ± 1.09
96.04 ± 3.82
16g
50.02 ± 11.41
43.92 ± 4.78
76.24 ± 4.15
73.23 ± 5.47
86.07 ± 2.02
25.36 ± 3.25
41.73 ± 8.79
95.47 ± 2.65
96.28 ± 5.24
81.22 ± 2.61
75.60 ± 4.72
92.67 ± 2.31
96.26 ± 2.44
89.24 ± 2.64
91.69 ± 3.34
40.02 ± 2.46
1
16h
1
2i
2j
16i
1
16j
1
2k
16k
1
2l
16l
1
6m
16o
1
6n
nintedanib

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Inhibitory effects against TGF-β-induced total collagen accumulation in NRK-49F cells at a
concentration of 10 μM. Cell survival rate is calculated by MTT assay. The results are the means ±
SD of at least three independent experiments.
As shown in Table 2, compared with 7a-7l, the collagen inhibition rate and cell
survival rate among 12a-12l and 16m-16o have been improved to some extent when
the NH-linker was replaced by S-linker or O-linker, respectively. Among them,
compound 12a in S-linker series and 16k in O-linker series showed good collagen
inhibition rate (80% and 86%) and low toxicity (13% and 4%). Compounds 7i, 12i
and 16i with the terminal Boc group demonstrated relatively acceptable in vitro
anti-fibrotic activities (73%, 72% and 76%). However, when the terminal Boc group
was removed (16m), it exhibited poor in vitro activity (58%), suggesting that a
hydrophobic group with a large spatial position at the terminal might be beneficial for
anti-fibrotic activity in vitro. In order to verify our suspicions, compound 16f and 16j
with similar steric hindrance groups and less susceptible to metabolism were
subsequently synthesized. Results showed that hydrophobic group with a large spatial
position at the terminal did play crucial parts in collagen accumulation inhibition.
Table 3. Collagen IR and cell SR of 20a-22o, 21a-21c, and 22a-22c.
2
37

Cpd
R
IR (%)
SR (%)
96.43 ± 2.47
87.18 ± 3.07
Cpd
20l
R
IR (%)
SR (%)
2
0a
0b
49.49 ± 6.52
38.27 ± 4.76
50.57 ± 5.93
60.27 ± 2.16
99.35 ± 1.12
93.31 ± 2.29
2
20m
2
0c
47.28 ± 7.12
90.77 ± 1.71
20n
53.92 ± 7.32
90.8 ± 1.44
2
0d
45.94 ± 2.55
73.87 ± 5.12
104.93 ± 3.04
98.95 ± 0.82
20o
21a
49.07 ± 4.19
34.57±3.54
89.61 ± 7.28
95.19 ± 3.13
2
0e
2
0f
87.53 ± 3.05
97.52 ± 2.26
21b
57.43±3.59
85.62 ± 2.31
2
0g
0h
42.73 ± 6.36
60.74 ± 6.01
89.13 ± 3.97
78.04 ± 3.94
58.81 ± 6.03
108.38 ± 3.04
98.59 ± 3.56
86.18 ± 5.65
93.11 ± 4.78
106.12 ± 2.32
21c
22a
22b
22c
61.53±3.21
56.59±2.67
76.37±2.78
67.72±5.45
95.47 ± 2.65
83.32 ± 4.24
98.23 ± 3.25
89.65 ± 2.36
89.77 ± 3.41
40.02 ± 2.46
2
2
0i
0j
2
2
0k
nintedanib
2
2
2
2
2
2
2
2
2
38
39
40
41
42
43
44
45
46
Inhibitory effects against TGF-β-induced total collagen accumulation in NRK-49F cells at a
concentration of 10 μM. Cell survival rate is calculated by MTT assay. The results are the means ±
SD of at least three independent experiments.
Next, compounds with fluorine atom substitution on benzene ring were designed
for both biological activity bioavailability improvements. As shown in Table 3, most
compounds with fluorine substitution demonstrated better anti-fibrotic effects and
lower cytotoxicity than their corresponding compounds 16a-16o. 20m possessed
weaker anti-fibrotic activity and less cytotoxicity than compound 20i owning Boc

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
group (similar to compounds 16i and 16m).
Finally, 21a-21c and 22a-22c were synthesized to investigate the anti-fibrosis
effect of substitution at different positions on the 2(1H)-quinolone. 21a-21c owning
substitution at C-6 position showed significant decrease in anti-fibrotic activity in
vitro, which is similar to compounds 22a-22c substituted at C-7 position. Compounds
20f, 20i and 20j with excellent bioactivity in vitro and low toxicities were further
selected as candidate compounds for further in vitro experiments.
In general, the initial structure-activity relationship could be concluded as
follows: firstly, the linker atom affects the cytotoxicity of these compounds.
Compounds with oxygen atom as the linker demonstrated the lowest cytotoxicity than
those with nitrogen atom and sulfur atom. Secondly, hydrophobic groups with larger
steric hindrance at the end of the piperazine ring contributed greatly to the
anti-fibrotic activities but caused greater cytotoxicities such as 7i, 12i, and 16i. Then,
the introduction of fluorine atom on the benzene ring was beneficial to enhance
anti-fibrotic activity and reduce cytotoxicity to some extent. Finally, substitution at
C-6 and C-7 position decreased anti-fibrotic activity and showed a little cytotoxicity
effect to NRK-49F cells. However, compounds 12a with tetrahydropyrrole and 16k
with diethylamine due to some certain exceptions also exhibited excellent anti-fibrotic
activity in vitro, which was inconsistent with the basic structure-activity relationship.
Taken together, five compounds of 12a, 16k, 20f, 20i, and 20j with potential
anti-fibrosis activities were selected for further determination of IC50 values with
potential anti-fibrosis activities. As exhibited in Table 4, those five compounds

2
2
2
2
2
69
70
71
72
73
showed an order of magnitude of inhibitory activity with nintedanib against
TGF-β-induced total collagen accumulation in NRK-49F cells with IC50 range
between 3.89 and 6.12 μM.
Table 4. IC values against TGF-β-induced total collagen accumulation in NRK-49F cells.
5
0
Entry
IC (μM)
Entry
16k
IC (μM)
5
0
50
1
2a
5.33 ± 0.35
3.89 ± 0.46
4.47 ± 0.52
5.27 ± 0.42
6.12 ± 0.32
1.10 ± 0.13
2
2
0f
0j
20i
nintedanib
2
2
74
75
The results are the means ± SD of at least three independent experiments.
2
2
2
2
2
2
2
76
77
78
79
80
81
82
Collagen fibers between cells were straightforward visualized through Sirius Red
staining. It was found that 12a, 16k, 20i, 20j and 20f could significantly reduce the
production of filamentous collagen (Figure 3). Moreover, it’s also worth noting that
the cell number of TGF-β-induced NRK-49F cells left after treatment of these
compounds was much higher than that of nintedanib, which was consistent with MTT
results, indicating that 12a, 16k, 20i, 20j and 20f were safer than nintedanib.

2
2
83
84
Figure 3. Picro-Sirius Red (PSR) staining for the total collagen accumulation induced by TGF-β
2
2
2
2
2
2
2
2
2
2
2
85
86
87
88
89
90
91
92
93
94
95
in NRK-49F cells.
3.2 12a, 16k, 20f, 20i and 20j inhibited TGF-β-induced fibroblasts migration
Fibroblasts migration to the fibrotic lessions stimulated by concomitant
cytokines (like TGF-β1) plays an important role in pulmonary fibrosis [26], so
inhibiting fibroblasts migration can delay the process of pulmonary fibrosis.
Accordingly, the effects of selected compounds on cell migration was accessed
by wound healing assays. The results indicated that 12a, 16k, 20f, 20i, 20j and
nintedanib could inhibit TGF-β-induced migration of mouse fibroblast L929 cells at
different level with a concentration of 10 μΜ, while 20f exhibited the best
anti-migration capacity (Figure 4). Hence, 20f was finally chosen as further

2
96
antifibrotic evaluation and mechanism investigation both in vitro and in vivo.
2
2
97
98
Figure 4. Wound healing assay were performed to access the inhibition effects of fibroblast
migration by 12a, 16k, 20f, 20i, 20j and nintedanib. L929 cells were treated with/without TGF-β
2
99

3
3
3
3
3
3
3
3
3
3
3
3
00
01
02
03
04
05
06
07
08
09
10
11
(5 ng/mL) and compounds (10 μΜ). Data were collected after 12 and 24 hours.
3.3 20f inhibited protein expression of collagen I and α-SMA in TGF-β-induced
NRK-49F Cells.
α-Smooth muscle actin (α-SMA) in myofibroblasts [27] and excessive collagen I
deposition [28] in extracellular matrix were hallmarks of fibrosis. Therefore, the
abilities of 20f to suppress protein expression of α-SMA and collagen I were
investigated in vitro. As shown in Figure 5, the protein expression levels of α-SMA
and collagen I were obviously over-expressed in TGF-β-induced NRK-49F cells,
whereas treatment with 20f or nintedanib significantly inhibited their expressions,
suggesting that both 20f and nintedanib inhibited α-SMA expression and collagen
accumulation in the response to TGF-β stimulation.
3
3
12
13
Figure 5. Protein expression levels of Collagen I and α-SMA were probed through western blot.
NRK-49F cells were treated with/without TGF-β (5 ng/mL) and compound 20f (20 and 40 μΜ)
3
14

3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
for 24 h. β-actin was used as a loading control.
3.4 20f exhibited anti-fibrotic activity by inhibiting TGF-β/Smad2/3-dependent
and independent pathways.
Since activation of Smad pathway is the primary downstream signaling pathway
of TGF-β and Smad proteins have been implicated in bleomycin-induced lung fibrosis
[29], the expression of total and phosphorylated Smad2 and Smad3 were further
detected by immunoblot to test the hypothesis that 20f might block the primary step of
TGF-β signaling via the regulation of Smads.
As shown in Figure 6A, TGF-β stimulation significantly increased the
phosphorylation levels of Smad2 and Smad3, whereas treatment with 20f significantly
reduced this phosphorylation in a dose dependent manner. However total Smad2 and
Smad3 proteins were constitutively expressed and were not affected by TGF-β or 20f
treatment.
Smad3-independent pathway including the ERK1/2 [30] and p38 MAP kinase
[31] has also been reported in TGF-β-induced fibrosis [32]. Activated ERK1/2 and
p38 in turn conveys the signal to the Smad2/3 via phosphorylation and active the
Smad signaling pathway. Therefore, the TGF-β/Smad2/3-independent pathway was
further explored. NRK-49F cells were cultured with or without TGF-β in the presence
or absence of 20f for indicated times. As shown in Figure 6B, TGF-β stimulation
obviously increased the phosphorylation level of ERK1/2 and p38, whereas
cotreatment with 20f significantly inhibited their phosphorylation levels of ERK1/2
and p38. Both total p38 and ERK1/2 were constitutively expressed and were not

3
3
3
3
3
37
38
39
40
41
affected by TGF-β or 20f treatment.
On the basis of these observations, it could be concluded that 20f ameliorated
fibrosis by inhibiting both TGF/Smad2/3-dependent and independent pathways,
including Smad2, Smad3, ERK and p38 phosphorylation.
3
3
42
43
Figure 6. (A) Effects of 20f on TGF-β-induced phosphorylation of Smad2 and Smad3 in
NRK-49F cells. NRK-49F cells were treated with/without 20f (20 and 40 μΜ) for 1 and 2 hours
and then treated with TGF-β (5 ng/mL) for 1h. (B) Effects of 20f on TGF-β-induced
phosphorylation protein expression of p38 and ERK1/2 in NRK-49F cells. NRK-49F cells were
treated with/without 20f (20 and 40 μΜ) for 1 and 2 hours and then treated with TGF-β (5 ng/mL)
for1h. GAPDH was used as a loading control.
3
3
3
3
3
3
3
3
3
3
3
44
45
46
47
48
49
50
51
52
53
54
3.5 Pharmacokinetic experiment of 20f.
The plasma concentration−time curves for 20f after a single dose in rats were
shown in Figure S1 (Supporting Information). The plasma concentrations of 20f
rapidly reached peak, and gradually declined after oral administration. As shown in
Table 5, the oral bioavailability of 20f was determined to be 41.55% (n = 5), much

3
3
55
56
better than nintedanib (approximately 12% in rats [20]).
Table 5. Pharmacokinetic parameters for 20f in SD rats.
2
0f
Compound
Intravenous injection
Oral administration
dose (mg/kg)
tmax (h)
5
5
0.08
4.33
t1/2 (h)
4.39
3.5
AUC0-t (µg/L*h)
F (%)
13149.76
5463.86
41.55
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
3.6 Effects of 20f on Bleomycin-induced pulmonary fibrosis model.
To further verify anti-fibrotic potency of 20f in vivo, two bleomycin induced
lung fibrosis models (one for prevention model and the other one for treatment model)
were employed in our study. Mice challenged with BLM (3 U/kg) at day 0 were
treated with 20f (50 or 100 mg/kg) and nintedanib (50 mg/kg) at day 1 for prevention
model and the whole administration process last for 30 days. For treatment model,
every experimental setting was the same except that both compounds were
administrated only from day 8 to day 21 (Figure 7A). Survival rates, hydroxyproline
(an indicator of collagen deposition) level, histologic analysis (Masson’s trichrome
staining, hematoxylin and eosin (H&E) staining and immunocytochemistry) were
further performed to evaluate anti-fibrosis effects of 20f.
As shown in Figure 7B, in prevention model, 20f (50 or 100 mg/kg)
demonstrated comparable survival rate with nintedanib (50 mg/kg). However, in
treatment model, 20f (50 mg/kg) showed better survival rate than nintedanib. BLM

3
3
3
3
3
3
3
3
3
72
73
74
75
76
77
78
79
80
induction resulted in a large amount of collagen to deposit in lung tissue in model
group, which was consistent with the hydroxyproline result (Figure 7C and 7D).
Both 20f and nintedanib could inhibit collagen deposition and reduce hydroxyproline
content in lung tissue in prevention model. However, when treatment started at day 8
(treatment model), there was no significant difference in the reduction of collagen
deposition in nintedanib group in comparison with model group. In contrast, 20f still
showed a significant blocking effect on collagen accumulation, particularly at high
doses (100 mg/kg) (Figure 7D).

3
3
81
82
Figure 7. The anti-fibrotic potencies of 20f in vivo. C57BL/6 male mice were intratracheal
instillated with BLM (3 U/kg) to induce pulmonary fibrosis, and the lungs were harvested at the
end of experiment for the following analyses. (A) Dosing schedule of prevention model (day 1-30)
and treatment model (day 8-21). (B) Survival curves of 20f and nintedanib for pulmonary fibrosis.
(C) Hydroxyproline contents in lung tissues of each group were measured as described in the
methods section. (D) Representative images demonstrating masson’s trichrome of lung tissues
3
3
3
3
3
83
84
85
86
87

3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
from experimental groups as indicated in Experiment section. Scale bars, 100 µm. Data were
presented as means (SD of the group. *p < 0.05, **p < 0.01, compared with the BLM treatment.
BLM, bleomycin; Nin, nintedanib.)
Both the Ashcroft scores and H&E staining were also used to evaluate the
anti-fibrosis effects of 20f. Results demonstrated that the lung tissues were fiercely
damaged in mice instilled with BLM (Figure 8A and 8B), whereas administration
with 20f and nintedanib notably protected the alveolar tissue structure (red arrow) and
ameliorated the infiltration of inflammatory cells (yellow arrow) in prevention model
(Figure 8A, upper panel). However, in the treatment model, delayed treatment with
nintedanib at day 8 failed to ameliorate fibrosis, 20f (100 mg/kg) still exhibited the
ability to protect structure of lungs (Figure 8A, lower panel). These findings
implicated that 20f showed potentially promising therapeutic effects in both
prevention and treatment models of pulmonary fibrosis.
To further validate the mechanism of 20f in vivo, the phosphorylation level of
Smad3 in lungs tissues was tested. As depicted in Figure 8C, BLM-induced mice
elevated level of Smad3 phosphorylation in lung tissues. Relative to nintedanib, oral
administration of 20f could more effectively suppress Smad3 phosphorylation level,
which was in accordance with the immunoblot results in Figure 6.

4
4
09
10
Figure 8. (A) Representative images demonstrating hematoxylin and eosin (H&E) staining of
lung tissues from experimental groups as indicated in experiment section. (B) Ashcroft scores of
lung tissues from experimental groups as indicated in experiment section. (C) Representative
images demonstrating p-Smad3 staining of lung tissues from experimental groups as indicated in
experiment section. Scale bars, 100 µm. Data were presented as means (SD of the group. *p <
0.05, **p < 0.01, compared with the BLM treatment. BLM, bleomycin. Nin, nintedanib.)
4
4
4
4
4
4
4
4
4
4
4
4
11
12
13
14
15
16
17
18
19
20
21
22
Taken together, 20f exhibited beneficial therapeutic effects, specifically in
survival rate improvement, collagen deposition decrease in lung tissue, as well as the
lung tissue protection. Further immunohistochemical experiments showed that 20f
also inhibited phosphorylated expression level of Smad3, which was downstream of
TGF-β pathway, and eventually contributed to its anti-fibrosis effect.

4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
4. Conclusion
In this study, a series of compounds bearing 2(1H)-quinolone scaffold inspired
by bioisosteres and scaffold hopping strategies were designed and synthesized, and
then their anti-fibrotic effects were evaluated in vitro. 20f was further selected to
perform the mechanism studies due to its excellent collagen deposition inhibition in
NRK-49F cells, low cytotoxicity, as well as decent anti-migration activity in L929
cells. Mechanism studies showed that 20f significantly suppress the expression of
fibrogenic phenotypic protein (such as α-SMA and collagen Ⅰ ) in vitro by western
blot analysis. Additionally, it also decreased the phosphorylation level of Smad 2/3,
ERK and p38, indicating that TGF-β/Smad signaling pathway played crucial roles for
collagen deposition reduction. Moreover, 20f (50 and 100 mg/kg/day, p.o.) effectively
alleviated collagen deposition in lung tissue and delayed the destruction of lung tissue
structure in both prevention and treatment models. It was worth noting that 20f
demonstrated better survival rate than nintedanib. Subsequent immunohistochemical
experiments further showed that the expression level of phosphorylated Smad3 in the
lung tissue cells significantly declined after treatment with 20f. All these phenomena
and results illustrated 20f effectively alleviated lung damage induced by intratracheal
instillation of bleomycin in mice, and these beneficial effects should attribute to its
inhibition of TGF-β/Smad dependent and in-dependent pathway as well as its low
cytotoxicity. Considering the excellent bioavailability (F = 41.55%) and suitable
eliminated half-life time (T1/2 = 3.5 h), 20f could be a potential drug candidate for the
treatment of IPF.